View Full Version : AMP6-Basic backpack stereo units
Here's a pic of three AMP6-basic backpack stereo units I've built to link in with Londonskate's street skate sound system (http://www.bassfreight.com/)
http://homepage.ntlworld.com/m.brown74/Photos/ASBOMBS/ASBOMBs.JPG (http://homepage.ntlworld.com/m.brown74/Photos/ASBOMBS/)
http://homepage.ntlworld.com/m.brown74/Photos/ASBOMBS/ASBOMBs-lights.JPG (http://homepage.ntlworld.com/m.brown74/Photos/ASBOMBS/)
The complete sound system consists of an 8 Freight cargo bicycle carrying the main amplifiers and speakers, plus a number of backpack repeater units carried by skaters which are linked by radio to the bike. There can be several hundred people taking part in the skates and the bodies tend to absorb the sound, so multiple synchronised units helps to provide a good spread of music throughout the skate.
I'm very pleased with the way the AMP6 works, it's excellent efficiency means we can get a good long run time with just 10 AA Ni-MH cells, and it sounds good. The units weigh only 2.5kg so they are easy to skate with. The speakers are Directed Audio SX-650 (6.5 inch) with neodymium magnets.
This forum has been a tremendously useful source of information, I'm no audio expert but thanks to all the helpful people who post on here I know a lot more now than when I started the project. Thank you all!
QUESTIONS:
I now want to build a more powerful mono backpack unit capable of providing a lot more bass. I'd like to use a 24V battery pack which means using either a bridged AMP4 or an AMP11-LV. Speakers will probably be a single Monacor SP-10A/250NEO (http://www.monacor.de/typo3/index.php?id=75&L=1&act=&act_sub=&artid=5779&spr=EN&typ=u) 8 Ohm 10 inch bass/mid range speaker and a pair of 4 Ohm Directed Audio SX-525 (5.25 inch) speakers (connected in series to give 8 Ohm together) with a simple crossover. Couple of questions:
1. As I'm using 8 ohm load will it be better to use AMP4 or AMP11-LV?
2. Will either of those amplifiers be able to provide the full 100W output when powered by 24V.
3. Assuming it can manage 100W output, what is the average current draw from 24V at full power?
4. Am I being over-optimistic about powering a 10 inch speaker from such a modest power supply? I've chosen it because it's very light for it's size (neodymium magnet), it's efficient (99db 1W/1M), Monacor seem to be a reputable brand, and it's just about affordable.
Any help appreciated!
If you are going to build mono...
Choose AMP11-LV...
AMP4 is already bridged and it can be parrallel.
Basically AMP11-LV is a parralleled AMP4.
With 24VDC ==> 17Vrms ==> 17 / 8ohm = 2.125Arms
2.125Arms x 2.125Arms x 8 ohm = 36.125Wrms
17 / 4 ohm = 4.25Arms
4.25Arms x 4.25Arms x 4 ohm = 72.25Wrms
If you need more power choose a 4ohm speaker driver ...
Thanks, mikechw. I hadn't realised it was that straightforward to work out. If anybody knows of a suitable speaker I'd appreciate any info. It has to have neodymium magnet to save weight and it has to be efficient.
Just out of interest, how do people manage to drive high power bass speakers in 12V car sound systems? Do they just use very low impedance speakers?
In order to have a higher voltage supply in car systems,
they either use a charge pump voltage regulator to boost the voltage or
use an inverter to transform DC into AC, then through transformer to get a suitable DC voltage.
Yes, auto sub-woofers can go down to 1 ohm (dual 2 ohm voice coils run in parallel)
Here's a 4ohm neo woofer (most neos are 8 ohm):
http://www.parts-express.com/pe/showdetl.cfm?Partnumber=264-833
Power handling: 120 watts RMS/240 watts max *VCdia: 2" *Impedance: 4 ohms *Re: 3.20 ohms *Frequency range: 30-800 Hz *Fs: 30 Hz *SPL: 84 dB 1W/1m *Vas: .66 cu. ft. *Qms: 4.26 *Qes: .31 *Qts: .29 *Xmax: 12 mm
You'd also get more power if your ran the direct audio 5.25" speakers at 4 ohms each - not in series.
The AMP6 can handle 4ohms just fine.
You should also look at some of krilli's posts on LiPo batteries in this forum (lighter and more power than AA NiMh).
***It's so easy, everyone should be doing it***
to produe sub bass, especially outside, you have to move a lot of air, and you're going to be pretty disapponted with any backpack sized sub, that 99dB efficency woofer is not going to be that efficent at sub frequencies (less than 100Hz) especially in an undersized box,
I think you're better off just placing a few folded horn subs in the middle and skate around them like these
http://www.billfitzmaurice.net/T39.html
at 14" width with a 1.8Kg Eminence basslite S2010
you could conceivably make a backpack that comes in at 10-15Kg, pity the bastard who has to skate around with that
or you could just fake it, with a sub harmonic harmoniser like this
http://www.waves.com/Content.aspx?id=337
really coo idea btw, and I wasn't expecting the lights :lol:
what sort of radio transmitter are you using? is it just FM?
ok I read the link, thats very very cool,absolutely love the mod bike, so scratch the subwoofer,
what you want is one or better two of these
http://www.billfitzmaurice.net/Jack.html
those car speakers can have very good sound, but they simply don't bring the SPL, what you really want is a PA type speaker, like the Jack10, it has really useful bass down to 60Hz, and wide dispersion, so in principle you wouldn't need the backpacks anymore, but they are so cool I'd keep them anyway, and you might find this hard to believe but you could actually comfortably power those with an amp6 thanks to the much greater efficiency, and you wouldn't have to cart around a huge battery either.
You don't actually need to produce frequencies below around 90Hz outside. If you have a slight rise in the midbassband around 120Hz the brain is automatically fooled to think there is subbass when in fact there is not. This trick only works outside. It doesn't work in a very large room, only outside.
It's because we have a inbuilt filter that disguish sound differently when outside or indoors, just like the brain also filters the input from our eyes differently when we're outside. You won't ever notice this because evolution has perfected it so that we don't ever see or hear it happening.
Thanks for all the replies, they're really helpful as I'm still finding my way with this stuff.
quote:Originally posted by gnome
really coo idea btw, and I wasn't expecting the lights :lol:
Thanks! The complete system was designed by StevieB and Sparky of the LondonSkate. I just copied the ASBOX backpack repeater idea and made these AMP6 powered lightweight versions to link up with the rest. And yeah, I couldn't resist the temptation to use those silver speaker cones as nice shiny diffuse reflectors :)
quote:Originally posted by gnome
what sort of radio transmitter are you using? is it just FM?
It's a Trantec S4000IEM (http://www.trantec.co.uk/html/p_InEar_Monitors_S4000IEM.htm) UHF system. The receivers are mounted on masts on the repeater units to get them above head height.
quote:Originally posted by Scratchy
You'd also get more power if your ran the direct audio 5.25" speakers at 4 ohms each - not in series.
The AMP6 can handle 4ohms just fine.
I think we're talking at cross purposes, the next version of the backpack is going to be a mono unit using the AMP11-LV. Originally I was thinking of 8 ohm load (2x Directed 5.25" 4 ohm in series plus 8 ohm woofer with crossover) but I've now decided to follow mikechw's suggestion of going with 4 ohm load so I'll need a pair of 8 ohm mid/high range speakers in parallel and a 4 ohm woofer.
quote:Originally posted by Scratchy
You should also look at some of krilli's posts on LiPo batteries in this forum (lighter and more power than AA NiMh).
I have looked into using LiPo/Li-ion batteries but I'm not entirely convinced that it's the way to go yet for a number of reasons. The AA Ni-MH I'm using are 2Ah Sanyo Eneloops which are robust and reliable, have a long life (if treated well) and have very low internal resistance so they can provide high peak current. Although Lipo would be lighter, the Tripath amps are so efficient that the batteries are already a small proportion of the total weight so we wouldn't save much. There's also the fire hazard when charging homebuilt Lithium battery packs. Laptops and their batteries have a lot of fail-safe protection circuitry built in but even they are known to catch fire occasionally. These units must be trustworthy and fool-proof, non-technical people must be able to just plug them into the charger and forget them without worrying about them catching fire if something goes wrong. And we've already got the peak detecting chargers for Ni-MH.
quote:Originally posted by gnome
ok I read the link, thats very very cool,absolutely love the mod bike, so scratch the subwoofer,
what you want is one or better two of these
http://www.billfitzmaurice.net/Jack.html
those car speakers can have very good sound, but they simply don't bring the SPL, what you really want is a PA type speaker, like the Jack10, it has really useful bass down to 60Hz, and wide dispersion, so in principle you wouldn't need the backpacks anymore, but they are so cool I'd keep them anyway, and you might find this hard to believe but you could actually comfortably power those with an amp6 thanks to the much greater efficiency, and you wouldn't have to cart around a huge battery either.
I'll pass that on to the bike guys, I know they're always looking at ways of improving efficiency, however I don't thing they're ready for a another full rebuild just yet.
The backpack units are definitely needed. The skate travels about 12 miles on average and although it's marshalled to keep everybody together, when there are several hundred skaters it can be 100 metres from front to back. Also, the skaters around the bike absorb the sound so the volume drops off quickly with distance. It's not like a stage setup where the speakers can project the sound above people's heads, we need the backpack repeaters to spread the sound throughout the skate. The present backpack units work very well, but it would be nice to have a couple more that put out more bass to fill out the sound.
DESIGN BRIEF:
With the 10x 2Ah AA cells powering the present AMP6 (2x25W) backpack we get about 6 hrs runtime which is more than enough. That means that 20x AA cells should be able to power the 100W AMP11 for a similar time, or maybe 21 or 22 cells to be able to drive 4 ohm load at full power. 22x AA cells weigh just over 600gms so that's not a problem. I'm aiming for a total weight of under 6 kg as that can be carried quite comfortably by a skater. So with those parameters fixed, I need a speaker/cabinet that can make the most efficient use of that power outdoors with a reasonable amount of bass. At a quick questimate I think I can build a sufficiently lighweight rigid cabinet with a net volume up to about 25 liters, although smaller would be nice, and preferably sealed as I think that generally means a smaller cabinet. I've used 9mm plywood for the present backpacks and it can be made very rigid with carefully designed internal bracing.
So it's a matter of finding a 4 ohm neodymium woofer that gives a reasonable compromise, and a pair of 8 ohm mid/high range that will be a good match for the woofer. There's probably no point in trying to get very low bass when there's less than 100W to play with, so Saturnus' idea makes a lot of sense. The Tang Band W8-1363SB Scratchy linked to looks interesting but unfortunately shipping is too expensive as I live in the UK, can't find anybody locally selling them.
LiPo cells are not standard Li-ion. They are completely safe and foolproof. They're even safer than NiMH batteries which also can catch fire when overcharged.
Just found the thread (http://www.41hz.com/Forums/topic.asp?TOPIC_ID=2740&SearchTerms=LiFePO4). Interesting, battery technology does seem to be moving along and I can see that for some applications it would be a good solution.
But they're still lithium which is highly flammable compared to Ni-MH cells, and although it's just about possible to cause a fire with Ni-MH you would need to use a very high charging current which would wreck the cells anyway after a few cycles. With any kind of lithium cell you are relying on the protection circuitry which can fail, which is why the Battery Space site includes a warning.
Assuming I've got my calculations right (big assumption :) ) I don't think it's an advantage for this particular project. For us the voltage/current requirements of the amp limits the amount of energy we need to store, there's no way we can get more than 100W at 27V using AMP11, and the skates last 2 hours so we don't really need more than 3 hours runtime. To get the 7A peak at 25.6V using LiPo we'd need need 2 of the 8x 18650 packs in series at 726gm, or 2 of the 4x 26650 at 670 gm. We could make a single series pack of 8x 18650 which would weigh 363gm but the peak current would be limited to 3.5A. Ni-MH will already give us double the runtime we need and weigh about 620gm so there's no advantage for us going to LiPo. We don't need long runtime, it's high voltage with high peak current we're after.
quote:Originally posted by MickB
But they're still lithium which is highly flammable compared to Ni-MH cells
Not the lithium ferrophosphate ones. They just aren't :)
quote:Originally posted by MickBWith any kind of lithium cell you are relying on the protection circuitry which can fail, which is why the Battery Space site includes a warning.
The protection circuity in LiFePO4 has a very different purpose from the usual lithium-ion protection circuitry. You can actually use them *without* the circuitry, both for charging and discharging. What the protection circuitry does is mostly three things - and it's mostly about protecting the cells, not small, flammable woodland creatures or children:
1) Stop overly high discharge currents, as LiFePO4 batteries can put out quite a lot of current! - But more to save the equipment they're connected to and prolong the life of the cells, not to avoid fires
2) Disconnect the cells when they get under a certain voltage, which can shorten their life a bit - but not to avoid fires, just to keep the cells good for longer
3) Balance the cells
The reason for the "must use PCM" disclaimer is AFAIK mostly because they advertise 1000-2000 cycles life -- and for customers to reach that kind of life, a prot. circuit is needed. But then again, 1000-2000 cycles is quite a lot!
While I'm not saying THESE ARE THE RIGHT CELLS FOR YOU!!1 or anything like that, I just had to point out some things here.
Actually Batteryspace guarentee minimum 2000 cycles on their LFP cells, if you're using a cell protection circuit.
LFP cells are completely nonflammable and nonexplosive. They're not just safer than regular lithium cells, they are safer than any other battery type. Peroid.
LFP cells also give far better peak current than NiMH cells, and most other battery types in general. For the type of use you're suggesting, NiMH cells are probably the worst choice possible. When you have high current draw on NiMH cells you degrade battery capacity and cycle life exponentially.
You don't "need" the protection curcuit on LFP cells really, they are drop in replacements for SLA cells. But just like SLAs, they don't like being overcharged and overdischarged, that's what the protection circuit is there for. It's to make sure you get the battery cycle life that is advertised. It's not to keep you safe.
quote:Originally posted by MickB
The backpack units are definitely needed. The skate travels about 12 miles on average and although it's marshalled to keep everybody together, when there are several hundred skaters it can be 100 metres from front to back. Also, the skaters around the bike absorb the sound so the volume drops off quickly with distance.
It's not like a stage setup where the speakers can project the sound above people's heads, we need the backpack repeaters to spread the sound throughout the skate.
Well the other problem you have is that sound drops off at 6dB for every doubling of distance, which is why you need the backpacks, there just isn't enough SPL to begin with.
Also if the skaters absorb the sound because the speakers are not up high, then the obvious solution is to put the speakers up high, on a pole, on the bike, say, this way you will need less power because you're playing sound into people's head's rather then their stomachs, and it will sound clearer, because you wil be getting less floor bounce.
yeah I know I'm not really helping here.
quote:
The present backpack units work very well, but it would be nice to have a couple more that put out more bass to fill out the sound.
DESIGN BRIEF:
With the 10x 2Ah AA cells powering the present AMP6 (2x25W) backpack we get about 6 hrs runtime which is more than enough. That means that 20x AA cells should be able to power the 100W AMP11 for a similar time, or maybe 21 or 22 cells to be able to drive 4 ohm load at full power. 22x AA cells weigh just over 600gms so that's not a problem. I'm aiming for a total weight of under 6 kg as that can be carried quite comfortably by a skater. So with those parameters fixed, I need a speaker/cabinet that can make the most efficient use of that power outdoors with a reasonable amount of bass. At a quick questimate I think I can build a sufficiently lighweight rigid cabinet with a net volume up to about 25 liters, although smaller would be nice, and preferably sealed as I think that generally means a smaller cabinet. I've used 9mm plywood for the present backpacks and it can be made very rigid with carefully designed internal bracing.
So it's a matter of finding a 4 ohm neodymium woofer that gives a reasonable compromise, and a pair of 8 ohm mid/high range that will be a good match for the woofer. There's probably no point in trying to get very low bass when there's less than 100W to play with, so Saturnus' idea makes a lot of sense. The Tang Band W8-1363SB Scratchy linked to looks interesting but unfortunately shipping is too expensive as I live in the UK, can't find anybody locally selling them.
you are going to struggle finding 4 ohm neo woofers with high efficiency, those are the domain of pro speakers, which are generally 8ohm, because the pro-market generally likes (needs) to parallel drivers.
those tangbands actually look like really decent speakers, but being a true subwoofers are useless to you, I think the Monacor unit you selected look a lot better, because you are not going to do anything below 100Hz, I'd team those with a couple or more piezos because they are light and loud, and actually sound really good in multiples, which will also simplify the crossover to the point where you don't need one, or just a coil across the woofer and a resistor in series with the piezos.
And I really think you're too hung up on the power rating, to go from 50W to 100W you gain just 3dB which is just barely audiable, which is why speakers are rated flat if they are within +/-3dB, besides if you actually were to use 100W flatout you'd be draining your battery in 20 minutes, those amp6 you have are really plenty.
IMO the only fruitful path to volume is to concentrate on speaker efficiency, which you get by either getting speakers of higher efficiency (and they always trade off bass response for it) or simply more speakers which increases the radiating area, incidently it is also the path to audio quality, also IMO.
quote:Originally posted by Saturnus
You don't actually need to produce frequencies below around 90Hz outside. If you have a slight rise in the midbassband around 120Hz the brain is automatically fooled to think there is subbass when in fact there is not. This trick only works outside. It doesn't work in a very large room, only outside.
It's because we have a inbuilt filter that disguish sound differently when outside or indoors, just like the brain also filters the input from our eyes differently when we're outside. You won't ever notice this because evolution has perfected it so that we don't ever see or hear it happening.
that's really interesting I've not heard that before, well I know that you can psychoacoustically conjure up bass, the box I linked to above does just that, but I wasn't aware that this would be more effective outdoors, can you point to any links that discuss this effect?
quote:Originally posted by krilli
Not the lithium ferrophosphate ones. They just aren't :)
My mistake, I assumed they were normal LiPo cells, I should have read it properly. Just when I think I know something about batteries somebody comes along and points out the gaping holes in my knowledge :)
quote:Originally posted by krilli
While I'm not saying THESE ARE THE RIGHT CELLS FOR YOU!!1 or anything like that, I just had to point out some things here.
Absolutely! I didn't know anything about that particular chemistry, thanks for the info. You have made me think again about those packs, although I think the cost is going to be too high with 2 packs plus charger.
quote:Originally posted by Saturnus
LFP cells also give far better peak current than NiMH cells, and most other battery types in general. For the type of use you're suggesting, NiMH cells are probably the worst choice possible. When you have high current draw on NiMH cells you degrade battery capacity and cycle life exponentially.
I agree with you about "normal" Ni-MH cells, from my experience they are fragile things which often degrade quickly even when you buy reputable brands and treat them carefully, and cheap off-brand ones are usually a complete waste of money. But I've been using low self-discharge Sanyo Eneloops which are rather different beasts. They have a much lower internal resistance and are happy to discharge at 2C continuous (4A for a 2Ah cell). Correct me if I'm wrong, but I think the main reason for cell degradation is heat build up with high currents, so the Eneloops low internal resistance should mean they'll be OK with intermittent peak currents well above 2C. However, if those LiFePO4 can manage 7A peaks without any problems then perhaps it is the way to go. Have to see how much everything else ends up costing.
quote:Originally posted by gnome
Also if the skaters absorb the sound because the speakers are not up high, then the obvious solution is to put the speakers up high, on a pole, on the bike, say,
Yeah, but there's a limit to how high you can get them without encountering serious stability problems. Above head height isn't really an option when you've got to ride the thing around the streets.
BTW, as people seem interested in these systems here are a couple more. The LFNS's Firebrox (http://homepage.ntlworld.com/m.brown74/Photos/Firebrox/Firebrox.jpg) which is used on the Friday and Sunday skates (different organisation to the LondonSkate although many of the same people actually run them and take part) and PaulC's Thumpers (http://www.bassfreight.com/index.php?option=com_content&task=view&id=17&Itemid=29) which are privately owned but are brought out for special occasions. The Thumpers are works of art with hand-beaten aluminium bodywork and powered by an electric motor from an invalid carriage.
quote:Originally posted by gnome
And I really think you're too hung up on the power rating, to go from 50W to 100W you gain just 3dB which is just barely audiable, which is why speakers are rated flat if they are within +/-3dB, besides if you actually were to use 100W flatout you'd be draining your battery in 20 minutes, those amp6 you have are really plenty.
Ah, that is very interesting. I'm beginning to realise that I know even less than I thought! One thing I still don't understand is why it would drain the batteries so much more quickly if the power output of the amp is doubled. Surely the power out is a function of the power in depending on the amp's efficiency. Why would a 100W amp use 36 times the battery power of a 50W amp (assuming the same efficiency)? Does it mean that the present speakers are only using a fraction of the potential output of the AMP6 at the moment, even at full volume? Apologies if this is very basic stuff but I'd really like to get a handle on it.
quote:Originally posted by MickB
BTW, as people seem interested in these systems here are a couple more. The LFNS's Firebrox (http://homepage.ntlworld.com/m.brown74/Photos/Firebrox/Firebrox.jpg) which is used on the Friday and Sunday skates (different organisation to the LondonSkate although many of the same people actually run them and take part) and PaulC's Thumpers (http://www.bassfreight.com/index.php?option=com_content&task=view&id=17&Itemid=29) which are privately owned but are brought out for special occasions. The Thumpers are works of art with hand-beaten aluminium bodywork and powered by an electric motor from an invalid carriage.
haha those thumpers are great, love it
quote:
quote:Originally posted by gnome
And I really think you're too hung up on the power rating, to go from 50W to 100W you gain just 3dB which is just barely audiable, which is why speakers are rated flat if they are within +/-3dB, besides if you actually were to use 100W flatout you'd be draining your battery in 20 minutes, those amp6 you have are really plenty.
Ah, that is very interesting. I'm beginning to realise that I know even less than I thought! One thing I still don't understand is why it would drain the batteries so much more quickly if the power output of the amp is doubled. Surely the power out is a function of the power in depending on the amp's efficiency. Why would a 100W amp use 36 times the battery power of a 50W amp (assuming the same efficiency)? Does it mean that the present speakers are only using a fraction of the potential output of the AMP6 at the moment, even at full volume? Apologies if this is very basic stuff but I'd really like to get a handle on it.
Nono power draw is quite linear 20 minutes is just a wild guess, well you can calculate backwards, if your 2Ah battery pack, or what is it? lasts 6 hours that's 300mA, or 3.6W continuous at 12V, bugger all. Well you could measure the current draw to get an actual picture of what is happening.
a 100W amp in't going to be much louder than a 50W amp, what it does give you is headroom, so you'll be less likely to clip the amp during loud passages, but quality is not going to be much of an issue since you'll be swinging the sound around constanly any nuances in sound quality are going to be quickly lost.
The power numbers are not actually that interesting, what is more important for you to understand is the efficiency numbers, those car speakers, sorry to be coming back to those I know you're set on building bigger backpacks, but I realy believe that you're trying to solve your problem in the least effective way, anyway, those car speakers will have something like 90-92 db/1W/1m, if you were to use a DR200 (http://www.billfitzmaurice.net/DR200.html) which you can build to be less than 12kg and has an efficiency of 104dB down to 110Hz and nearly 180 degree horizontal dispersion.
then with just 300mA or 4W you can reach 110dB!! to get to the same level with the car speakers would take 100W! it's game set and match, a more sensitive system is going to trounce a more powerful one every time.
quote:Originally posted by gnome
I know you're set on building bigger backpacks, but I realy believe that you're trying to solve your problem in the least effective way,
No really, forget about the bike, pretend I never mentioned it. Rebuilding it again isn't an option and it doesn't want to be any louder anyway for practical reasons. I understand what you're saying about efficiency, and if it was a static system sitting in the middle of a field I'm sure that would be the right way to go, but a street skate winding through narrow London streets imposes a completely different set of constraints. If the guys had the time, money and enthusiasm to rebuild the bike yet again (which they don't) I'm sure they'd like to make it more efficient, but only to make it lighter, not louder. There already tends to be a clear space immediately around the bike as it's uncomfortably loud when you get too close, making it even louder would just make the ears of the people near it bleed without doing anything for the skaters 50 metres ahead and behind around a couple of corners. And before you suggest it, no we can't build a second bike for all sorts of reasons!;)
The wireless linked backpack system is a success, it's practical, flexible and works as intended. So far there are 6 of them altogether and they provide good area coverage. As I said before, all I'm trying to do is come up with an additional design which can handle more bass but is under the 6kg weight limit. I'm trying to design a loud efficient backpack, not a loud efficient bike system.
There are actually 3 designs of backpack at the moment. The first two units, ASBOX I & II, were proof-of-concept and consist of a rucksack frame fitted with a small Class A-B car amp, a couple of of 6.5" ferrite car speakers and a 7Ah SLA battery and weigh 6 or 7kg. They are robust, cheap to build and they do the job but they're not very efficient and they're quite heavy, although still light enough to be carried by a skater. I designed mine (christened ASBOMB due to the prototype's resemblance to a suicide bomb belt!) to be a lightweight version of the ASBOXs. They are more expensive and far more time consuming to build but they sound better and they weigh only 2.5kg so you can almost forget you're skating with it, well...apart from the noise. I've built 3 ASBOMBs so far and will probably make another 2 when I get the time this winter. There is also an ASBOX III which is simply a beefed up version of ASBOXs I & II, it's got a more powerful car amp, bigger speakers and bigger batteries. Again, it works as intended as it's significantly louder than the other ASBOXs and ASBOMBs with a lot more bass, but it's weight makes it unwieldy and it's hard to find anybody who will carry it. I want to make a lighter more efficient version of ASBOX III.
BTW, gnome, I really do appreciate you taking the time to share your expertise, you obviously know far more about this stuff than I do. I just didn't want to veer off in a different direction :)
I think the best design for you would be a sort of line-array using 4 or more small 3" or 4" full-range speaker mounted vertically in a backpack. This would make the sound spread where you need it, and the sound will mostly be near-field. It will also be easy to make it very effective that way, and very sturdy.
I'll see if I can make a sketch of what I'm thinking.
Thanks, Saturnus.
Actually I was just looking at those P-Audio speakers that V-Bro used for his little AMP32 system (http://41hz.com/Forums/topic.asp?TOPIC_ID=1789&SearchTerms=macaroni). Would small speakers be able to provide much bass compared to the 6.5" Directed speakers I've used for my ASBOMBs? I admit that my ignorance of audio matters is extensive, but I was under the impression that the smaller the speaker the less bass it can handle. Also the efficiency is also quite low for the SPH-30 at 87 SPL.
Right, this is another area which I thought I understood but now I'm not so sure...
quote:Originally posted by gnome
Nono power draw is quite linear 20 minutes is just a wild guess, well you can calculate backwards, if your 2Ah battery pack, or what is it? lasts 6 hours that's 300mA, or 3.6W continuous at 12V, bugger all.
...which sounds as though you're saying that I'm only using a fraction of the AMP6's potential continuous output. But in another AMP6 thread Saturnus wrote this about power consumption...
quote:Originally posted by Saturnus
357mA on a stabilized 12 volt supply at maximum music output (-12dB) into 4 Ohms both channels loaded.
A standard sized 7Ah 12V SLA lasts about 20 hours of playing time (at maximum output).
... and I've seen several other references on this forum to the AMP6 never drawing much more than 300mA average at full volume, although peak current will obviously be much higher. It's what I used to decide on the original 10x AA battery pack for my ASBOMBs. I assumed it means that the 2x25W rating of the AMP6 was a maximum value which is never held continuously for any length of time when playing music in real life. I've also seen references here to the actual power draw of an amp being roughly 1/10 of it's rated RMS output value which is consistent with that.
And yet gnome also says this...
quote:Originally posted by gnome
And I really think you're too hung up on the power rating, to go from 50W to 100W you gain just 3dB which is just barely audiable, which is why speakers are rated flat if they are within +/-3dB, besides if you actually were to use 100W flatout you'd be draining your battery in 20 minutes, those amp6 you have are really plenty.
...which makes it sound as though the AMP11 can actually be used continuously at 100W output, which means a continuous current draw of about 4.5A at 24V.
Does an amplifier's RMS rating mean that it's possible to sustain that power continuously when playing music, with a consequently high current draw, or is 1/10 of the RMS rating the actual approximate average power output at full volume in real life?
quote:Originally posted by MickB
Also the efficiency is also quite low for the SPH-30 at 87 SPL.
Quite low!?! 87db is INSANE for a 3" driver!! And I've measured them myself, it is REAL efficiency. A lot of manufacturers mess with the units/figures to make it look impressive. I can tell you from experience Monacor specs don't tell lies!
Try to find one in the whole world that can do better! AND have the same resonance frequency of course!!
Don't forget that placing drivers close to eachother doubles the efficiency and more bass is easily achieved by more air movement or in other words adding filtered drivers to perform as "sub" addition....and placing the stereo set in one chamber, most bass is mono anyway so this will help too (also because it is in a larger enclosure with lower resonant frequency than one single driver would be)
As long as you are using enough amplifier channels this is easily achieved by PLLXO filtering.
------X-max. can be several meters on any driver, too bad it can be done only once...------
quote:Originally posted by MickB
quote:Originally posted by Saturnus
357mA on a stabilized 12 volt supply at maximum music output (-12dB) into 4 Ohms both channels loaded.
A standard sized 7Ah 12V SLA lasts about 20 hours of playing time (at maximum output).
... and I've seen several other references on this forum to the AMP6 never drawing much more than 300mA average at full volume, although peak current will obviously be much higher. It's what I used to decide on the original 10x AA battery pack for my ASBOMBs. I assumed it means that the 2x25W rating of the AMP6 was a maximum value which is never held continuously for any length of time when playing music in real life. I've also seen references here to the actual power draw of an amp being roughly 1/10 of it's rated RMS output value which is consistent with that.
That's very accurate. :D Remember, music is (fortunately) not sinus curves. RMS output is THE maximum continuous output which requires full volume half the time and no volume half the time on average, effectively that gives -3dB compared to maximum peak output, or precisely half that.
Music is fortunately mixed to -12dB average, if it upholds the RIAA norm. Because compressing it much more is virtually impossible without seriously affecting sound quality. -9dB is usually called pink noise (it's the maximum average level where different sounds can be detected) and -6dB is pure white noise.
87 dB/W/m of real sinsitivity for a 3" real full-range driver that can actually play bass well is impressive to say the least. 4 of those vertically mounted and coupled in series/parallel gives a 93 dB/W/m speaker. I don't think you can get much better than that.
As a side note, classical music is usually recorded at average level of either -18db (semi-compressed) or -24db (full dynamics range).
quote:Originally posted by V-bro
Quite low!?! 87db is INSANE for a 3" driver!! And I've measured them myself, it is REAL efficiency. A lot of manufacturers mess with the units/figures to make it look impressive. I can tell you from experience Monacor specs don't tell lies!
Try to find one in the whole world that can do better! AND have the same resonance frequency of course!!
Erm... I said something wrong there, didn't I ;)
quote:Originally posted by V-bro
Don't forget that placing drivers close to eachother doubles the efficiency and more bass is easily achieved by more air movement or in other words adding filtered drivers to perform as "sub" addition....and placing the stereo set in one chamber, most bass is mono anyway so this will help too (also because it is in a larger enclosure with lower resonant frequency than one single driver would be)
As long as you are using enough amplifier channels this is easily achieved by PLLXO filtering.
Thanks, that's all very useful info. Just to make sure I've understood correctly, if I go with Saturnus's plan for 4x 3" Monacor/P-Audio drivers I'd mount them all in a single chamber and I would need 4 amp channels. 2 for a stereo pair and 2 with PLLXO low pass filters to drive the other pair as subs. Is that right? In that case an AMP9-Basic would do the job.
I just looked up PLLXO filters and one potential problem seems to be that they cause an insertion loss. The output from our radio receivers is already very low and I've had to use 100k input resistors with the AMP6 to obtain sufficient gain. Can I raise the gain even higher with the AMP9, or I would a need to use a preamp to boost the amplifier input?
Nah, I'd drive them all full-range and let the PLLXO handle the equalization needed if any. Those 3" drivers are also very easy loads if you remember to filter out frequencies they can't reproduce, so you could do it with a single amp6basic and drive a pair into 2 Ohm in parallel. It'd just require some extra attention to cooling.
Also note that a supply voltage much over 12V is highly discouraged in that case since the output power is near the absolute limit of the amp.
Heheh, just educating the masses man, I'm not angry at all! :D
Here's perhaps something to get inspired on: http://41hz.com/forums/topic.asp?TOPIC_ID=2576
The PLLXO filters are rather to perfect the sound, but without them it sounds fairly good already. At larger distances perhaps even better. These speakers I rather designed for nearfield operation. The loss of my filter (yes every PLLXO filter design has different loss figures...) I compensated for by using step-up transformers from Neutrik (NTE-4) which are tiny and cheap and perform pretty well. They have a 1:4 voltage step-up ratio, so gain is basically times 4, but the loss of my filter is perhaps /3 on average. The net result is still a little extra :)
I chose smaller input caps on the amp to filter the anyways non-existing bottom end (to also protect the little drivers) 220nF or 330nF will work nicely. Less if you want them to be able to handle more, higher if you want more lower extension, between 100nF and 470nF can be experimented with.
To cut corners:
The reason why I chose the filter was partially the sharpish high frequency response. With four drivers I think you shall probably get a nice balance if you just put a lowpass filter on two of the drivers at around 8Khz... You can still go for the transformers to just up the gain massively, it'll also pull up the bass quite a bit with the NTE-4 transformers (you need one for each channel by the way)
AMP9 won't get you anywhere for this cause, forget about it...two AMP6 will be much more efficient for such a portable system.....
I have helped someone out on the design of speaker columns equipped with 15 SPH30X per channel and the SPL and SQ of this system is still causing many people to drool all over the place.... You can literally "feel" things happening. Still I first advised other drivers, but he liked the looks as well, it's just that for serious HIFI there are better drivers available. The SPH-30X is just mechanically VERY strong and has unbeatable efficiency...
------X-max. can be several meters on any driver, too bad it can be done only once...------
quote:Originally posted by V-bro
I have helped someone out on the design of speaker columns equipped with 15 SPH30X per channel and the SPL and SQ of this system is still causing many people to drool all over the place.... You can literally "feel" things happening.
Interesting, how is the connection topology? Serial / parallel / final impedance, I mean.
Originally posted by MickB
quote:
but only to make it lighter, not louder. There already tends to be a clear space immediately around the bike as it's uncomfortably loud when you get too close, making it even louder would just make the ears of the people near it bleed without doing anything for the skaters 50 metres ahead and behind around a couple of corners. And before you suggest it, no we can't build a second bike for all sorts of reasons!;)
The wireless linked backpack system is a success, it's practical, flexible and works as intended. So far there are 6 of them altogether and they provide good area coverage. As I said before, all I'm trying to do is come up with an additional design which can handle more bass but is under the 6kg weight limit. I'm trying to design a loud efficient backpack, not a loud efficient bike system.
Well that's the thing, you could put the single speaker on a normal bike, with the whole thing weighing no more than 12Kg, including battery and portable music player, and you don't have to be louder, in fact you can turn down the volume because you have the far superior dispersion.
quote:
BTW, gnome, I really do appreciate you taking the time to share your expertise, you obviously know far more about this stuff than I do. I just didn't want to veer off in a different direction :)
yeah I know, but I think it would be remiss of me not to tell you a superior option, I'm just content to put the seed of what's possible in your mind [}:)]
Although I appreciate that form can be just as important than function, I love the backpack, and the bike.
quote:Originally posted by MickB
...which makes it sound as though the AMP11 can actually be used continuously at 100W output, which means a continuous current draw of about 4.5A at 24V.
no I was speaking hypothetically, just making the point that you don't need that much power, or want to because that means having to cart a bigger battery.
I'm getting excited now, feels like the design is taking shape! I'm so glad I talked to you guys first instead of just jumping straight in.
A single AMP6-B would be nice as it's cheaper and simpler and I can use the same battery pack as the ASBOMB. In fact everything can be the same except the speakers and cabinet, and it'll probably be even lighter as well as louder. I can just buy the speakers to start with, build the cabinet and bolt it on to an existing ASBOMB harness to test it out.
From what gnome said about +3db being barely noticeable, I assume there'd be no advantage to using 2x AMP6, especially as the speakers would then be the limiting factor so it would be less than +3db anyway. It's just a way of obtaining 4 channels which I don't really need.
As I'll be using a single vertical array, I assume there's no point having any kind of stereo separation. So rather than drive a pair of paralleled speakers on each channel, would it be better to parallel the AMP6 and parallel all 4 speakers for a single full range load? If so, should I use 4x 8 ohm speakers to get a 2 ohm load, or 4x 4 ohm speakers for a 1 ohm load?
quote:Originally posted by V-bro
I chose smaller input caps on the amp to filter the anyways non-existing bottom end (to also protect the little drivers) 220nF or 330nF will work nicely.
Are those values for use with the standard 22k input resistors?
The input transformers look neat. I'm leaning towards not using a PLLXO filter, in which case is there any advantage to using the transformers to increase the input rather than 100k input resistors to increase the gain?
quote:Originally posted by V-bro
AMP9 won't get you anywhere for this cause, forget about it...two AMP6 will be much more efficient for such a portable system.....
Just out of interest, is that because the AMP6 is inherently more efficient? Or is it because the AMP9 would be working at only a fraction of it's rated output?
EDIT: And if all that works well, how about a 6 speaker vertical array wired in parallel/series driven by an AMP11-LV for the next version? :D The extra height could just about be fitted onto the backpack design. I realise that gnome's comments about power doubling still apply, but would the fact that it's a 50% larger array make the efficiency gain worth the extra cost? Or is a 4 speaker array the power/weight/efficiency/cost sweet spot for this sort of application?
quote:Originally posted by MickB
From what gnome said about +3db being barely noticeable, I assume there'd be no advantage to using 2x AMP6, especially as the speakers would then be the limiting factor so it would be less than +3db anyway. It's just a way of obtaining 4 channels which I don't really need.
I was being a little over dramatic there, 3dB is noticeable, and nice,
and in principle you want to eck out very single decibel that you can, just that increasing amp power is one of the least effective paths to do that. If you want to get a feel how much that is, take two speakers and put them next to each other play some music and toggle one on and off. That's 3dB difference. BTW if you connect the two speakers to the same amp channel in parallel, you will increase output by 6dB which is very significant, because not only are you doubling the surface area but your halving the impedance, the amp sees. That's why arrays of speakers are so seductive.
quote:
As I'll be using a single vertical array, I assume there's no point having any kind of stereo separation. So rather than drive a pair of paralleled speakers on each channel, would it be better to parallel the AMP6 and parallel all 4 speakers for a single full range load? If so, should I use 4x 8 ohm speakers to get a 2 ohm load, or 4x 4 ohm speakers for a 1 ohm load?
I'm not sure you can run the Amp6 down to 2ohm, maybe 3ohm?
quote:
Just out of interest, is that because the AMP6 is inherently more efficient? Or is it because the AMP9 would be working at only a fraction of it's rated output?
no it's because they are both limited by the same voltage (V^2/R), well you can run the amp9 at 24v, but that again means that you need a bigger battery, you just can't get away from the fact that the more power you use the bigger the battery will have to be.
quote:
EDIT: And if all that works well, how about a 6 speaker vertical array wired in parallel/series driven by an AMP11-LV for the next version? :D The extra height could just about be fitted onto the backpack design. I realise that gnome's comments about power doubling still apply, but would the fact that it's a 50% larger array make the efficiency gain worth the extra cost?
as arrays go, 50% bigger is good, but only for sound quality (less stress on individual drivers and smoothing frequency response) and directivity, increase in output is not significant, 4 is a good number and you're not really going to have much from directivity and sound quality with constantly moving speakers, I also wouldn't dismiss your inital idea of using that 10" monacor you mentioned, it certainly has the oompf, but is going to be much more unwieldy than 4 3" drivers.
quote:Originally posted by gnome
I'm not sure you can run the Amp6 down to 2ohm, maybe 3ohm?
You can use it down to 2 Ohm - just don't push the voltage too high, and make sure you cool it. Roughly: you can use it without a heatsink with 4 Ohm loads, but you must heatsink it with 2 Ohm loads, at least with any reasonable power output.
I ran my Amp6-B for a long while on 2.66 Ohm, by accident kind of. Not really heat-sunk well enough though, but it worked fine, and still does.
The max current output is around 6 Amperes, don't remember exactly. This is the main constraint. Plus the heat. Stay within the three max ratings, V A and C, and you'll be OK.
quote:Originally posted by gnome
I'm not sure you can run the Amp6 down to 2ohm, maybe 3ohm?
I meant a 2 ohm load if the AMP6 inputs and outputs were paralleled to produce mono output. I've no need for stereo if it's a single vertical array, and I was under the impression (perhaps wrongly) that it's best if all the speakers are trying to play exactly the same signal if they're part of an array and are sharing the same chamber.
quote:Originally posted by gnome
as arrays go, 50% bigger is good, but only for sound quality (less stress on individual drivers and smoothing frequency response) and directivity, increase in output is not significant, 4 is a good number and you're not really going to have much from directivity and sound quality with constantly moving speakers,
Thanks, that all makes a lot of sense. So 4 it is.
quote:Originally posted by gnome
I also wouldn't dismiss your inital idea of using that 10" monacor you mentioned, it certainly has the oompf, but is going to be much more unwieldy than 4 3" drivers.
Hmm, just when I thought I'd decided... :)
If the increase in volume and bass response is sufficiently great compared with 4x 3" then I can deal with a certain amount of unwieldiness. As I mentioned before, 6 kg is quite manageable on skates and I reckon I can keep it get under that weight with up to a 25 litre cabinet for the 10", a couple of small high range speakers of some description, and double the present size of battery. Wasn't the fact that the 10" Monacor is only available as 8 ohm a problem when we talked about it before? I am still interested in looking at that option though if you think it might be worth it.
I was also looking at some 10" Celestion neodymium bass guitar speakers which are a little lighter, almost as efficient (allegedly) and available as 4 ohm: BN10-300S (http://www.lean-business.co.uk/eshop/index.php?main_page=product_info&cPath=20&products_id=327) BN10-300X (http://www.lean-business.co.uk/eshop/index.php?main_page=product_info&cPath=20&products_id=542) (pdfs here (http://professional.celestion.com/bass/pdf/BN10-300S4.pdf) and here (http://professional.celestion.com/bass/pdf/BN10-300X4.pdf)). I contacted Celestion about them and they say they really need ported enclosures of 35 - 40 litres to get the best out of them though which I think is going to be a bit too big and heavy.
My present plan is to buy some of those Monacor/P-Audio 3" anyway once we've finalised the details, and build the vertical array cabinet. I can try it out with the AMP6 I've already got here without having to buy anything else. I suspect the design is going to be better than the present ASBOMB design and even a little lighter, so I can use it for the next 2 ASBOMBs instead of the Directed 6.5" speakers. If I feel it's still a bit lacking compared with the large ASBOX III then I can maybe try a 10" design.
quote:Originally posted by krilli
quote:Originally posted by V-bro
I have helped someone out on the design of speaker columns equipped with 15 SPH30X per channel and the SPL and SQ of this system is still causing many people to drool all over the place.... You can literally "feel" things happening.
Interesting, how is the connection topology? Serial / parallel / final impedance, I mean.
A big row of dedicated power amp channels...[}:)]
------X-max. can be several meters on any driver, too bad it can be done only once...------
quote:Originally posted by MickB
A single AMP6-B would be nice as it's cheaper and simpler and I can use the same battery pack as the ASBOMB. In fact everything can be the same except the speakers and cabinet, and it'll probably be even lighter as well as louder. I can just buy the speakers to start with, build the cabinet and bolt it on to an existing ASBOMB harness to test it out.
From what gnome said about +3db being barely noticeable, I assume there'd be no advantage to using 2x AMP6, especially as the speakers would then be the limiting factor so it would be less than +3db anyway. It's just a way of obtaining 4 channels which I don't really need.
As I'll be using a single vertical array, I assume there's no point having any kind of stereo separation. So rather than drive a pair of paralleled speakers on each channel, would it be better to parallel the AMP6 and parallel all 4 speakers for a single full range load? If so, should I use 4x 8 ohm speakers to get a 2 ohm load, or 4x 4 ohm speakers for a 1 ohm load?
quote:Originally posted by V-bro
I chose smaller input caps on the amp to filter the anyways non-existing bottom end (to also protect the little drivers) 220nF or 330nF will work nicely.
Are those values for use with the standard 22k input resistors?
The input transformers look neat. I'm leaning towards not using a PLLXO filter, in which case is there any advantage to using the transformers to increase the input rather than 100k input resistors to increase the gain?
quote:Originally posted by V-bro
AMP9 won't get you anywhere for this cause, forget about it...two AMP6 will be much more efficient for such a portable system.....
Just out of interest, is that because the AMP6 is inherently more efficient? Or is it because the AMP9 would be working at only a fraction of it's rated output?
EDIT: And if all that works well, how about a 6 speaker vertical array wired in parallel/series driven by an AMP11-LV for the next version? :D The extra height could just about be fitted onto the backpack design. I realise that gnome's comments about power doubling still apply, but would the fact that it's a 50% larger array make the efficiency gain worth the extra cost? Or is a 4 speaker array the power/weight/efficiency/cost sweet spot for this sort of application?
To cut corners you may be fully happy with just one AMP6 driving 2 speakers per channel and on each channel one driver is lowpassed at 8Khz with a coil. Only problem may be the phase shift the coil brings along, but that could not be much trouble in practise...
You could use three 8 ohm drivers per channel too.... then LP filter two drivers per channel...
AMP9 is designed for low impedance and is much more powerful. It has more potential than this project needs and due to that it has higher quiescent current, thus less efficiency for this project. If you were to fully blast on at least 18Vdc and low impedance speakers you may wanna try it :)....
The cap values I tried with 47K Rfbk and 22K Rin.
------X-max. can be several meters on any driver, too bad it can be done only once...------
Still two AMP6-B will give you better filter options without too much phase shift problems. And the NTE-4 transformers I find more or less a must with the SPH30X. The combination of PLLXO-NTE4-SPH30X just sounds great together.....
------X-max. can be several meters on any driver, too bad it can be done only once...------
quote:[i]Originally posted by MickB
If the increase in volume and bass response is sufficiently great compared with 4x 3" then I can deal with a certain amount of unwieldiness. As I mentioned before, 6 kg is quite manageable on skates and I reckon I can keep it get under that weight with up to a 25 litre cabinet for the 10", a couple of small high range speakers of some description, and double the present size of battery. Wasn't the fact that the 10" Monacor is only available as 8 ohm a problem when we talked about it before?
with the higher efficiency of the 10" you're still ahead, spl wise, over the 3" array.
also there will be more bass with the 10", for a rough apples to apples comparison, calculate the swept volume of the drivers,
(ie effective cone area * xmax), so for the 10" monacor that's 106cm
and for 4 of the sph30x that's 24cm which gives you a rough idea of bass capability.
quote:
I was also looking at some 10" Celestion neodymium bass guitar speakers which are a little lighter, almost as efficient (allegedly) and available as 4 ohm: BN10-300S (http://www.lean-business.co.uk/eshop/index.php?main_page=product_info&cPath=20&products_id=327) BN10-300X (http://www.lean-business.co.uk/eshop/index.php?main_page=product_info&cPath=20&products_id=542) (pdfs here (http://professional.celestion.com/bass/pdf/BN10-300S4.pdf) and here (http://professional.celestion.com/bass/pdf/BN10-300X4.pdf)). I contacted Celestion about them and they say they really need ported enclosures of 35 - 40 litres to get the best out of them though which I think is going to be a bit too big and heavy.
and I thought the Monacor was impressive, 1.6kg is amazing, the first one is going to be louder and the second one is going to give you more bass,it's always the trade off, always.
It think if you were to make a curved box you could get to 30L and you can use 6mm ply, I wouldn't have a any qualms about going to 3mm, because as soon as you curve a flat surface it becomes incredibly stiff, it's how those JBL EONS et al. can get away with their 3mm plastic casing.
maybe cut a ring of 12 or 15mm ply for the driver to mount on and glue that to 3mm ply, make the ends out of PVC pipe, maybe something like this, you can put piezos on the sides to really spray the sound around
http://farm3.static.flickr.com/2585/3841427744_2e916f1462.jpg
in fact if you brace it up like this it will be so stiff
you will be able to stand on it
http://farm3.static.flickr.com/2482/3840703605_868a12c7e2.jpg
hehe I'm really getting into this
That's excellent! I'm very tempted to give it a go. Hadn't realised 3mm ply would be stiff enough even when curved, and that's a neat idea to use PVC pipe for the sides, it's not only stiff but it'll probably take the inevitable knocks better. I'm itching to make a cardboard mock-up now to see how bulky that volume is when it's on my back. Thanks gnome, you're inspiring me!
I'm guessing the 300S would be better for this as I haven't got a huge amount of power.
quote:The bloke at Celestion said
Although the speakers will work in a closed box they deliver more punchy bass in vented enclosures. Typical volume for the 10 is 35-40litres and the driver itself will take up about a litre so probably isnt worth worrying about too much.
How critical is enclosure volume for a speaker like this? Would 30L be about the minimum I can get away with? I realise it's all a matter of compromises and there are no "correct" answers.
I know even less about piezos than I do about normal speakers. I'm planning on doing a spot of searching later, what sort of specs should I be looking out for?
quote:Originally posted by gnome
hehe I'm really getting into this
lol, I know just what you mean. Feeling slightly resentful that I've got to do some proper work today instead of playing speaker cabs. When the project bug bites...
quote:Originally posted by MickB
quote:The bloke at Celestion said
Although the speakers will work in a closed box they deliver more punchy bass in vented enclosures. Typical volume for the 10 is 35-40litres and the driver itself will take up about a litre so probably isnt worth worrying about too much.
How critical is enclosure volume for a speaker like this? Would 30L be about the minimum I can get away with? I realise it's all a matter of compromises and there are no "correct" answers.
30 litres will probably be about right, 25 may be better, to raise the corner frequency to above 100Hz else you're just wasting amp power, even 20 may be okay.
quote:
I know even less about piezos than I do about normal speakers. I'm planning on doing a spot of searching later, what sort of specs should I be looking out for?
doesn't really matter that much actually, because you want to use many, in singles they can be really rough at high volume, but the more you string together the smoother frequency response becomes, and the less excursion there is going to be, and then they sound truly incredible, I mean actually hifi
like this one don't be fooled by the price, they are the goods
http://www.parts-express.com/pe/showdetl.cfm?Partnumber=270-011
the nice thing is that you don't need a crossover, because they act like a capacitor and therefore a high pass filter, actually this can be a problem because some amplifiers don't like purely capacitive loads, don't know how the amp6 fares but it is good practice to put a 4 ohm non inductive resistor in series to prevent those problems.
Been playing with an online speaker box calculator (http://www.bcae1.com/spboxad2.htm) which agrees with you as it recommends a 29L ported enclosure for the 10" Celestion. I can knock together a couple of simple chipboard boxes of different volumes just to test it before building the real thing.
The calculator says the efficiency is 95.2db, a little lower than the claimed 97db and considerably less than the 10" Monacor's 99db (V-bro says Monacor's specs can be trusted). Still higher than the 93db of the 4x 3" array but not by a huge margin, although presumably the high efficiency of the piezos has to be factored in too.
----
EDIT: Hmmm... just fed the 10" Monacor's specs into the calculator and it's says it's efficiency is 93.5db so not sure what that means.
----
There are lots of places in the UK selling Skylon 85 x 85 mm piezos which look very similar to the one you linked to. 1.61GBPounds a pair here (http://www.lifeismusic.co.uk/Electronic/Piezo-horn-tweeter-85-x-85-x-7-12450.aspx). It's not often hi-fi stuff comes so cheap!
*goes off to research porting*
quote:Originally posted by MickB
(V-bro says Monacor's specs can be trusted)
so far so good....
I can't guarantee this counts for all of them as I haven't tried them all. Monacor sells many speaker drivers from many different makes....
------X-max. can be several meters on any driver, too bad it can be done only once...------
quote:Originally posted by MickB
Been playing with an online speaker box calculator (http://www.bcae1.com/spboxad2.htm) which agrees with you as it recommends a 29L ported enclosure for the 10" Celestion. I can knock together a couple of simple chipboard boxes of different volumes just to test it before building the real thing.
it's not that critical actually, all the volume of the cabinet controls is how the bass rolls off, nothing else, you don't want to produce notes below 100Hz anyway, because they waste power and you can't really hear them anyway because they'll be pretty thin, there is also the problem of the null produced by the floor bounce, because the speaker is 1.5m or so above the ground, but presumably you'll be playing music and not sine waves to amuse yourselves, so it's not an issue to get antsy about, actually everybody seems to have one particular bee in their bonnet when it comes to speaker building and they will spend vast sums of time and money to overcome one issuse, neglecting 10 other issues in the meanwhile (mine is sensitivity btw)
quote:
The calculator says the efficiency is 95.2db, a little lower than the claimed 97db and considerably less than the 10" Monacor's 99db (V-bro says Monacor's specs can be trusted). Still higher than the 93db of the 4x 3" array but not by a huge margin, although presumably the high efficiency of the piezos has to be factored in too.
----
EDIT: Hmmm... just fed the 10" Monacor's specs into the calculator and it's says it's efficiency is 93.5db so not sure what that means.
----
okay I'm not quite sure about this, I know that the reference efficiency is useful to compare different drivers with how they will produce bass, but not sure how they calculate the dB from that or at what frequency , but don't worry about it too much, if you look at the frequency response graph you'll see that a single figure efficency number is meaningless anyway, and that changes again if you put it in a box, but it all sounds about right, the Monacor has double the impedance which makes the amp run cooler and more efficiently, there's always a compromise to bemade, and no 10" is going to give you 99dB with 1W unless you put it in a horn, also you're certainly not going to get 93dB from 4 of those little Monacor at 100Hz. I'd go for the Celestion simply because it is lighter and because the amp6 will deliver more power to it, which is okay in this instance because it's free, well sort of
Ah, so it's the old problem of trying to quantify complex behaviour with a single number. Yeah, I see what you mean about the "free" power. It's slightly frustrating as I need only half the present run time and I could easily carry double the size of battery, so in theory with a suitable amplifier I could provide 4 times the power and the speaker could easily take it to give me +6db, but the combination of voltage, peak current requirements and available speaker impedance mean I can't actually drain the batteries fast enough even if I used LifePO4 cells. Damned physics getting in the way!
That's interesting about box volume only affecting corner frequency, I assumed it affected all sorts of other stuff as well. I feel like I'm getting a free intensive course on audio system design and I very much appreciate the patience of you guys teaching me this stuff. Yeah, I know it's because you're interested too, but it would be so much harder if I was struggling on my own.
OK, if you guys will bear with me once more, a quick recap of what I'm trying to achieve with the 10" and it's enclosure (please correct me if I'm going wrong anywhere)...
1. For this particular application there's no point trying to get anything much below 100Hz, so presumably I can filter out frequencies below that by choosing suitable input capacitor values and/or use PLLXO filters together with a smaller than recommended enclosure volume as suggested by gnome. (probably not PLLXO filters though, see below)
2. Saturnus mentioned that a small sensitivity peak at 120Hz will give the subjective impression of extended bass response. How should I achieve that? Is it a matter of playing with port tuning?
3. I've done a bit of reading up on porting and I think I've got a rough idea of what's going on. For a 3" (76mm) port the speaker box calculator tells me length should be 1"(25mm) with a 29 litre box. As I'll probably be using a 25 litre box or less, which will raise resonant frequency, I assume I'll need to change the port proportions as well. None of the calculators I've seen allow me to set box volume as a variable, so is there any way of finding out suitable port length?
-----
EDIT: I've downloaded AJ Design's Vented Designer (http://www.ajdesigner.com/speaker/ajvented.php) and that recommends a 19 litre enclosure anyway with SBB4 alignment. From what I've been able to find out, alignment is a way of characterising particular patterns of response, and SBB4 seems like it's the sort of thing I'm after. Does that sound reasonable? It all ties in with gnome's comments on box volume. Fb = 80.7 Hz (same as Fs) and F3 = 82 Hz. For a 76mm dia port it recommends a length of 53mm.
-----
And finally (yeah, right :)): To be able to feed most of the AMP6 output to the single 10" I assume I want to parallel the inputs/outputs. I suspect I can't just parallel my source outputs (mp3 player/radio receiver/whatever) by wiring them together. I'm guessing that the input transformers V-bro suggested would isolate them from one another but I'm reluctant to buy them as they're going to cost me 25 quid (US$40) including shipping and I'm already spending more than I ought. Is there a cheap way of doing this without causing insertion loss?
I can assure you that cutting down on the cost of the transformers is a big mistake, they soup-up the sound quite nicely and a project like this can use a little of this kind of extra sauce....
You will agree with me once you've done a "with" and "without" test...
------X-max. can be several meters on any driver, too bad it can be done only once...------
Fair enough, V-bro. If it's going to be worth it then I'm happy to spend it, and the receiver outputs could do with a boost anyway. So there's no problem paralleling the inputs after the transformers?
Heh, I've just remembered I was happily talking about twices as much battery and more expensive amp, but because this is little I don't want to spend money on it even though it's a lot cheaper. But it's just as important as any of the other bits because it's handling the entire signal near the beginning of the chain, so it affects everything that comes after it. It's finally sunk in that you've really got is a little self-powered pre-amp consisting of the transformers, to boost the signal voltage, and the PLLXO to trim it and shape it so that it best suits the following components. The amp and speakers then behave more efficiently and sound better. So at the cost of drawing a miniscule amount of power and adding a tiny amount of weight for the pre-amp, you actually make the whole system more efficient. That's nice!
To start with I just wanted to make a music box that was light and loud and Tripath amps seemed the best way of doing it. So I'd bung one together - bosh! - job's done. But now I'm beginning to understand why it could be a very addictive game in it's own right.
quote:Originally posted by V-bro
I can assure you that cutting down on the cost of the transformers is a big mistake, they soup-up the sound quite nicely and a project like this can use a little of this kind of extra sauce....
You will agree with me once you've done a "with" and "without" test...
------X-max. can be several meters on any driver, too bad it can be done only once...------
I think you need to keep the end use and environment in mind.
I'm not sure you will notice a difference with a single speaker playing behind you in an open area.
Using a single transformer with 2 inputs will automatically sum the signals (due to the common iron core) but keep the signals electrically isolated.
Here is a speaker box calculator that allows you to adjust the volume and calculate the vent size accordingly:
http://www.linearteam.dk/default.aspx?pageid=winisdpro
<edit> This also has a crossover calculator as well as a signal generator, so you can test and tune your design once it's built. </edit>
***It's so easy, everyone should be doing it***
Well, there will be a lot of people with the speaker in front of them and facing them, the music isn't primarily for the benefit of the person carrying it, but it is a noisy environment and the box is moving around so I don't want to get too carried away chasing sound quality. The nicer the better within reason though as it's likely to get used in a static set-up occasionally, although that's not it's main purpose.
Thanks for the calculator link.
It is possible that just one with dual primaries'll work satisfactory, but from my experience with my Macaroni the sound was incredibly stereo from just the one unit. The transformers indeed play as an effective preamp to compensate for the PLLXO loss. The net result is so much better than without these compensations.
------X-max. can be several meters on any driver, too bad it can be done only once...------
Well, the test box is up and running, and it sounds superb! The sound quality is streets ahead of the ASBOMBS as well as producing lots more bass. Portable hifi :D
Tried 19 litres to start with but it chopped off rather a lot of bass so I've upped it to 24 litres which sounds about right, so you were spot on, gnome. Took me a while to work out why there wasn't any top end to start with until I worked out that I should have connected the tweeters in parallel, not series [:I]
If anything it's still sounds a little bit light at the top end but that's playing it in the house, haven't had a chance to turn it up loud outdoors yet as I don't think the neighbours would appreciate it at this time of night. Hope to give it a side-by-side test with the ASBOMB outside tomorrow.
quote:Originally posted by MickB
Well, the test box is up and running, and it sounds superb! The sound quality is streets ahead of the ASBOMBS as well as producing lots more bass. Portable hifi :D
Tried 19 litres to start with but it chopped off rather a lot of bass so I've upped it to 24 litres which sounds about right, so you were spot on, gnome. Took me a while to work out why there wasn't any top end to start with until I worked out that I should have connected the tweeters in parallel, not series [:I]
If anything it's still sounds a little bit light at the top end but that's playing it in the house, haven't had a chance to turn it up loud outdoors yet as I don't think the neighbours would appreciate it at this time of night. Hope to give it a side-by-side test with the ASBOMB outside tomorrow.
Wow, you're not pissing about, that's fanatastic, this is just a normal wooded box I presume?
Now the frequency response is going to be far from flat and, using a 10" driver with piezo's is really stretching things in the crossover region and the sound quality is pretty much all a consequence from the efficency but there is a few things you can do, try swapping polarity of the tweeters (which are all wired in parallel I hope?) compared to the woofer, also line (not stuff) the cabinet with 30mm of open cell foam, that stops sound coming back through the cone, try going to 6 piezos, the more drivers the smoother the response.
Great man! Congratulations on that!
Please post us some pics as soon as you feel it is worth a peek!
------X-max. can be several meters on any driver, too bad it can be done only once...------
Hehe, being driven by the enthusiasm of a new project. And it's much more fun than renovating the windows of my house which is what I should be doing at this time of year.
Not much to look at yet, it's just a rectangular box which I knocked up from some MDF from my junk pile, very ugly, although I think the tweeter arrays look cool.
Behold the beautiful craftsmanship...
http://homepage.ntlworld.com/m.brown74/Photos/ASBOMBS/Test_Box.JPG (http://homepage.ntlworld.com/m.brown74/Photos/ASBOMBS/)
I'm using it to get the volume, porting and general arrangement right before I start on the real thing. I've only got the 10" and tweeters so far but the input transformers are on order. I'm running it off one channel of the AMP6 in the ASBOMB at the moment with a dummy load on the other channel.
That's a good idea about swapping the polarity to see if it makes a difference, gnome. It did cross my mind when I was wiring them up but I assumed it didn't matter because of the different output ranges, but of course there must be some overlap. It was just a first impression though, and listening to it again this morning I'm not sure it really needs anything more. I'll see what it sounds like outside. Also I'm thinking of angling the tweeters forwards (backwards) a bit on the real thing.
There's an annoying buzz when I turn the bass up loud which seems to be coming from the tweeters. It does it whether the tweeters are connected or not and I have a suspicion it's the tweeter terminal blades vibrating. On the finished article I can smother them in hot melt glue, but I can't test it on this box as I need to be able to unsolder them to remove them.
That WinISD calculator that Scratchy linked to is excellent. It's nice being able to see the response curve change as you adjust different parameters and then use the signal generator to hear what effect it has, gives a good feel for what's going on.
NICE
I'd wear that on my back even if it was silent.
cut the piezoes down so the elements are closer together that way you can fit six next to the woofer like this
glue them together using pipe cement
it will sound better, because of less stress on individual elements and you're cutting down on comb filtering because the centers are closer together
and you will be able to use them for casual listening, having tweeter and woofer horizontally next to each other isn't ideal, but speaker building is all about compromise :)
it's certainly far superior than the horizontal array.
http://farm4.static.flickr.com/3449/3864904088_695854e64d_o.jpg
test them individually by putting a couple of watts (that'll be really loud) of pink noise through them that way you can easily tell the funky ones
because they are so cheap there is not a huge budget for quality control, get 20% more piezos than you require
and that you're hearing vibration from them tells me that they are leaking, and your idea of slathering them with hotglue is in fact the usual method of fixing that problem
LOL. You might need a hand lifting it though...
Tested it outside in the garden and it sounds very good although it's amazing how quickly the sound drops off outside compared to indoors. Couldn't quite manage a side-by-side test with the ASBOMB as I had to swap speaker leads over between tests, but subjectively it sounds significantly louder with a lot more bass and much better overall sound quality. So basically it fulfils it's design brief.
However it doesn't have quite as much top end as the ASBOMB and I do wonder whether the piezos are paying their way for the weight they're costing. Their efficiency is nice and they provide a spot of icing on top, but if I disconnect the 10" so just the piezos are playing they're contributing very little to overall sound level. They are so cheap that it was well worth trying them, but 8 of them weigh nearly 500gm which is about the same as 3x Monacor SPH-30X. I'm wondering whether it's worth going with gnome's idea of a single vertical array beside the bass, but using 3x 3" instead of piezos. They'll be less efficient but I've got plenty of spare battery power and amplifier output which I'm not making use of so that probably doesn't matter. I could use one channel of the AMP6 for the 10" and the other for the 3x 3" array with a PLLXO on each channel to make the most efficient use of each.
What do you guys reckon? I'm beginning to come to terms with the fact that just doubling power output doesn't necessarily make a huge difference to subjective sound level. It's going to be a more expensive option so if it's only going to make a marginal difference then I'm not so keen. And will it miss the very top end which the piezos provide? However if the improvement is likely to be significant then it would be worth it. I also realise that there's no substitute for live testing so it's hard to be sure about any of this anyway until I try it.
The SPH-30X can be very loud with the three of them and since they have a little bump in the top end (I normally notch-filter this away) I'm sure you won't complain about having too little top end.
Combined with the 10" you can put them in very small closed inside cabinets and filter much of the bottom-end to get good power handling.
That said you could also use some of those tiny 2,5mm soft-dome mini hornloaded tweeters from Monacor (although they are much more expensive than the piezos of course)
They ARE very small and damned efficient though!
http://www.impactaudio.co.uk/product.asp?ref=970&session=1da77d27fe9c9937aa28bf28e2f39a9b
I feel like some Monacor affiliated salesman here, I MUST add that I have no affiliation with them whatsoever. I just have had most good experiences with their products whereas for instance Visaton (also have many cheap units, but perform likewise) has always left me down....
------X-max. can be several meters on any driver, too bad it can be done only once...------
Those soft dome tweeters look interesting, and don't worry V-bro, if I buy them I'll tell Monacor you recommended them to me so you'll get your usual percentage ;). Seriously though, it's really nice talking to people who are experienced with this sort of thing and are willing to pass on their hard-earned knowledge.
How many do you think I'd need to give an output that would be a reasonable match for the 10"? And are they highly directional? Sorry to keep asking all these questions, I know the final decisions are all down to me, just trying to get as good a result as possible without the learning process costing me too much.
They have a reasonable good spread, of course also much dependent on where they are situated in the casing. I wouldn't be suprised if one is enough, but a line array of course raises the level (+6db) quite a bit for each added unit, they have a BIG advantage in being small to avoid too much comb filtering effects and for your project the other advantage is they are extremely light weight.
I don't want to kick into anyone's shins, but I have NEVER hear good piezos myself. Last time I tried them is a LONG time ago though and they are perhaps better now, and I have more experience too. But I have sworn never to use them any more myself.....
Those domes have very low resonant frequency, sound very nice, are small, light, efficient and don't cost too much for all these nice features if you ask me, so why bother with those crappy piezos is my thought......
------X-max. can be several meters on any driver, too bad it can be done only once...------
Yes, the small size and low weight are very tempting features, especially as they seem a better match for the 10" which seems to cope very well with the midrange. I'm guessing the SPH-30X would duplicate a lot of what the 10" is doing anyway as well as being a lot bulkier, heavier and less efficient than these tweeters. As I want to spread the sound out to the sides, at least to some extent, I'm tempted to angle the sides of the box at 45 degrees and try just one on each side.
I think I might order a pair and play around with the placing to see what gives the best results. I can always get 1 or 2 more later if I think they're needed. Will they need their own closed chamber too?
I deleted the other post, it is all due to the EEEXXXXTRRREEEEEEEMMMMEEEEEELLLLLLYYYYYYYYYY SLLLLLOOOOOWWWWW internet I have here in a hotel in Turkey...
------X-max. can be several meters on any driver, too bad it can be done only once...------
No, they are closed already. They only need a simple 12db/oct HP filter to protect them from low frequencies. When you filter higher than 3K you can even get away with a 6db/oct filter...
You can try a 0,82mH aircore coil and 5,6uF cap (MKT/mylar quality at least) to start with. Then you filter roughly at 2,5Khz and 12db/oct.... Maybe you want to filter higher, I haven't looked into the frequency response of your 10" so far as I have too slow internet here to have a closer look. From tomorrow I'm in another hotel and probably have faster internet....
------X-max. can be several meters on any driver, too bad it can be done only once...------
http://www.snapdrive.net/files/361770/SNC11456.JPG
------X-max. can be several meters on any driver, too bad it can be done only once...------
Thanks, V-bro, that's a really useful table. Saved!
The official Celestion spec is 70 - 4k, and the freq response curve matches that as it dives steeply from 95db just above 4k.
Actually it's not just your hotel connection, the forum site has been behaving very erratically today.
Oh, and I meant my double post, not yours :) It wouldn't let me delete mine so I edited it.
quote:Originally posted by V-bro
I don't want to kick into anyone's shins, but I have NEVER hear good piezos myself. Last time I tried them is a LONG time ago though and they are perhaps better now, and I have more experience too. But I have sworn never to use them any more myself.....
You are absolutely correct, piezo's are crap, if you use them in singles, they are far from flat, and they have hat screechy high distortion hat really grates you, pretty much everybody who has tried them has exactely the same experience, however if you put them in a line this all changes, the frequency response does smooth out and they become extremely efficient. I was pretty skeptical as well when I first tried them, but the proof is in the listening, I have a speaker with a line of eight piezos and it is the equal of any high end speaker I have heard, it is crisp and buttery smooth at the same time with not a hint of siblance, it's how I like my high end to be, whch is why all my other speakers have ribbon drivers, but the piezos are just as good, listeing to them is pure joy.
Mick, even with 4, those piezos should be really loud, there is something funny going on, maybe one of them is bad and sucks all the power or one of them is wired out of phase? in fact one could have the wrong polarity wired internally, test them individually then in pairs next to each other by swapping leads one way will sound louder then the other, the louder way will be in phase.
Also I suggested mounting them to each side to overcome the narrow dispersion of the 10" and before I knew that wanted to use them while stationary, the rule is that you shouldn't cross 10" drivers lower than 1.6kHz because they start to beam above that frequency, so when stationary you're not getting good integraion between the two, and it might well produce frequencies up to 4kHz but they aren't going to be that pretty, so a low pass filter even just 1st order is not a bad idea.
If you cut the tweeter housing down as I suggested you can save 25% or more weight on each and be able to fit six in a line next to the woofer without looking out of place.
You might well get better integration between a line of SPH-30X and the 10", because you can cross them lower than the piezos, but getting enough volume out of them to match the 10" is going to be touch and go, I think you should try.
Well if you can cross over at such a high frequency, then you should be able to use a much smaller "tweeter".
Have a look at some small drivers here:
http://zaphaudio.com/smalltest/
http://zaphaudio.com/smalltest/compare.html
And tweeters here:
http://zaphaudio.com/tweetermishmash/
http://zaphaudio.com/tweetermishmash/compare.html
I'd recommend some of these neo tweeters crossed at 2500hz;
http://www.parts-express.com/pe/showdetl.cfm?Partnumber=275-035
Dayton ND20FB ($8) - Very smooth, usable response. Nifty rear mount design. Good distortion performance above 3.5kHz. Super small and super cheap but definitely usable. Note that this "F" version has ferrofluid. Old/Other tests available. Off axis response curves available. Tested April 2008.
I know they are only available in the US, but at this price, they might be affordable in the UK...
***It's so easy, everyone should be doing it***
those Dayton neos look great, but the Fs at 2070Hz is too high to cross at 2500Hz, the general rule is to cross one octave above Fs
and check this out instead of paying $10 for one, try 200!
http://www.parts-express.com/pe/showdetl.cfm?Partnumber=275-035
if shipping for me didn't cost $120 I'd get these in a flash
quote:Originally posted by gnome
those Dayton neos look great, but the Fs at 2070Hz is too high to cross at 2500Hz, the general rule is to cross one octave above Fs
and check this out instead of paying $10 for one, try 200!
http://www.parts-express.com/pe/showdetl.cfm?Partnumber=275-035
if shipping for me didn't cost $120 I'd get these in a flash
Hey gnome, not sure what you found at PE, but that's the same link for the Neo's that I posted?
Since the Celestion goes up to 4k, perhaps you could try crossing these higher ~ 3k as V-bro has suggested.
http://www.zaphaudio.com/tweetermishmash/Dayton_ND20FB-4-FR.gif
These are also for sale in Canada now (Solen now carries Dayton products):
http://www.solen.ca/pub/cms_nf_catalogue_fiche.php?id=2097&recherche=&numRows=&manufacturiers=42&niveau1=&niveau2=&niveau3=
***It's so easy, everyone should be doing it***
ANY driver placed in a line array becomes "smoother" or better said the distortion decreases dramatically because each driver has to move less to get the same SPL. I believe you when you say that this works out quite nicely with piezo's. Still I would dare to make a bet that a line array of decent tweeters outperforms the piezo array.... Only it'll cost much more....
------X-max. can be several meters on any driver, too bad it can be done only once...------
Been playing around with the piezos on their own but still not getting much out of them. Haven't tested every single one individually but I've tried various combinations without any change. I had a flash of inspiration when I thought the 4ohm wire wound resistor I was using in series might be acting as a low pass filter. I was pretty sure it was non-inductive but I swapped it for some metal film ones just in case, still no difference. Bottom line seems to be they're just not outputting much, and what they are producing isn't particularly clear. Seems to piss off the cat though :) Maybe these ones are just too cheap. At that price it was worth a try though.
I've decided to go with a couple of the Monacors V-bro suggested next. A pair of them won't break the bank and the size and weight are a big plus for a portable box.
quote:Originally posted by Scratchy
Hey gnome, not sure what you found at PE, but that's the same link for the Neo's that I posted?
[:I]
these
http://www.parts-express.com/pe/showdetl.cfm?Partnumber=269-810
quote:Originally posted by V-bro
ANY driver placed in a line array becomes "smoother" or better said the distortion decreases dramatically because each driver has to move less to get the same SPL. I believe you when you say that this works out quite nicely with piezo's. Still I would dare to make a bet that a line array of decent tweeters outperforms the piezo array.... Only it'll cost much more....
yeah maybe, but there simply isn't a lot of room up there
the ones I have sound so good that I simply couldn't imagine how they could sound better, it certainly isn't worth spending 20 times more to find out that there is a slight difference.
And you can't beat the simplicity of the piezo, so maybe spending another $50 on crossovers isn't that complicated but if you don't need to it is better.
quote:Originally posted by MickB
Been playing around with the piezos on their own but still not getting much out of them. Haven't tested every single one individually but I've tried various combinations without any change. I had a flash of inspiration when I thought the 4ohm wire wound resistor I was using in series might be acting as a low pass filter. I was pretty sure it was non-inductive but I swapped it for some metal film ones just in case, still no difference. Bottom line seems to be they're just not outputting much, and what they are producing isn't particularly clear. Seems to piss off the cat though :) Maybe these ones are just too cheap. At that price it was worth a try though.
that's very strange they should blow your head off at just moderate volume. Maybe you can take a picture of how you wired them up?
quote:
I've decided to go with a couple of the Monacors V-bro suggested next. A pair of them won't break the bank and the size and weight are a big plus for a portable box.
yeah those do look really great, pretty jealous of the amazing choice you have in europe. Wish we had Monacor here.
I've chopped the wires so I could pull the tweeters out of the box, gnome, but they were all wired properly in parallel. I double and triple checked as I know how easy it is to make stupid cock-ups with this sort of thing. The 4 ohm resistor was wired in series between one terminal of the amp output (I'm using a single channel for testing) and the common connection for one side of the tweeters. The other amp terminal was wired straight to the common connection for the other side of the tweeters. Before removing them I tried each bank of 4 individually, a few combinations of pairs and several individual tweeters on their own. With a single tweeter it sounded rather like sitting next to somebody on the train with their earbuds turned up very loud. With them all going it was a fair bit louder but nothing like there should be and there was no real clarity. I think they are just duff units for whatever reason.
quote:Originally posted by gnome
yeah those do look really great, pretty jealous of the amazing choice you have in europe. Wish we had Monacor here.
And not only that, but V-bro linked direct to very cheapest supplier in the UK :D
I'd like to make sure I've sorted out what I'm doing before I order a selection of capacitors for filtering. I realise I need a range of values so I can experiment and I'd like to get them all in one go as shipping per order looks like it could be expensive.
As I see it I've got several options using the AMP6. I'll definitely be using V-bro's input transformers with a PLLXO highpass filter to get rid of frequencies below 100 Hz before the amp. It also seems best to connect the 8ohm tweeters in parallel for a 4 ohm load to match the 4 ohm impedance of the 10".
1. Parallel both channels for a mono signal and use the 2.5k high pass filter V-bro suggested after the amp to protect the tweeters.
2. Use one channel for the 10" and the other for the tweeters in which case it seems more sensible to use the PLLXO to do all the filtering: 2.5k high pass for the tweeter channel, and 100 Hz high pass for the bass channel.
3. Same as 2. but parallel left and right signals between the input transformers and the PLLXO so that the tweeters are playing a composite signal from both stereo inputs instead of just one side. Not sure whether there's enough difference between channels for it to matter for this sort of application though.
BTW, I fed the specs of the Directed 6.5" ASBOMB speakers into WinISD and it looks like I can get much better bass performance out of them. Purely by luck I happened to get the box volume about right, but it seems they're better suited to a vented enclosure. Fiddling with port proportions gives a nice smooth peak of +3db at 140Hz, double the tuning frequency which Saturnus said is a good profile to aim for with outdoor speakers. That compares with a much higher tuning frequency and flatter profile with the sealed box which is already dropping below -0.5db at 140 Hz. I'm going to add input transformers and PLLXO too. I'm enjoying this game :)
And the audience likes the way you play it.
I'm not playing it so well actually, krilli. I just realised I made a very stupid mistake calculating the volume of the existing ASBOMB box: forgot to divide by 2 because it's triangular, not square [:I] Porting it will still improve bass response but only by a small amount. It did seem a bit too good to be true! On the other hand it shows that building a bigger box for it would make a big difference to the sound without adding a lot of weight.
:)
The part I like best is that I'll be dredging good info out of here for months to come.
Do pay well attention to the fact that paralleling the tweeters on just one filter changes the values of the capacitor and inductor! Have a look at the table I posted!
------X-max. can be several meters on any driver, too bad it can be done only once...------
Thanks V-bro, yeah I did realise that. Your table is most useful, nice to see all the values at a glance instead of having to calculate each one.
I'm inclined to go with the tweeters on one channel and the 10" on the other with the PLLXO handling all the filtering as it seems the most elegant solution to my uneducated mind.
Is there any practical difference between that method, and the alternative method of paralleling both channels with the crossover after the amp? Just for example, does the amp handle the power to the bass better if it's shared between channels?
The greatest advantage of PLLXO filtering is in low frequencies as a filter after the amp needs MUCH larger values. Thus bulkier, heavier and more expensive parts. Earlier on we were discussing the use of the SPH-30X and there is a big advantage to PLLXO filtering on these drivers...
In your case PLLXO might not have such great advantages. A tweeter you MUST filter after the amp as this protects it from for instance the turn on pops or other source related pops. The bottom end needs no real filtering with the 10", but if you would have wanted to you just would have needed to lower the value of the input cap on the amp, effectively creating a PLLXO filter...
Paralleling both channels could be a bonus for the bass, but could also diminish the quality of the high range. I would not be able to predict what will result in the best sound between both set-ups. The one channel tweeter, the other the bass set-up could still have the advantage of being able to use the PLLXO lowpass for the 10"... This saves an inductor and allows you to use a much smaller cap. There may be a small weight advantage there too...
And about the input transformers, with this set-up you ONLY need input step-up transformers in case the source has very low output voltage, but can drive enough current. For instance from an MP3 player, designed to drive headphones the transformers are a great upgrade to the sound. If you are connecting the source to the same battery as the amp there is another advantage to the transformers as they prevent ground loops from occurring. But you may not really need to compensate for a PLLXO filter any more...
Ah, that's filled in lots of gaps for me. In fact I think that's enough for me to make all my filtering decisions now, so I hope to stop pestering you all. Thanks for your patience!
The Monacor tweeters that V-bro recommended have arrived at last and they're excellent! Lovely clear sound. The dispersion angle is about 60 degrees with quite a sharp drop-off outside that so I think it needs one more. With one pointing straight forward (well, back really) and a couple angled at the sides that will give nearly 180 degree spread which is just about right.
Which leads to yet another question...
3x 8 ohm parallel will give a 2.7 ohm load. Should I attenuate the output to the tweeters to avoid overloading the amp? Or can I get away with it as there are no low frequencies (cross-over corner frequency is 2.5kHz ). BTW, I'll probably use an AMP11-LV for the final version as I've been looking into the LifePo4 packs that krilli found and I've realised I can run it off 24V with very little weight penalty. That's a 6db increase over AMP6 at 12V so I think it would be worth doing.
The cabinet is looking even better now. I'm sure you'll agree that this honest and unpretentious approach to design results in a simple elegance that could become a design classic. Should I paint it? Or would that be guilding the lily?
http://homepage.ntlworld.com/m.brown74/Photos/ASBOMBS/Test_Box-2.JPG (http://homepage.ntlworld.com/m.brown74/Photos/ASBOMBS/)
Nice :)
You should be able to get away with 2.7 Ohm impedance:
- The chipset in the Amp11-LV is rated for over 100W - current handling in the Amp11-LV is the same as in the "heavier duty" ones - the TP2050 datasheet indicates a 6 Ampere current limit. Using 24-28V from an 8-high LiFePO4 pack puts 4.5A max through 2.7 Ohm if I'm doing the math properly.
- An Amp6 should similarly be able to handle 2.7 Ohm at 12-14V, given proper cooling (IIRC current handling in this situation was calculated / estimated in the forum a while ago - look it up, it was good info).
Ps. - Credit should go to Saturnus for discovering batteryspace.com and telling us about the LiFePO4 goodies
I say paint it construction orange so you won't lose it in the forest
You could perhaps use a separate capacitor for each tweeter and keep the impedance somewhat milder for the amp, although like Krilli said I don't think it is a problem.
Great to hear the tweeters turned out good, I always feel bad when I advise something and read later that someone is not happy with it. But the advise was more than just a hunch, I have used these tweeters more often...and yes they are great value for money and have amazing sensitivity, however a little at the price of dispersion. But that's always a bit the case with horn loaded tweeters, even such small ones....
------X-max. can be several meters on any driver, too bad it can be done only once...------
Mildly off-topic:
I've never really understood why high dispersion is necessary. I've never been that bothered by the sound changing a bit when I'm off to the side of the speaker. Now I may just always have had speakers with good dispersal and am simply missing the comparison, and thus the point.
Now in a mini-PA like the system in the thread, of course you want reasonable sound dispersion ... but, yeah, can't explain my question better, but can you guys describe a situation where there wasn't enough dispersion? Or how dispersion saved the day?
Humbly yours.
Poor dispersion can be a big plus! You can radiate sound in the right direction, be near it yourself (as the musician for instance) without getting hearing damage...
With a system like this indeed dispersion in the high frequency range is wanted. Dispersion by the way is never such a big problem with bass as these waves naturally radiate more or less all around, where high frequencies -due to the small wavelength- suffer more from the speaker box "standing in the way" to radiate around. So you get either a cardioid radiation pattern, or a forward beam. If you want to hear full spectrum all around it is important that all wavelengths are radiated in all directions equally.
If you ever wondered why in frequency diagrams often several lines are printed? It's to show the drop-off at certain angles, so you know more or less what to expect...
http://www.linkwitzlab.com/images/graphics/p6.gif
------X-max. can be several meters on any driver, too bad it can be done only once...------
quote:Originally posted by krilli
Now in a mini-PA like the system in the thread, of course you want reasonable sound dispersion ... but, yeah, can't explain my question better, but can you guys describe a situation where there wasn't enough dispersion? Or how dispersion saved the day?
Humbly yours.
Well, my home speaker. The Fostex horns have absolutely crappy dispersion, and no more than 2 people next to each can fit in the "sweet spot". That doesn't matter though as we're usually not more than 2 people when there's actually serious listening going on.
In my Boominator however it is paramount to have ridiculously high dispersion, it has 100x100 degree tweeter horns, because the listening position changes constantly and is often at off-angles. And though you could say it's not actually hi-fi in the strictest sense of the word, it was very important for me that as many people as possible could have a reasonable listening experience.
quote:Originally posted by MickB
The cabinet is looking even better now. I'm sure you'll agree that this honest and unpretentious approach to design results in a simple elegance that could become a design classic. Should I paint it? Or would that be guilding the lily?
http://homepage.ntlworld.com/m.brown74/Photos/ASBOMBS/Test_Box-2.JPG (http://homepage.ntlworld.com/m.brown74/Photos/ASBOMBS/)
That is unquestionably one of the best looking & well constructed speaker enclosures I have seen in a long time.
I would put it on par with these speakers:
http://i28.tinypic.com/43rrp.jpg
;) [:p] [:o)]
***It's so easy, everyone should be doing it***
Scratchy, you did know this box was just the proto for this one right?
http://farm3.static.flickr.com/2482/3840703605_868a12c7e2.jpg
------X-max. can be several meters on any driver, too bad it can be done only once...------
I love the idea of gnome's thin skin box design but I've done some calculations and sadly I think it's not going to be practical for this application for a number of reasons. The back panel can't have much curvature as it has to sit against the skater's back so would have to be much thicker material. For the sides, PVC drainpipe 2.3mm thick weighs 3.5 kg/sq metre which is almost the same as the 9mm ply I've got. The 10" bass speaker covers much of the width of the box and would need a fairly deep mounting ring to give a flat mounting surface on the curved front panel. Taking those things together means it wouldn't be much lighter than an all 9mm plywood box but it would take a lot more work to build.
I've also got to consider the effects of falling over backwards onto it (which will happen sooner or later) and I think a 3mm thin skin box will be too fragile. It will also be awkward to repair as the bonding between plywood and PVC is likely to be a weak point, and if the PVC shatters it will mean a major rebuild.
By contrast 3.6kg/sq metre 9mm ply is very tough. I've fallen over backwards onto my Asbomb and just made a few gouges in the surface. And if it does get more seriously damaged it's quite easy to repair. I've worked out that I can make a 9mm plywood box about 1.8kg so the whole system should be under 5kg.
BTW, I've found that 9mm exterior ply is considerably lighter and stiffer than 6mm birch ply, although the density does vary a lot so it's worth selecting the lightest stuff. I've also found out that a highly effective way of annoying the staff at the local timber merchant is to weigh all 50 of their sheets of 9mm ply with a spring balance.
One of the other skaters has access to carbon composite materials at his university and we are having a discussion about maybe using that, but I suspect there will be too many obstacles to pull it off. It's a shame as double skin carbon with honeycomb core would make a wonderfully light and rigid speaker box!
quote:Originally posted by V-bro
Scratchy, you did know this box was just the proto for this one right?
------X-max. can be several meters on any driver, too bad it can be done only once...------
You did see my smiley face icons at the bottom of my post?
;) [:p] [:o)]
***It's so easy, everyone should be doing it***
quote:Originally posted by MickB
One of the other skaters has access to carbon composite materials at his university and we are having a discussion about maybe using that, but I suspect there will be too many obstacles to pull it off. It's a shame as double skin carbon with honeycomb core would make a wonderfully light and rigid speaker box!
That is Very Interesting. I must say I think you could get some kind of academic R&D grant for this, maybe putting it within reach.
Unfortunately, carbon fiber especially sandwich structures are much too brittle for this use. It will likely shatter if you fell on top of it. No serious speakers companies use it for very good reason. They use molded ABS plastic cabinets instead which have most of the qualities of wood while being easier (and cheaper in sufficient numbers) to shape into whatever shape you want.
I'd definitely recommend a plywood box for this as it's cheaper, lighter and much stiffer and resilient to abuse than any of the choices you have as a hobbyist/small scale manufacturer.
Plywood baffle+rear panel and PVC sides...
------X-max. can be several meters on any driver, too bad it can be done only once...------
That's very useful to know, Saturnus. In a perverse way I'm pleased about that as a carbon box would be a huge amount of work even if we could get the materials cheaply, and now I know I'm not actually missing out anyway. I've thought about all sorts of different materials and it always seems to come back to plywood. It's interesting that such old material technology can still provide the best combination of qualities.
quote:Originally posted by MickB
That's very useful to know, Saturnus. In a perverse way I'm pleased about that as a carbon box would be a huge amount of work even if we could get the materials cheaply, and now I know I'm not actually missing out anyway. I've thought about all sorts of different materials and it always seems to come back to plywood. It's interesting that such old material technology can still provide the best combination of qualities.
Look at it this way ... wood is a carbon fiber structure, one that has taken nature billions of years to perfect. :D
I used to work at a company that made specialized transport boxes for the military, aerospace and automotive industries. And all those boxes were made entirely of plywood because no other materials can even come close to matching woods properties when it comes to cost, stiffness, flexing, flexibility, resistance to stress, temperature and moisture, and the list just goes on and on.
The density of pvc is 1.4g/cm and density of plywood is ~0.5-0.7g/cm
so comparing 9mm ply to 2.3mm PVC, the PVC has the edge, you can make the ends from 9mm ply but I would still be tempted to make the outside face curved from 3mm,if you use internal braces than the side facing the back can be straight, in fact even the speaker side can be straight, but you you would have to add a third brace at least.
And I think if you fall the most likely thing to break is the woofer itself or the grill being mangled, you are using grill I presume?
Also I recommend you put the tweeters together because than you can use them as stationary speakers, placing them (speaker drivers) within 1/4 wavelength let's them couple the sound for a bonus 3dB,
ie you're getting 6dB, three from doubling the surface area and 3dB from sound coupling. adding a third in line is going to give you another 1-2dB so not that significant (although it could be just the right amount to make them sound more balanced)
put them next to the woofer if you can bear the width, or above or below the woofer if you can bear the height. Being able to use them while stationary is a better cause then dispersing HF sound for while they are moving around imo.
btw stiffness goes up by the cube of the thickness which is why 9mm is so much stiffer than 6mm, and also the reason why lateral braces are such a fait accompli no matter how puny they are.
The exact reason Gnome gave above is why I recommended 3-4 small 3" driver in line instead of one big. You'd also not need tweeters in such a set up.
Braces btw mostly only have to cope with stress in one direction/plane which is why they are effective almost no matter how thin they are.
quote:Originally posted by Saturnus
The exact reason Gnome gave above is why I recommended 3-4 small 3" driver in line instead of one big. You'd also not need tweeters in such a set up.
touch and go, four 3" drivers still wouldn't deliver the bass of a 10", but then you're not going to get that much bass from a small volume outside anyway.
We used to have plywood chairs in the school I was on back when I was a kid. The teachers one day went out to choose them and came back with a demo which was made of no more than 6mm thick bent plywood. The seat and back were out of one piece. A bit like this:http://img.tradekey.com/images/uploadedimages/offers/2/9/A1743406-20090414014429.jpg The salesman said that you could put the chair upside down and stand on the bend as a grown man, which was true. Our teachers literally tried to break it but couldn't manage to do so, even though they bounced up and down on it on a very smooth floor.
I used to have those http://23.media.tumblr.com/tumblr_kpfjunRMMb1qzukcpo1_500.jpg chairs, a good example of how strong plywood is.
I don't master making plywood at all, so I won't know what glue and methods are best to use for a strong construction. I agree with Saturnus that it is probably the best material for the whole structure, but also quite a lot of work if you have to laminate the plywood yourself. I don't know if you can make the tight bends on the sides from pre-laminated plywood. Maybe it can crack, perhaps in time. I do feel it would be possible to make VERY strong home-laminated plywood from single layers of wood.
For the bracing you can probably get away with quite thin material, but the surface to glue it to the sides it could become uncomfortably thin. Probably best not to overdo it and when there are a bunch of large holes in them the weight factor is not going to be problematic after all... Perhaps you can use some tempex to glue the braces on first, then you'll have some more flesh and it would perhaps keep them better in place.
Anyways, perhaps you can make a small model or piece with a bend to try out the strengths of some different methods and woods first...
------X-max. can be several meters on any driver, too bad it can be done only once...------
quote:Originally posted by gnome
The density of pvc is 1.4g/cm and density of plywood is ~0.5-0.7g/cm
I've managed to find some ply that is even lighter, less than 0.4g/cm^3. The weight/cm^2 is virtually the same as the PVC, I've weighed both. I agree that PVC drainpipe is a very useful material as it's so versatile and can be heat moulded. I use it for receiver cradles, stiffeners and various other components.
quote:Originally posted by gnome
I would still be tempted to make the outside face curved from 3mm,if you use internal braces than the side facing the back can be straight, in fact even the speaker side can be straight, but you you would have to add a third brace at least.
It would be nice to save more weight but I'm surprised that even well braced 3mm could be used flat, that's just going by intuition though which is notoriously unreliable. Is it possible to say what the maximum unsupported square area is that one can get away with using flat 3mm ply? It's easy to noticeably depress the centre of quite a small flat panel of 3mm with just finger pressure. I don't know how much pressure it has to resist in practice but I've noticed that people use bracing even when using 12mm or 19mm MDF.
quote:Originally posted by gnome
And I think if you fall the most likely thing to break is the woofer itself or the grill being mangled, you are using grill I presume?
I've had lots of practice at falling :) After nearly 20 years skating I've found that one very rarely lands flat on ones back, the most likely area of impact will be somewhere along the bottom edge of the baffle, and usually you manage to twist round a bit after going over backwards in order to try to get a hand down so it's probably going to be on one corner. Trouble is it's very hard to be sure how it will stand up without actually building it and falling on it to see what happens. I do like to test stuff before I build a final version but I'm not quite as dedicated as that.
I will be using some sort of grill for protection in a crash and while transporting it around. I've got a few ideas but I'm sure I'll think of a whole string of questions when I get to that stage ;)
quote:Originally posted by gnome
Also I recommend you put the tweeters together because than you can use them as stationary speakers, placing them (speaker drivers) within 1/4 wavelength let's them couple the sound for a bonus 3dB,
ie you're getting 6dB, three from doubling the surface area and 3dB from sound coupling. adding a third in line is going to give you another 1-2dB so not that significant (although it could be just the right amount to make them sound more balanced)
put them next to the woofer if you can bear the width, or above or below the woofer if you can bear the height. Being able to use them while stationary is a better cause then dispersing HF sound for while they are moving around imo.
It's primary purpose is for moving along in a street skate, any static use will be very occasional and shouldn't detract from it's main function. Obviously 360 degree dispersion at all frequencies would be ideal but given that's impractical with something this size and weight then 180 degrees is a good target to aim for. I've tested it outside and it does pretty well manage that, or rather it will do with 3 tweeters. There's no need for louder high frequency in any one direction as a single tweeter is a good match in loudness for the 10".
quote:Originally posted by Saturnus
The exact reason Gnome gave above is why I recommended 3-4 small 3" driver in line instead of one big. You'd also not need tweeters in such a set up.
I would like to try that sometime if only to see how it compares to the 10" bass + tweeters. It would make a very light compact unit. For now I'm committed though as I've bought the speakers, and I am very pleased with the way it sounds. You guys have so many good ideas backed up with sound technical knowledge that I want to try them all! If only I had unlimited funds...
Once I've finished this unit I'm hoping that one of the other skaters will will buy it off me for the cost of the parts, it's the way I'm doing it with the Asbombs. I'm mainly interested in building them and then seeing them in action, I don't particularly want to retain ownership as long as they're being used on the street skates. If that works out then I'll be able to afford to build the 4x 3" vertical array. Just out of interest, what sort of dispersion angle can one expect from that? It does need to be around 180 degrees for street skates. In the meantime I might knock up a box to take a spare pair of those Directed 6.5" concentrics in a vertical array just to see what it sounds like.
I've tried doing some research on vertical arrays but still haven't got a clear picture of what's going on. I found this (http://www.meyersound.com/support/papers/line_array_theory.htm) page on line array theory but it's a bit dense for me and is talking about huge PA systems and seems to concentrate more on the vertical focusing rather than horizontal spread.
quote:Originally posted by V-bro
I don't master making plywood at all, so I won't know what glue and methods are best to use for a strong construction. I agree with Saturnus that it is probably the best material for the whole structure, but also quite a lot of work if you have to laminate the plywood yourself. I don't know if you can make the tight bends on the sides from pre-laminated plywood. Maybe it can crack, perhaps in time. I do feel it would be possible to make VERY strong home-laminated plywood from single layers of wood.
the glue to use is a polyurethane construction adhesive, one that cures with moisture an expands slightly in the process, it has this amazing ability to ceep into the wood making an amazingly strong bond, whch is stronger than PVA wood glue, and much more flexible, so the glue joint can take a lot more stress, you also get a 100% airtight seal, it also sticks great to non porous surfaces like plastic or metal, I just love the stuff and have comptletely abandoned wood glue.
I've bend 3.4mm ply over 12cm radius, but that was pretty much the limit, you can get ply wood that has all laminations with the grain at the same orienation, which is designed to be bend.
quote:Originally posted by MickB
quote:Originally posted by gnome
The density of pvc is 1.4g/cm and density of plywood is ~0.5-0.7g/cm
I've managed to find some ply that is even lighter, less than 0.4g/cm^3. The weight/cm^2 is virtually the same as the PVC, I've weighed both. I agree that PVC drainpipe is a very useful material as it's so versatile and can be heat moulded. I use it for receiver cradles, stiffeners and various other components.
quote:Originally posted by gnome
I would still be tempted to make the outside face curved from 3mm,if you use internal braces than the side facing the back can be straight, in fact even the speaker side can be straight, but you you would have to add a third brace at least.
It would be nice to save more weight but I'm surprised that even well braced 3mm could be used flat, that's just going by intuition though which is notoriously unreliable. Is it possible to say what the maximum unsupported square area is that one can get away with using flat 3mm ply? yeah hard to say
quote:
quote:Originally posted by gnome
Also I recommend you put the tweeters together because than you can use them as stationary speakers, placing them (speaker drivers) within 1/4 wavelength let's them couple the sound for a bonus 3dB,
ie you're getting 6dB, three from doubling the surface area and 3dB from sound coupling. adding a third in line is going to give you another 1-2dB so not that significant (although it could be just the right amount to make them sound more balanced)
put them next to the woofer if you can bear the width, or above or below the woofer if you can bear the height. Being able to use them while stationary is a better cause then dispersing HF sound for while they are moving around imo.
It's primary purpose is for moving along in a street skate, any static use will be very occasional and shouldn't detract from it's main function. Obviously 360 degree dispersion at all frequencies would be ideal but given that's impractical with something this size and weight then 180 degrees is a good target to aim for. I've tested it outside and it does pretty well manage that, or rather it will do with 3 tweeters. There's no need for louder high frequency in any one direction as a single tweeter is a good match in loudness for the 10".
that is a pretty impressive tweeter, but Im a little nervous about the low crossover, especially if you want to use the amp11, I think you have a good chance of rooting the tweewer through over excursion.
the other problem is that placing the tweeters the way you have does denature the sound somewhat, it would be better to put them at an angle to each other I'm on a netbook so I can't draw a picture for you, but like this /\, but at 90deg to each other, then you will geta good 150 deg of dispersion and you get the mutual coupling benefit and just a bit of combfiltering above 9kHz which is an acceptable compromise.
quote:Originally posted by gnome
the glue to use is a polyurethane construction adhesive, one that cures with moisture an expands slightly in the process, it has this amazing ability to ceep into the wood making an amazingly strong bond, whch is stronger than PVA wood glue, and much more flexible, so the glue joint can take a lot more stress, you also get a 100% airtight seal, it also sticks great to non porous surfaces like plastic or metal, I just love the stuff and have comptletely abandoned wood glue.
I've bend 3.4mm ply over 12cm radius, but that was pretty much the limit, you can get ply wood that has all laminations with the grain at the same orienation, which is designed to be bend.
I agree.
Here in North America it goes by the name of "gorilla glue".
I dampen the joints with water before I apply this glue, and it grabs the pieces like a gorilla once it's cured. Very sticky stuff.
***It's so easy, everyone should be doing it***
try and find something like this Scratchy it's an order of magnitude more awesome
http://www.stickwithpl.com/Products/detail.asp?PLProductID=14
gorilla glue while essentially the same stuff is runnier, bubbles more and isn't as strong as the stuff you apply with a caulking gun, which dries very hard.
Problem with that is that for curved shapes we actually don't want a hard bond but a flexible one as it absorbs vibrations better. Gorilla glue is the best you can get ... but only if you can't get PVA hardener. PVA glue mixed with hardener creates an even stronger bond than gorilla glue and remains much more flexible.
The problem with very hard glues is that they are brittle, under contineous vibration it cracks, and loses it's bonding.
I flexible joint is good, but when joining porous surfaces like MDF, I like to get some penetration of the glue into the wood.
Because it is so thick, the PL will only bond to a very thin layer of the MDF.
It will work much better with hard or non-porous materials.
As for gluing plywood, it is not as porous as MDF, and so might be a better choice for that.
PL is also ideal for sealing the inside corners of an MDF speaker box once it's all glue together.
***It's so easy, everyone should be doing it***
quote:Originally posted by Scratchy
I flexible joint is good, but when joining porous surfaces like MDF, I like to get some penetration of the glue into the wood.
As I said, gorilla glue is excellent but PVA glue with PVA hardener is far superior (and at least 3 times more expensive per volume, the hardener costs an arm and a leg) for all these purposes. PVA hardener is an epoxy resin specially developed to react with PVA glue.
quote:Originally posted by gnome
the other problem is that placing the tweeters the way you have does denature the sound somewhat, it would be better to put them at an angle to each other I'm on a netbook so I can't draw a picture for you, but like this /\, but at 90deg to each other, then you will geta good 150 deg of dispersion and you get the mutual coupling benefit and just a bit of combfiltering above 9kHz which is an acceptable compromise.
That's interesting. So for this coupling effect to happen the speakers don't have to be pointing the same way? Does that mean I've inadvertently achieved it with the triangular box of the Asbombs, or do the speakers need to be closer to the front corner? (picture on page 1 of this thread). Does it work for any angle?
I've got another tweeter on the way anyway, so if I mounted the 3 of them an the front and both sides of a small cube so they're at 90 degrees to one another would I get more than 180 degree dispersion as well as a db boost from the coupling effect?
quote:Originally posted by gnome
the other problem is that placing the tweeters the way you have does denature the sound somewhat, it would be better to put them at an angle to each other I'm on a netbook so I can't draw a picture for you, but like this /\, but at 90deg to each other, then you will geta good 150 deg of dispersion and you get the mutual coupling benefit and just a bit of combfiltering above 9kHz which is an acceptable compromise.
This is indeed quite a smart method, you'll have the centre of the tweeters very near eachother, so indeed minimal comb filtering and maximal coupling. It would have the extra benefit of them being somewhat enclosed to protect them from curious fingers and damage from dropping. A small piece of mesh would than even protect them completely!
------X-max. can be several meters on any driver, too bad it can be done only once...------
quote:Originally posted by V-bro
It would have the extra benefit of them being somewhat enclosed to protect them from curious fingers and damage from dropping. A small piece of mesh would than even protect them completely!
BTW, the only reason they're sticking out of the sides on the prototype is because it's a quick and easy way of mounting them so I could experiment with different angles. For the real thing I was planning on mounting them flush into the chamfered edges of the box with another in the baffle pointing straight forwards (backwards).
I knew it wasn't the definitive construction... Of course you would mount them much alike the piezo's, you already knew about the coupling and comb filtering. But what Gnome mentioned does seem to make sense...
------X-max. can be several meters on any driver, too bad it can be done only once...------
Aha! It's just clicked, gnome's talking about a 90 degree internal angle between the tweeters, so they're angled towards each other and then the pair is recessed into the baffle. I thought he meant they were angled away from each other. Your reply makes sense now. Time to annoy the neighbours with another test I think.
to be truthful you will get better coupling when they are pointing the same way, the crossfiring is mainly for bringing the centres together as v-bro said, it makes a significan't difference to the sound quality and course dispersion, but no need to go overboard with that since the woofer is also limited as it nears the crossover point.
quote:Originally posted by gnome
Im a little nervous about the low crossover, especially if you want to use the amp11, I think you have a good chance of rooting the tweewer through over excursion.
Is it possible to quantify the risk, however roughly? The tweeter is rated at 50Wrms and Monacor's recommended crossover frequency (12db) is 2kHz. I've got a 12db crossover at 2.5kHz. It will be turned up loud a lot of the time though. We turn these units up until they start to distort noticeably then ease back a bit, but the music varies a lot in level and content, and when you're concentrating on skating in a noisy environment it might not get noticed for a while. But I would like to keep the crossover as low as possible to improve dispersion at the top of the mid range, those same frequencies will be getting very directional if the 10" is producing them. I do want to keep wide dispersion near the top of the list of many conflicting requirements.
they reccomend 2K? hang on a minute
ok I just checked the spec, the resonant frequency is 1200Hz, or some reason I thought it was
much higher, so forget what I said, 2.5Khz is just fine.
that really is a superb tweeter
Well, the speaker box is built. I'm sure there are lots of mistakes from an acoustic point of view but there comes a point where one just has to get on and build it. I'd be glad to hear any criticisms though as I may decide to build another one. I was tempted to go with gnome's suggestion of a pair of tweeters set at 90 degrees, but in the end decided to go with 3 splayed drivers to achieve wider dispersion at the expense of some efficiency. I've arranged them on a common arc to minimise comb filtering in the overlapping regions, does that make sense? The 2 slightly larger holes at the corners of the baffle are for ports. The box has a slight "ring" to it when it's tapped and I wonder whether I've inadvertently constructed a bell with the ring of bracing around the woofer opening and too much symmetry. It is very rigid though.
I've also been asked to make a couple more Asbombs, one is identical to the previous units but the other one has a larger volume speaker box which according to WinISD will give better bass response in return for a little more weight. Be interesting to hear what difference it makes in practice.
Sorry, bit late. BUT: it looks great! Just marvelous!
Thanks, FFF! The 10" mono AMP11 unit has been on hold while I've been building a couple more of the original Asbomb units for other people. I'll post some pics later.
Just about ready to get back to the 10" mono unit now and I'm getting all excited about it again.
alexlangberg
22-Jul-10, 12:50
Hey!
This is pretty cool, I was working on pretty much the same idea for festival use, also based on the 10" NEO from Monacor. It's so light and not too big. I have one question though. How do you drive a single bass from the AMP6? What is the best solution for this? For instance if I want to connect 1 10" bass and 4 piezos. Could you
A) Just connect the bass to e.g. the right channel and hope it works? Would it require a resistor or something on the other channel?
B) Bridge the amp in some way so it only plays mono? Stereo doesn't make much sense anyway?
C) Do something completely different?
Anyways, very cool design, I don't think a lot of people have the patience for that kind of plywood work. :)
Could you
A) Just connect the bass to e.g. the right channel and hope it works? Would it require a resistor or something on the other channel?
B) Bridge the amp in some way so it only plays mono? Stereo doesn't make much sense anyway?
C) Do something completely different?
You can do A, yes - i.e. if you're talking about using only one channel. Then you need to "close" the unused channel by jumpering its input to ground, and putting an artificial-load resistor on the output. Not sure of the value, but somewhere around 1-10KOhm I'd think? It's answered in the forum.
You can basically do B. You can't bridge those amps that are suitable for battery-power, but you can parallel them into a single load AFAIK. But you only gain anything by paralleling if you're creating a load with an impedance lower than 4 Ohm, to push more current through.
C? Maybe :)
What's the impedance and efficiency of all the drivers?
alexlangberg
22-Jul-10, 18:46
Thanks for the fast reply, krilli.
I was thinking about using a Monacor SP-10A/250NEO (http://www.monacor.dk/produkter/hoejttalerenheder-pa10-12/vnr/104600/) along with 4-6 Monacor MPT-016 (http://www.monacor.dk/produkter/hoejttalerenheder-padiskant/vnr/110490/) piezos. If this is stupid, let me know, it just seems like they might fit well, taking their frequency response into account. It also seems like people prefer several piezos per channel before they sound good at which point the piezo efficiency of the MPT-016 seems to fit well.
Well, I'm honestly not very fond of solution A. It seems a bit "wrong" to me, because you might loose important parts of the music. A panning bass would just go silent when moving to the other channel and such. I doubt it's much of a problem, I honestly have no idea how well it would work in practice.
I have very little knowledge about solution B, I'm pretty sure the 99dB efficiency of the bass should push enough sound in a standard configuration so the only reason I'd do it is to make sure it plays all the music from both channels so you don't miss any of the sound.
Solution C is the magical solution where the AMP6 is easily converted to a stereo input mono output amp. Or where the channels are combined before the amp, feeding both channels on the amp a combined mono signal of the source. I have no idea. :)
Thanks for helping. :)
NP :)
One quick thing: It's pretty easy to create a mono sum from a stereo input: Just solder the L and R wires together :) ...
alexlangberg
22-Jul-10, 19:23
NP :)
One quick thing: It's pretty easy to create a mono sum from a stereo input: Just solder the L and R wires together :) ...
Hahaha, well, I guess that was solution C. :P
I wasn't sure if this was possible. So this means that I could basically take a simple switch and connect it to the L and R of the input jack of the AMP6? So that if I flip the switch the two channels are "shorted" and thus combined to mono?
I guess this should then just be combined with plan A and everything should work? :)
Yes indeed, that would be my understanding!
I must say that I only know very little about how to connect multiple high- and low-frequency drivers together, so I'll have to explicitly say that I have only been talking about about what's possible, amp-wise.
You can bunch all the drivers together on one channel, or you can have the woofer on one channel and the tweeters on the other, or a mix ... I have no idea what will work out best.
EHm, you can connect together whatever amplifier inputs, but certainly not combine source outputs. This should really be buffered in some way, or impedance corrected at least...
Some resistors can do the trick, I think you might be able to parallel the input channels right on the chip pins... There will be the 2.5V bias voltage, but since the potential is the same I guess this is OK. At least then you will have the 22K input gain resistors in place, so the source outputs will at worst see each other with 44K resistance in between....
Yes, if you are going to parallel the outputs, then they need to have the exact same signal source.
Any differences in output between the 2 channels when they are paralleled will become instant heat in the TA2020 chip.
That's also why it's also important to make sure the DC offset on each channel is as close to 0V as possible before you connect them together.
alexlangberg
23-Jul-10, 12:10
In order not to hijack this thread any further, I moved my discussion to here (http://www.41hz.com/forums/showthread.php?3479-Stereo-amp-mono-output). :)
alexlangberg
27-Jul-10, 02:15
Another question, what do you use for the straps? Did you just take apart some backpacks for the straps or is it something special? :)
Hi alexlangberg, thanks for the kind words! I make the harnesses myself. One of my day jobs is repairing windsurf sails which means I have a couple of industrial sewing machines in my workshop and all the necessary bits and pieces. No reason why it shouldn't be done on a domestic machine though as the materials are not very thick.
I use foam cut from a camping mattress for padding, polypropylene webbing (strong enough for this purpose and much lighter than polyester) and black nylon cloth. I make the individual straps by cutting a strip of foam, stick it to a length of webbing with double sided tape, wrap the cloth around them both then run a couple of rows of stitching along the length to hold it all together. The vertical part of the shoulder straps have plastic stiffeners pushed up the middle to stop the music box sagging away from your back. I've also put a short stiffener in the back of the waist belt (as that also carries the battery pack on the small units) and I've made a pivot where the waist belt joins the back/shoulder straps so that there's less restriction when you bend your body.
All the stiffeners are made from PVC roof gutter which I heat with a hot air gun until it goes soft, flatten it out and then cut to whatever shape I want. It's incredibly useful stuff as you can heat mould it to whatever shape you want, I've used it for all sorts of fittings on these systems.
Cutting up an existing back pack would be a lot less work though. I plan to use that method in the future as we're planning a much simpler design which hopefully retains most of the good features of the Asbombs but cuts down on the huge amount of work that goes into each unit at the moment.
Just say if you want more details of the harnesses and I'll post some pictures of the construction.
Re parallel inputs: I've used a pair of Neutrik NTE-4 audio transformers (http://uk.farnell.com/neutrik/nte4/transformer-audio-1-4/dp/724609) for the AMP11/Monacor 10" mono system that V-bro recommended. They're small and light and isolate the source outputs from one another and also give a 1:4 signal boost.
alexlangberg
28-Jul-10, 14:58
Wow, I was expecting something along the lines of "Yeah, I make the harnesses myself". :D
I think I'm going to use a cut up backpack or something like that. Unless somebody has a better idea. I was thinking of maybe using the harnesses bodyguards use for guns/walkies. I'm not sure yet.
Also, I'd like to throw the E-MU Pipeline (http://www.thomann.de/gb/emu_pipeline.htm) into the idea box here. They are battery-powered audio transmitter/recievers with a delay of only 10ms and a range of 15m. They might not be useful for your project as I can imagine you need much more range but these units would be very useful for smaller projects. :)
By the way, what do you use to dummy load an AMP6 channel? Just a resistor?
Finished at last! Well, it'll probably need a few tweaks but it's ready for it's first outing at Bognor Carnival this weekend. Huge thanks again to everybody on this forum for all the help and advice, it wouldn't have turned out nearly as well without your invaluable input. And of course to Jan for his excellent amplifier! Particular thanks to V-bro for his fine workmanship building the AMP11 and for recommending the tweeters and all sorts of other stuff.
I'll write up a proper article later but in the meantime here are some pics:
After much trial and error and frustration I've at last worked out how to paint things properly!
http://homepage.ntlworld.com/m.brown74/Bassbomb/Painted_box.JPG (http://homepage.ntlworld.com/m.brown74/Bassbomb/)
Amp bolted to heat transfer block and heatsink.
http://homepage.ntlworld.com/m.brown74/Bassbomb/Amp_Heatsink.JPG (http://homepage.ntlworld.com/m.brown74/Bassbomb/)
Battery pack and cross over. For reasons I'll explain in the article I've gone with 20 cells arranged as 2 banks of 10 cells in series for 24V nominal. Each bank will be charged separately.
http://homepage.ntlworld.com/m.brown74/Bassbomb/Cross_over.JPG (http://homepage.ntlworld.com/m.brown74/Bassbomb/)
These are all the parts made from various sizes and thickness of PVC pipe and gutter. It's incredibly useful stuff!
http://homepage.ntlworld.com/m.brown74/Bassbomb/PVC_parts.JPG (http://homepage.ntlworld.com/m.brown74/Bassbomb/)
Small surface mount LEDs are glued to the underside of the grille with wires soldered directly to them. As their forward voltage is only 2V and I've got 24V supply they all run in series so total current draw is only 25mA. I've sprayed the speaker cone silver which acts as a reflector.
http://homepage.ntlworld.com/m.brown74/Bassbomb/Grille_LEDs.JPG (http://homepage.ntlworld.com/m.brown74/Bassbomb/)
http://homepage.ntlworld.com/m.brown74/Bassbomb/Harness.JPG (http://homepage.ntlworld.com/m.brown74/Bassbomb/)
http://homepage.ntlworld.com/m.brown74/Bassbomb/Internals-1.JPG (http://homepage.ntlworld.com/m.brown74/Bassbomb/)
http://homepage.ntlworld.com/m.brown74/Bassbomb/Internals-2.JPG (http://homepage.ntlworld.com/m.brown74/Bassbomb/)
Bit of a squeeze for all the wiring as I decided to add a digital voltmeter after I'd built the box. One of the small boards screwed to the bottom of the recess carries the input transformers and the other is the low voltage cut-off which protects the batteries from being damaged by over-discharging. The cut-off consists of a MAX8212 voltage monitoring chip and a small relay controlled via a transistor. When the voltage drops below about 21.5V the MAX8212 cuts the power to the transistor which in turn cuts off the relay. The chip is supplied via a diode with a capacitor across the supply which acts as a reservoir so that the cut-off isn't triggered by momentary voltage dips caused by big bass notes.
http://homepage.ntlworld.com/m.brown74/Bassbomb/Internals-3.JPG (http://homepage.ntlworld.com/m.brown74/Bassbomb/)
http://homepage.ntlworld.com/m.brown74/Bassbomb/Bassbomb-1.JPG (http://homepage.ntlworld.com/m.brown74/Bassbomb/)
http://homepage.ntlworld.com/m.brown74/Bassbomb/Bassbomb-2.JPG (http://homepage.ntlworld.com/m.brown74/Bassbomb/)
http://homepage.ntlworld.com/m.brown74/Bassbomb/Bassbomb-3.JPG (http://homepage.ntlworld.com/m.brown74/Bassbomb/)
http://homepage.ntlworld.com/m.brown74/Bassbomb/Bassbomb-4.JPG (http://homepage.ntlworld.com/m.brown74/Bassbomb/)
I added the voltmeter due to the necessity of charging the 2 banks of batteries separately. There's a 3 position toggle switch above the meter which selects the reading for the individual banks in the left and right positions, with the middle position showing the total supply voltage. It makes it easy to check that both banks have been charged before use, and the total voltage reading is useful for judging the remaining run time. The rotary ON/OFF switch incorporates a disk which covers the 2 charging sockets. The 2 holes ensure that the sockets are only accessible in the OFF position. The switch below the voltmeter is for the lights.
http://homepage.ntlworld.com/m.brown74/Bassbomb/Bassbomb-5.JPG (http://homepage.ntlworld.com/m.brown74/Bassbomb/)
http://homepage.ntlworld.com/m.brown74/Bassbomb/Bassbomb-6.JPG (http://homepage.ntlworld.com/m.brown74/Bassbomb/)
http://homepage.ntlworld.com/m.brown74/Bassbomb/Bassbomb-7.JPG (http://homepage.ntlworld.com/m.brown74/Bassbomb/)
http://homepage.ntlworld.com/m.brown74/Bassbomb/Lights.JPG (http://homepage.ntlworld.com/m.brown74/Bassbomb/)
Wow that's just incredible workmanship.
It must have taken a few hundred hours of designing, building and finishing.
I really like the charging socket interlock disc, ingenious but simple.
Kudos Mick - NICE!!!
That is just AWESOME craftsmanship!!http://www.radioparadise.com/scripts/tiny_mce/plugins/emotions/img/icon_notworthy.gif
I'm glad I could be of some help to this project!http://www.radioparadise.com/scripts/tiny_mce/plugins/emotions/img/icon_hug.gif
The charge socket locking disc is indeed a very smart feature! I guess one gets creative when one is in need...http://www.radioparadise.com/scripts/tiny_mce/plugins/emotions/img/icon_frustrated.gif
So, how's the sound and how are the responses?http://www.radioparadise.com/scripts/tiny_mce/plugins/emotions/img/icon_yell.gifhttp://www.radioparadise.com/scripts/tiny_mce/plugins/emotions/img/icon_eyes.gifhttp://www.radioparadise.com/scripts/tiny_mce/plugins/emotions/img/icon_guitarist.gifhttp://www.radioparadise.com/scripts/tiny_mce/plugins/emotions/img/eusa_dance.gif
Thanks guys! I really appreciate the comments coming from people who know about this stuff. Must admit I was quite pleased with the interlock, one of those little flashes of inspiration which seem to come out of nowhere but which are probably subconsciously stolen from somewhere else :) I'm trying to make these units as simple, reliable and foolproof as possible as they're often used by people who aren't technically minded so I try to arrange it so that the operator has to make a very special effort to get things to go wrong. Ideally all the attention they should ever need is recharging and occasional repair of accident damage, until components eventually wear out of course. I don't think I've quite achieved that yet but hopefully it's getting close. Having the 2 banks charged seperately isn't ideal in that regard but there are good reasons for doing it with this particular unit.
Regarding the sound I've been alternating between being absolutely delighted and mildly disappointed, and at the moment I'm somewhere in the middle but mostly pleased. That probably sounds rather odd so I'll explain...
I've never been a hi-fi buff so most of this is new to me. For me music is an accompaniment to my other activities so it's either my mp3 player and earbuds or the sound systems on the street skates or at roller discos. When I hear a nice hi-fi system at friends' houses I can tell they sound a lot better than what I'm used to but I don't really know what it is I'm hearing.
With this system the test box sounded very good to me because it was a lot better than the Asbombs. The clarity is excellent and it's also very loud! Mission accomplished, or so I thought. I then built the final box and tested it again but still without having sorted out the cross-over properly, it was really just something to protect the tweeters from low frequencies. After some more listening I realised that it wasn't a very flat response and there was too much midrange. It actually sounded better with the woofer connected in opposite polarity which I assumed was because of overlap between the low and high ranges and it sounded better to have it cancelling than reinforcing.
Next step was a proper cross-over with values calculated for 2.25khz and that sounded much better so I was delighted again. More listening and I started to realise that although there's a hefty thump down to a certain frequency it was rather lacking in bass. Some tunes sounded excellent, balanced throughout the range and very clear, but others sounded rather "shouty" (for want of a better word). I'd used a pair of 50mm dia x 45mm long port tubes as WinISD Pro showed a fairly flat frequency response curve with them. So I made some 35mm and 25mm long tubes and found it sounds best with the 25mm even though WinISD Pro shows a 2db hump above 100hz.
Just out of interest I tried connecting the woofer in opposite polarity again, and again it sounded better but it's hard to put my finger on what the difference is. I've tried it with and without stuffing and I think it sounds best with a moderate amount against the back of the box but it's very hard to be sure without a proper A/B test. Filling it with stuffing seems to deaden the sound a bit.
I'm a numbers sort of bloke, I just don't trust these subjective judgements when trying to make incremental improvements and I like to be able to measure things so I decided to be a bit more methodical. I used a tone generator to create a series of sound files of single frequency tones and played them in turn starting at 40hz and working up. It's still subjective but at least it's easier to judge the relative volume of single notes. That was interesting as the response appears to be far from flat with a very pronounced peak around 500hz which would probably explain the shoutiness in some songs. As expected with this size box there's not a lot below 100hz which is fine as it's meant for outdoor use. Interestingly it sounds quite flat over the cross-over region even though the speakers are still connected in opposite polarity. I always assumed that speaker manufacturers all use the same polarity convention but now I'm not so sure. Above 3k it's noticeably louder so presumably that means I should attenuate the tweeters, although the high frequencies don't seem too intrusive when listening to music.
Here are the curves from WinISD Pro. First with 45mm long port tubes...
http://homepage.ntlworld.com/m.brown74/Bassbomb/Freq_curve_45mm_vents.jpg (http://homepage.ntlworld.com/m.brown74/Bassbomb/)
...and with 25mm tubes - V-bro, look away now ;)
http://homepage.ntlworld.com/m.brown74/Bassbomb/Freq_curve_25mm_vents.jpg (http://homepage.ntlworld.com/m.brown74/Bassbomb/)
Finally here's a very approximate sketch of what I hear when I play the series of single tones through my PC soundcard (SB AudigyII). Don't take too much notice of the db numbers as it's just a guess but it does give a rough idea of the shape. The high end peak is probably less pronounced than I've shown but the 500hz peak is very noticeable. Any ideas what might be causing it? Presumably it's some sort of resonant frequency caused by the box.
http://homepage.ntlworld.com/m.brown74/Bassbomb/Bassbomb_freq_curve.jpg (http://homepage.ntlworld.com/m.brown74/Bassbomb/)
LOL!!
Ok, so what's the schematic of your speaker filter plus drivers? Can you draw that up for me? I'm quite an expert on speaker filters if I may say so, maybe I can help out there!
Also the response of a portable system is always hard to judge because the environmental acoustic properties are different all the time, and usually not very ideal. It being a backpack unit already gives away that the bass will suffer the most of this. For outside response bass needs more power compared to an in room system. Especially when you take away that last bit of acoustical reinforcement, which is the floor. What you are actually looking for is a system that shows a good free air response with virtually no help of environmental acoustics...
I have made some very small portable speaker systems with astonishing bass for the size, but then mainly when placed on the floor. When lifted up higher than a foot they start giving in on the bass response. It's still good, but not great either...
There is often a lot to improve on when efficiency can be sacrificed. Just shape down the excessive tones an you can be amazed! It's more or less a compromise between the two. That's why I like active systems and choose the drivers I expect that need most filtering in a very high efficiency. This enables much more possibilities to shape the sound to suit the needs.
I'm sure this system is about as perfect as can be, but still a tiny change of an overlooked detail can make all the difference. So I hope this is the case and we can find more improvement!
Cheers mate!
Also the neutrik transformers, did you place an RC network on the secondaries?
Here's a copy of the NTL1 datasheet:
http://i.imgur.com/4Avyn.jpg
Especially the 150pF and 15K network I am talking about here. For the NTE4 other values apply and I have personally tailored a network that makes them nicely linear. You need 22nF and 2.7K for them. The NTE4 doesn't really need the 1.65K input resistors and 680pF decoupling caps for as far as I have experienced...
I think this will tame a lot of the high range sharpness!
Here's the cross over filter schematic. Frequency is 2.25kHz. The 3 tweeters in parallel have 2.67 ohm impedance so I added a 1 ohm resistor to bring it up to 3.67 ohm. The resistor I'm using is one of these (http://uk.farnell.com/tyco-electronics-cgs/mpc1081r0j/resistor-10-watt-1r0-5/dp/1265148?Ntt=1265148) which is allegedly non-inductive.
http://homepage.ntlworld.com/m.brown74/Bassbomb/Crossover_schematic.jpg (http://homepage.ntlworld.com/m.brown74/Bassbomb/)
There is often a lot to improve on when efficiency can be sacrificed. Just shape down the excessive tones an you can be amazed! It's more or less a compromise between the two. That's why I like active systems and choose the drivers I expect that need most filtering in a very high efficiency. This enables much more possibilities to shape the sound to suit the needs.
I'm sure this system is about as perfect as can be, but still a tiny change of an overlooked detail can make all the difference. So I hope this is the case and we can find more improvement!
Hehe, far from perfect I suspect. It's showing a lot of promise though and I'd love to get these final things sorted out so your help would be greatly appreciated, yet again! I can manage the construction details but I'm getting well out of my depth when it comes to this sort of tweaking.
I'm quite happy to sacrifice some efficiency as I think there is some volume to spare at the moment, it really is very loud when I've tested it in my garden. I know it's mostly going to be used in a noisy environment but I'd love to get the sound properly balanced.
Re the Neutriks: No, I didn't even think about the RC network, I've just used the transformers on their own. I'll get straight on to that.
EDIT: Will the fact that I've got the secondaries paralleled for mono input to the amp affect the RC values?
Any thoughts on that 500Hz peak? It does seem very pronounced.
Firstly the tweeter polarity SHOULD be swapped because it is a 12db/oct filter... It's rather brilliant that you've heard that actually! Means you have pretty good absolute hearing ability. It's actually quite well possible to design a speaker system with ears as measuring instruments, as long as you know how it's supposed to sound...
Look at the schematics here and the dot on the 'magnet', it's clear...
http://i.imgur.com/IjKIt.jpg
Then the values (despite the resistor) of the tweeter filter could perhaps be chosen somewhat different. Like 18-22uF for the capacitor and 0.15-0.20mH for the coil.. Then when looking at the table above the woofer values might also be slightly off, but that could not make much difference...
Then the resistor could better be an L-pad configuration, this ensures minimal resistance and parasitic properties in the signal path...
http://home.broadpark.no/~onorstra/speakers/img/lpad.png
Here's a values sheet:
http://i.imgur.com/n9VhY.jpg
The cap for the woofer could best be a raw bipolar electrolytic cap quality and the tweeter cap a nice film quality...(don't know what's in there now?)
http://www.visaton.de/de/chassis_zubehoer/bauteile/elko_rauh/index.html
http://www.visaton.de/de/chassis_zubehoer/bauteile/mkt_kondensatoren/index.html
You can parallel them to get the value you want...
The transformers should each have the R/C I proposed and can be paralleled after...
Then the 500hz bump, it's probably baffle step, for which you can also make an excellent correction filter (simulation with the freeware program called Edge) But let's fix the transformer correction and filter phase and values first. It's also possible that it's a resonance, we'll fix it!
http://www.tolvan.com/edge/help.htm
Do you need the dampening of the series resistor? Otherwise you could dampen it another way that will affect the sound much less.
Add a 8 Ohms resistor and treat it as a tweeter, then run them all (the 3 actual tweeters and the 8 Ohms resistor) in series parallel.
It's questionable whether or not you should actually reverse polarity of the tweeters in your application. I would have made a 12 dB / 6 dB filter instead to get a 90 degrees phase shift at the x-o, and avoid the problem all together.
alexlangberg
30-Aug-10, 13:19
Wow, that's amazing! I thought this thread had died but just... wow. :)
You say you went through much trial and error on how to paint things. I'm going through the same thing (although with a case for a solar panel) and ended up spraying it with polyurethane/paint/polyurethane in that order. It seems to work ok although I'm not sure if the outer layer of polyurethane is a waste. May I ask what you use and how?
biggerbyfar
01-Sep-10, 02:06
Hi all. Just been reading through 15 pages of posts on my favourite topic! I didn't realise how long it had been since I looked at the 41hz site since purchaing a few kits around 2007 (Amp6, Amp6-Basic, Amp-9 BASIC).
MickB - very nice work and good skills. You've got far more patience than me! I'm also wondering if our paths have crossed already as I've got a few photos of some of the systems you mentioned (there's a flickr group just for this stuff):
http://farm3.static.flickr.com/2397/1535650093_f28d8ca588.jpg (http://www.flickr.com/photos/biggerbyfar/1535650093/)
Yours? (http://www.flickr.com/photos/biggerbyfar/1535650093/)
http://farm3.static.flickr.com/2192/1536553720_db1cb59a01.jpg (http://www.flickr.com/photos/biggerbyfar/1536553720/)
Bassfreight (http://www.flickr.com/photos/biggerbyfar/1536553720/)
http://farm3.static.flickr.com/2128/1536481596_c9fc90af30.jpg (http://www.flickr.com/photos/biggerbyfar/1536481596/)
Thumper (http://www.flickr.com/photos/biggerbyfar/1536481596/)
I occasionaly bring some tunes on the London Critical Mass cycle ride in the form of this shabby looking beast. http://farm3.static.flickr.com/2467/3914212368_f75cc28b43_m.jpg (http://www.flickr.com/photos/biggerbyfar/3914212368/)
White box bicycle sound system PA (http://www.flickr.com/photos/biggerbyfar/3914212368/)
I was intersted in the older posts about power vs batteries. I originally used an AMP6 but found it lacking in oompf so I switched to an AMP9. This was all at 12v with a 5Ah SLA. The Amp9 used to heat up much more and so received a heatsink upgrade. Then I knackered the two batteries I had by leaving them discharged (they weren't even deep cycle GEL types). So recently I aquired two slightly bigger 12v 7.5Ah deep discharge GEL SLA's, but I also decided to run them in series ~24v. Something that I didn't see mentioned, maybe I missed it, was the issue of discharge rates in relation to capacity. So a 5Ah 12v battery says 5Ah, but this figure is often quoted at a 10-hour rate. So it will do 0.5A for 10 hours, but it will fall well short of sticking out 5A for an hour.
In theory doubling your battery capacity not only doubles the capacity, it also puts you on a different relative discharge rate maybe 0.5C instead of 1C. Pulling power at 1C i.e. 5A from a 5Ah battery can result in halving the effective capacity of the battery compared to its stated 10 or 20 hour rate. So doubling battery capacity could potentially give you maybe three times original capacity in real-world usage. This is true for sealed lead acid, dunno if it holds for LiPo, LIon and variants?
See http://www.powerstream.com/battery-capacity-calculations.htm
I recently ran at 24v for the first time with an AMP9. It really improved things at high volume, really clear. However, the amp ran hotter and unfortunately seemed to blow after about 20mins on almost full chat. Now it just emits some white noise and heavily distorted audio output. It also now makes a strange squelch/chirp when powered up, I haven't looked into it yet properly.
Regarding dispersion, the 4x arrangement on the box really throws quite a concentrated beam, which is good news for the rider and is pretty much what I wanted. The 4 drivers are on their own channel as it's a 4x amp.
Lastly, the 500hz hump you have. Based on the the short-backed shape of the box, I would have though that it would promote resonances bouncing straight back through the cone off the relatively close rear panel. V-bro mentioned baffle-step, not sure if thats the same thing?
I recently ran at 24v for the first time with an AMP9. It really improved things at high volume, really clear. However, the amp ran hotter and unfortunately seemed to blow after about 20mins on almost full chat. Now it just emits some white noise and heavily distorted audio output. It also now makes a strange squelch/chirp when powered up, I haven't looked into it yet properly.
If the amp9basic runs even slightly hot, you're running it in AM (class-A/B) mode. You have to switch it to class-D mode. But since you've run it at 24V in AM mode you have completely destroyed the chip, so you'll need a replacement.
V-bro mentioned baffle-step, not sure if thats the same thing?
No, baffle step is something different than resonance. Baffle step is the lack of baffle surface supporting the wavelengths longer than the size of the baffle influencing the socalled 'diffraction' of the system. Many designers forget this (usually) creates a shelve lowering all bass output beneath that point.
Here's how it theoretically looks:
http://www.t-linespeakers.org/tech/bafflestep/bafflestep.gif
The 500hz peak could be caused by both baffle step and resonance. In a real life situation you get the natural response and the 'stepped' response added up and this can result in a sum that looks quite different! For instance when there is actually a large peak due to high Qtc of a closed box, the bafle step might show itself as a dent between the lower peak and the higher raise in output...(just an example)
http://www.gem.hi-ho.ne.jp/katsu-san/audio/Speaker/Baffle_step_filter.html
Some people fight it with a notch filter:
http://www.htguide.com/forum/showthread.php4?t=27583
The response 'graph' drawed by MickB is probably rather an exaggerated one because it's a decibel scale. It is often overlooked that what seems like a gigantic peak is actually just a few db you're looking at. This might look like a generally rather flat response, unless you scale it differently.
What really helped me to realise this and enhanced my feeling of making the right analysis was simulating and measuring situations. I have used LT Spice and Edge to draw up filters and get simulation graphs, then compared them to measured results. In Spice you can alter the scaling and often simulations looked like they weren't gonna have much effect, until I changed the scaling it showed it was often rather major!
biggerbyfar
01-Sep-10, 12:08
If the amp9basic runs even slightly hot, you're running it in AM (class-A/B) mode.
Oh balls, will have a look tonight to see how I set it up. Also saw a post elsewhere about leaving the unmute permanently connected results in a high offset on the output, which is exaclty what I have done. Will report back when I've pulled it out.
biggerbyfar
01-Sep-10, 23:27
If the amp9basic runs even slightly hot, you're running it in AM (class-A/B) mode. You have to switch it to class-D mode. But since you've run it at 24V in AM mode you have completely destroyed the chip, so you'll need a replacement.
Unfortunately, you are spot on. No jumper and so in AM mode, how stupid. It explains a few things. Still at least I've got some significant gains to look forward too!
unfortunately, you are spot on. No jumper and so in am mode, how stupid. It explains a few things. Still at least i've got some significant gains to look forward too!
Just don't toggle the jumper with voltage on the board!! It will kill the TAA4100A chip!!!!!!!!!!!
Hi biggerbyfar, the systems in those pics look very familiar! The backpack isn't mine, it's one of the original pair of ASBOxs built by StevieB and Sparky of the Londonskate to act as repeaters for Stevie's Bassfreight. They were the inspiration for my Asbombs as they made me want to build a lightweight version.
Your beast looks good. There are a bunch of us on skates and bikes joining in the London Sky Ride this Sunday (5th Sept) with as many linked sound systems as we can muster. We had a blast when we did it the year before last and it went down well with the cyclists. Will we see you there with yours?
The Bassbomb has been out for 2 outings now and has been a great success, Bognor Carnival on Sunday and the London street skate last night. I'm excited about starting on the mods that V-bro has suggested as I think it's going to sound very nice with a bit of tweaking, but unfortunately I've been negelecting the money earning work recently and I've got to spend a few days catching up. Updates as soon as I've got a bit of time to spare...
Still been too busy to start on the tweaks yet, which is frustrating as I'm itching to try them, but I hope to get some time next week. At least I've been getting to use it a lot, both on it's own and linked up with the other systems. Anyway, here's a quick review as this project has been very much a joint effort with the 41hz forum:
Now that I've used it a lot I can tell it definitely needs some adjustments to the sound but it has the potential to sound very nice. A lot of people have commented on the sound quality and loudness compared to the smaller units, and everybody loves the look of it. Playing it at home through my PC and making some coarse adjustments with the equaliser in VLC has been a revelation. To my uneducated ears it's hifi and I've been going through all my music and hearing all sorts of things I never knew were there, it has real depth and clarity to it. Listening to my mp3 player now sounds like I've got cotton wool stuffed in my ears. You bastards, you've infected me with hifi! I thought I was immune :)
Also the response of a portable system is always hard to judge because the environmental acoustic properties are different all the time, and usually not very ideal. It being a backpack unit already gives away that the bass will suffer the most of this. For outside response bass needs more power compared to an in room system. Especially when you take away that last bit of acoustical reinforcement, which is the floor. What you are actually looking for is a system that shows a good free air response with virtually no help of environmental acoustics...
I have made some very small portable speaker systems with astonishing bass for the size, but then mainly when placed on the floor. When lifted up higher than a foot they start giving in on the bass response. It's still good, but not great either...
I can see exactly what you mean now. While we were waiting for Bognor Carnival to start I placed it on the ground with it's back against the wall of a building as there was a smooth tarmac car park in front so we could have a little impromptu roller disco. The difference in sound was amazing! Considerably louder with much better bass response, it really filled the space. It shows that it can make a very nice static system if we choose the right spot. It also works well skating along the streets when there are tall houses close to road on each side, it sounds very loud then even on it's own.
The other revelation was how easy it is to skate with. I've removed the waist strap, and with the shoulder straps pulled fairly tight it sits firmly against your shoulders without wobbling around and moves together with your upper body. You notice the extra weight at first but after about 10 mins your brain adjusts to the extra inertia and you almost forget you're carrying it. It doesn't affect my balance and I can dance around like normal.
I built it for Londonskate as I didn't think I'd want to carry it myself, but now I've used it a lot I've fallen in love with it and I want to make another one for myself! A couple of other people have been inspired by it and want to build similar units too which means I need to ask some more advice:
The Celestion BN10-300S-4 (http://professional.celestion.com/bass/pdf/BN10-300S4.pdf) bass guitar speaker I used is no longer available. A possible alternative which I can buy is the FaitalPRO 10PR300 (http://www.faitalpro.com/img/products/schede/PS/10PR300/10PR300_datasheet.pdf). It's 2.1kg which is 0.5kg heavier than the Celestion but I can live with that. Those specs only show the 8 ohm version but it's also available in 4 ohm. I've tried entering the TS parameters into WinISD PRO but the results seem a bit odd. The cabinet it recommends is only 12 litres with 2 x 100mm dia x 500mm long (approx) vents. The frequency response curve drops away sharply below 200hz down to -4.5db at 100 hz, totally different to the Celestion. Increasing cabinet volume doesn't really help as the curve goes a very funny shape. I've checked the parameters I entered but can't see any mistakes.
Does anybody have an opinion on whether the Faital speaker would be suitable?
EDIT: Well, AJ Vented produces the same sort of response curve and cabinet specs so presumably that's what it is. Just seems surprising that there's not more bass response from a speaker of this size which is recommended for bass guitar.
EDIT (again): After playing around with the volume and vents a bit more I've come up with this which looks more respectable. That's a 16 litre box with 2 x 50mm dia x 19mm ports:
http://homepage.ntlworld.com/m.brown74/Bassbomb/Faital-16L-response.jpg (http://homepage.ntlworld.com/m.brown74/Bassbomb/)
EDIT (again, again): Sod it, I couldn't resist the temptation and I've ordered a couple, one for myself and one for another guy who's just caught the music backpack building bug. Done a bit of searching around and Faital seem to produce good quality gear, and I had a chat with the supplier who gave me a reasonable deal on a pair. In fact he generously agreed to refund if they turned out to be totally unsuitable which helped swing it. First thing will be to try it in the existing Bassbomb (23 litres) without changing anything and see how it compares to the Celestion, but the response curve above suggests it should work better in a smaller box which would be a bonus and help to make up for the extra weight of the driver. I'll report back.
biggerbyfar
17-Sep-10, 14:30
Re: outdoors. From being on your back I guess you're going from near freespace, to halfspace (on the floor) and then to quarter space (floor + wall).
p.s. Drivers - how about a Volt Radial 18". At a mere 17Kg it would be ideal for a backpack... Look kinda 'serious'.
http://www.voltloudspeakers.co.uk/Loudspeakers/Pro_PA_Range/body_pro_pa_range.html
biggerbyfar
26-Sep-10, 11:46
If the amp9basic runs even slightly hot, you're running it in AM (class-A/B) mode. You have to switch it to class-D mode. But since you've run it at 24V in AM mode you have completely destroyed the chip, so you'll need a replacement.
Unfortunately, you are spot on. No jumper and so in AM mode, how stupid. It explains a few things. Still at least I've got some significant gains to look forward too!
I take it back! It wasn't in AM mode. I built up a new amp9-basic kit which is now running, however, when I removed the old one completely (rather than looking into the box from above) and was able to see the underside of the pcb, I found that I had previously bridged pins 5-6 of J2 permanently closed and so always running in class-D mode.
biggerbyfar
26-Sep-10, 11:47
... this also means that it was running hot in class-D mode also.
Got some time at last so I can have another play...
Firstly the tweeter polarity SHOULD be swapped because it is a 12db/oct filter... It's rather brilliant that you've heard that actually! Means you have pretty good absolute hearing ability. It's actually quite well possible to design a speaker system with ears as measuring instruments, as long as you know how it's supposed to sound...
Look at the schematics here and the dot on the 'magnet', it's clear...
Doh! I never noticed that. As you say, though, it's encouraging that I actually heard it even though I didn't know what I was hearing. It gives me the confidence to trust my ears rather more than I did.
Then the values (despite the resistor) of the tweeter filter could perhaps be chosen somewhat different. Like 18-22uF for the capacitor and 0.15-0.20mH for the coil.. Then when looking at the table above the woofer values might also be slightly off, but that could not make much difference...
I did use that exact same table as you kindly posted it for me previously. As I'm aiming for 2.25k crossover I took the values for the woofer filter midway between those for 2k (12uF/0.47mH) and 2.5k (10uF/0.39mH) which gave me 11uF/0.43mH. Is that not a good way of doing it?
For the tweeter filter, if I later find that I have to attenuate the tweeters further using resistors will that affect the filter component values? Just wondering whether to order a range of different value capacitors all at the same time so that I don't have to put in another order later on. Don't want to buy lots of stuff I don't need though.
The cap for the woofer could best be a raw bipolar electrolytic cap quality and the tweeter cap a nice film quality...(don't know what's in there now?)
Vishay MKP (http://uk.farnell.com/jsp/search/productdetail.jsp?SKU=1413921)
http://www.visaton.de/de/chassis_zubehoer/bauteile/elko_rauh/index.html
http://www.visaton.de/de/chassis_zubehoer/bauteile/mkt_kondensatoren/index.html
The only bipolar electrolytic caps I can find at Farnell are the Panasonic SU series (http://uk.farnell.com/jsp/search/browse.jsp;jsessionid=RJM0GARAN5Z10CQLCIPZN4Q?N=0&Ntk=gensearch_001&Ntt=bi-polar+capacitor&Ntx=mode+matchallpartial&suggestions=false&ref=globalsearch&_requestid=346218) which are a lot cheaper than the ones you have linked to. Are they likely to be good enough quality? Doesn't say on the data sheet whether they are raw or plain.
Do electrolytics do a better job than MKP for the woofer filter? Or is it just that they're more compact?
Powered by vBulletin® Version 4.2.0 Copyright © 2013 vBulletin Solutions, Inc. All rights reserved.