View Full Version : Anti-Thump procedure
I'm trying to design a very simple anti-thump circuit for AMP15. I was first wondering whether this circuit is actually necessay or not. e.g. what happens if I provide supply to the amplifier (in unmuted status) with the speakers connected to the output ? Do You hear a massive BUMP ?
If the answer to the above question is positive, what is the correct power-on procedure ?
1. Speakers disconnected, ampli muted, provide supply to ampli, unmute ampli after a few secs, then connect the speakers
2. Speakers disconnected, ampli muted, provide supply,connect the speakers,unmute ampli after a few secs.
3. Speakers connected at all times, ampli muted, provide supply, unmute ampli after a few seconds
and what about power off ?
Any suggestion ? [?]
My amp15 is not finished yet, but from my experience with several other 41hz amps the turn on thump is not loud at all.
So you provide supply to the ampli unmuted and with the speakers connected?
Hi I have randomly tried several turn on/off procedures and never got a frightening turn on thump. Actually I don't recall getting one at all....
------X-max. can be several meters on any driver, too bad it can be done only once...------
With my Truepath, the amplifier initially starts in muted state and unmutes after a few seconds, but that might be due to the delay in the rail supply coming up. On mine, the transition from muted to unmuted and vice versa is mostly silent.
However if the power is cut without moving the amp to muted state first I have gotten a nasty sound, which is more like a screech rather than a pop. Basically, you don't want 5V to drop before the rails if the amp is not muted.
I guess 5V on the Amp15 would be supplied from the bulk caps so you may not have this issue with it, as 5V will be maintained for several seconds when the supply is disconnected.
To get around this on my Truepath I wired the second pole on my power on-off switch to the mute jumper so that the amp goes into mute when I turn off the power. I also have a speaker protection relay circuit, which disconnects the speakers on turn-on and when the amp goes into mute.
I highly recommend using speaker protection, but this might be useful:
http://www.41hz.com/Forums/topic.asp?TOPIC_ID=1361
My Amp15's turn on thump is more of a slight click, not bad at all and no relays.
what value of input capacitor are you using ??
2.2uf Stock kit polyprop. I have some 2.2uf and 3.3uf Panasonic FM electrolytics, I may try them too.
I used a speaker protection module with a speaker on/off delay, no thumps at all. I didn't see any change in the sound quality, and also it has an add of speaker protection.
audiophool
07-Feb-09, 10:57
quote:Originally posted by col_s
With my Truepath, the amplifier initially starts in muted state and unmutes after a few seconds, but that might be due to the delay in the rail supply coming up. On mine, the transition from muted to unmuted and vice versa is mostly silent.
However if the power is cut without moving the amp to muted state first I have gotten a nasty sound, which is more like a screech rather than a pop. Basically, you don't want 5V to drop before the rails if the amp is not muted.
I guess 5V on the Amp15 would be supplied from the bulk caps so you may not have this issue with it, as 5V will be maintained for several seconds when the supply is disconnected.
To get around this on my Truepath I wired the second pole on my power on-off switch to the mute jumper so that the amp goes into mute when I turn off the power. I also have a speaker protection relay circuit, which disconnects the speakers on turn-on and when the amp goes into mute.
I highly recommend using speaker protection, but this might be useful:
http://www.41hz.com/Forums/topic.asp?TOPIC_ID=1361
Here is an explanation of the results of my experimentation with Truepaths and several speaker types which led to the current arrangement of the tbrick circuit. It may help others who are similarly experimenting.
Pop avoidance for class D amplifiers is not a trivial exercise. The latest class D amps have inbuilt timers other circuitry to slowly ramp the signal and avoid pops.
Tripath didn't publish any information about startup or shutdown of their chips. TA3020 datasheet goes so far as to assure us that if we have pops then it's our own fault.
Startup pop is not solely caused by caps charging because pops happen when the power supplies are stable before unmuting.
When you power up a normal amp, the mute led always comes on for a second. When the led goes off and the amp unmutes, and despite what the TA3030 datasheet says, there is a click/pop/thump on the speakers. The actual sound depends on which speakers are used and some speakers don't make minimal or no sound.
When you switch off the amp, V5 drops and this causes the amp to mute and the mute led comes on. I'm not sure why the mute comes on but perhaps it the reduction in V5 makes the VPP/VNN (which dosen't drop quickly due to the large bulk caps) look higher to the TA3020 and this triggers an overvoltage mute. When V5 drops further the mute led goes off again. The speaker click happens when the mute led comes on. Again, both the sound and significance of the click varies from speaker to speaker.
I don't need or want any facility for muting the amp. I just switch the power on or off, therefore my Truepaths all work with the mute jumper always in.
At first I wasn't concerned about the click but after a while it became annoying.
Manually switching the mute at power on and power off was not an option because it's too much hassle and messy.
Here's what I did.
The first thing I expected to stop the startup click was the speaker protection relay driven from Trueapth's mute signal. I was surprised to find the relay it didn't reliably stop either the startup or shutdown click.
I considered Col's suggestion of extending the up-time V5 by adding a bigger bulk capacitor. This would require a pretty big V5 cap or some additional circuitry to discharge (or disconnect) the VPP/VNN, and would only stop the shutdown click.
Aleka's circuit (referenced by Col above) uses a power supply monitoring circuit typically used in PCs (basically a comparator) to force the mute when V5 is below 4.6V. Seems like a good idea and has good reports. I went so far as to buy some for testing but never ended up using them (does anyone want them?). I'm not confident that this will work as a generic solution across the entire power supply range with all speaker types.
Col's arrangement of using the power switch to switch the mute is similar to Aleka's except it is surely a generic solution to shut-down pops because it mutes the amp before V5 drops at all. This arrangement alone however doesn't address the start up pops. I have another concern about this arrangement and that's safety; having 240V and V5 on the switch puts them close together. A loose connection at sometime in the future could be disastrous.
Tbrick circuit monitors the V5 supply before the V5 rectifier and when the power switched off, the mute before V5 collapes.
Could it be that after umute there is some sort of inrush into the Tbrick output circuitry that causes the startup pop? Startup pops are avoided if there is a delay after umute before the speaker protection relay closes. I did this on the latest tbrick circuit.
Here's the tbrick circuit.
http://lh4.ggpht.com/_zjwqFDsgXoQ/SW9Biv_21SI/AAAAAAAAD50/1sklHvy0Alc/s800/TBrickTrV5r1_2%20Schematic.jpg
The pop suppression is in the top right corner. If either HMUTE is high or the Power-Off Detect diode has no signal then the Unmute Delay Cap is kept discharged and the speakers are kept disconnected.
After V5 is up and HMUTE falls then a the Ummute Delay Cap charges via the 100k resistor. This provides a 1 second delay after HMUTE drops before the speaker relay closes.
I hope this helps.
Rgds,
Thankyou audiophool
this is what I was searching for.. also the explanation confirms what I thought.. would you mind attaching an high res pic of the schematic (zipped maybe)? this way is very difficult to read some of the components values..The power off detect uses the positive pulses of the transformer to keep C11 discharged.. so if no more positive half waves are coming from the transformer, voltage on C11 raises and the mosfet shorts Q14 base to GND and the relay goes off...
I was thinking to a similar circuit.. anway maybe there's a better option still very functional but easier to build.. I'm working on it, if it works well I'll post it in the next days.
quote:From ta3020.pdf
[u]Mute</u>
When a logic high signal is supplied to MUTE, both amplifier channels are muted (both high- and low-side
transistors are turned off). When a logic level low is supplied to MUTE, both amplifiers are fully operational.
There is a delay of approximately 200 milliseconds between the de-assertion of MUTE and the un-muting of the TA3020.
[u]Turn-on & Turn-off Noise</u>
If turn-on or turn-off noise is present in a TA3020 amplifier, the cause is frequently due to other circuitry
external to the TA3020. While the TA3020 has circuitry to suppress turn-on and turn-off transients, the
combination of the power supply and other audio circuitry with the TA3020 in a particular application may
exhibit audible transients. One solution that will completely eliminate turn-on and turn-off pops and clicks is to
use a relay to connect/disconnect the amplifier from the speakers with the appropriate timing at power on/off.
The relay can also be used to protect the speakers from a component failure (e.g. shorted output MOSFET),
which is a protection mechanism that some amplifiers have. Circuitry external to the TA3020 would need to be
implemented to detect these failures.
The mute logic prevents a simple RC-link in series with a double function power switch that will mute the amp when you power off the amp.
There is not much one can do about this, lets just hope it works out of the box. I really need this will be flawless since I will be bi-amping, and don't want to risk the tweeters when turning the amp on or off.
> There is not much one can do about this
There is definitely something you can do.. you can use a tiny circuit that connects your speakers to the AMP after some seconds after you tunr it on and then disconnects the speakers immediately when power is removed. In my AMP I have an SMPS as power supply, so I cannot disconnect on the behalf of missing sinusoid pulses coming from the transformer, but I should have found a very good alternative to detect immediately when power is removed and deenergize the speakers relais. It's still in beta testing, it's working well but I still have to finish the AMP to see how much it's effective (I would like a totally noise-free turn on and turn off)
quote:Originally posted by hawui1
> There is not much one can do about this
There is definitely something you can do.. you can use a tiny circuit that connects your speakers to the AMP after some seconds after you tunr it on and then disconnects the speakers immediately when power is removed. In my AMP I have an SMPS as power supply, so I cannot disconnect on the behalf of missing sinusoid pulses coming from the transformer, but I should have found a very good alternative to detect immediately when power is removed and deenergize the speakers relais. It's still in beta testing, it's working well but I still have to finish the AMP to see how much it's effective (I would like a totally noise-free turn on and turn off)
You are right, what I meant was that there is nothing one can do simply by doing stuff to the mute function. I wasn't even thinking about speaker relays.
Ah Ok, infact with themute function I guess there's nothing more we can do.. but the relays should fix the transients on both power-up and power-down if correctly driven by the appropriate circuit..
thomaseliot
04-Mar-09, 00:03
For who want to try a simple and cheap solution there is the Velleman K4700 board. It delays output speaker relays on startup and disconnects them on power off. Besides it has another interesting feature: it detects if there is more than +/-1V DC on output, in which case it disconnects speakers. This is a real protection, because it can happen a short between main rails and output: on chip overvoltage/current protection will detect normal voltages so the amp will unmute normally and main rails voltage will pass to speaker frying them. K4700 would disconnect speakers detecting more than 1V DC on outputs.
I use a K4700. It works well and has saved my speakers.
If you don't like the relays in the signal path, Amplimo makes a special relay (called LRZ) specifically designed for audio use with gold contacts and high current capability. A problem with normal relays is that they can corrode over time and introduce distortion.
If anyone in the Netherlands would like to do a bulk buy of these relays from Amplimo and sell them on, then I would be interested in at least four of them.
Just bought one.. this will save me from having to create the printed circuit board..
Thanks for the advice col_s !
regarding the relays, where did you found them? I searched Amplimo website but I couldn't find any :(
They used to sell them from directly from their website, but they recently changed their website and I don't see them there anymore, but i'm sure they still sell them as they use them in some of their amplifier modules. I think you probably need to email them.
I found an online shop that sells them, but it's in dutch:
http://shop.idd-electronics.com/product_info.php?products_id=11456
It's a nightmare to understand what is in the above site.. but I was able to register and I ordered a couple of these relays.. (at least I hope I did this and I didn't do anything wrong !!) let's see what happens :)
quote:Originally posted by hawui1
It's a nightmare to understand what is in the above site.. but I was able to register and I ordered a couple of these relays.. (at least I hope I did this and I didn't do anything wrong !!) let's see what happens :)
Need translation of anything in particular?[8D]
------X-max. can be several meters on any driver, too bad it can be done only once...------
They've been very kind, I stated in the notes of the order I'm just English/Italian speaking and they were so kind to send me a message in English to confirm the order was processed correclty and that I didn't do any mistake.. so now I'm just waiting for the relais to arrive.. (plus the K4700 kit)..
Anyway thankyou for offering support ;)
That's great but I really should have mentioned that I have not used one of these relays with a K4700 [:I]
Looking at the K4700 schematic it uses two relays with 12V coils. The Amplimo LRZ relays are 24V, so they are probably going to need something like at least 20V to turn on (maybe less, you would need to check, some 24V relays can work at 12V).
The manual including the schematic for the K4700 is here:
http://www.velleman.be/downloads/0/illustrated/illustrated_assembly_manual_k4700_rev1.pdf
This page talks about a few mods that can be made, including the suggestion of adding an external transformer:
http://www.anidian.com/audio/deconstruct/k4700.shtml
The transformer supplied with the K4700 is 2 x 6VAC, with produces a split supply for the opamp and then the transistors that switch the relay operate between the +ve and -ve rail, ie. about 12-16VDC. Maybe increasing the transformer to 2 x 9VAC would provide enough voltage to power the relay, as that would give about 2 x 12VDC (ie. 24V to the relay). If you did this, the values of several resistors would need to be recalculated and altered.
You also need to look at what the resistance of the relay coil is to ensure that it won't need too much current. I think it should be OK as the resistance of the coil in the velleman relay is 360ohm, and resistance of the LRZ is 900ohm. Anyway, in the article I linked above it says "Replacing Relays - The K4700 will safely supply quite a bit of current, especially if the AC power supply is replaced. T4 is rated for 500mA, quite enough to power larger, higher rated relays.".
The best specs I could find on the LRZ relay are:
"This unique AMPLIMO relay has been designed for switching loudspeakers. It contains a tungsten leading contact which can switch 100A (at 50V) loudspeaker current. After closing, this contact is bridged by a gold plated silver contact, giving negligible contact resistance even at small currents.
When switching off the loudspeaker the silver contact opens before the tungsten contact. The silver contact is protected by the tungsten contact, it never switches high currents, so it will not burn. The gold-plating prevents oxidizing of the silver contact, so even years later the contact resistance is very low.
Coil voltage 24VDC nominal, at 70C ( 20.4V to 33.5V ).
Coil resistance 900 ohm
Ambient temperature 20 to 70C
Convenient for power amplifiers of 5 to 200W"
No problems.. I saw that from the beginning..
but I have a solution at hand.. don't worry ;)
Anyway thanks for the additional info.. this idea of the two contacts is really good..
thomaseliot
10-Mar-09, 00:21
quote:Originally posted by hawui1
No problems.. I saw that from the beginning..
but I have a solution at hand.. don't worry ;)
Anyway thanks for the additional info.. this idea of the two contacts is really good..
Can you share your (tested) solution?
1. http://www.geofex.com/Article_Folders/ampprot/dcprot.htm
2. http://www.tkk.fi/Misc/Electronics/circuits/ampdelay.html
3. http://www.elecfree.com/electronic/loudspeaker-protection-with-soft-start/
4. http://users.otenet.gr/~athsam/protection_1.htm
More on 4th option -> This is tested by me with following mods.[:p]
The last one which I build ~2.5 years back and still working fine. In place of relay driver Q1-Q4, I used 2N5294 transistor. Rather driving relay from Q1, I used Q1 to pull down AMP1 mute pin. Q2 drives output relay for speaker cutoff. Q4 drives buzzer in case of DC(while muted AMP). Q3 not used so far.
One wire from AMP's fault logic pin goes through series of diode 1N4148 and resistor of 10K each to junction of R8-9, R14-15. SO this responds to AMP's fault too. Diode 1N4148 is there to protect reverse voltage from module to AMP's fault logic pin.
To power this module, I wound 50+50 turns of 28SWG copper wire to get center tap 12-0-12, 500mA on 200VA toroidal trafo itself.
Here is the picture. Upper half between trafo and relay you can see everything.
http://www.41hz.com/gallery/omishra/dsc00201.jpg
> Can you share your (tested) solution?
the solution is VERY SIMPLE, I have a separate 24V from my SMPS, the relay on the board of K4700 will activate the "special" ones (Amplimo RLZ) using this 24V as bare switches.
The delay switching off in case of protection triggering is negligible given the relay on the protection board are switching very very low currents.. and at any rate for me the idea was to obtain a delay connecting the speakers on switch on and an immediate disconnection of the speakers at amplifier switch off exactly what I'm going to obtain.. addittionally these LRZ seem to be very good quality not to affect signal quality..
I'm still waiting two things : an aluminium plate and the K4700.. when they arrive I'll start building the AMP.
I finished to build K4700 yesterday evening.. Fortunately it's working perfectly.. (the kit is very good quality.. ) I tested also the DC protection and is working fine..
Only one final doubt: Anybody knows if switching on/off the AMP15 (or AMP7) with no load at the output can cause damages ?
some amplifiers doesn't like at all to remain with no load at the output at any time.. even if the speaker gets disconnected when they're idle the Ampli can get damaged.. so just wanted to know if anybody have odd experiences on this side..
Thanks in advance ;)
Powered by vBulletin® Version 4.2.0 Copyright © 2013 vBulletin Solutions, Inc. All rights reserved.