View Full Version : Nice output filter resistors...
I've found these resistors to replace the two 10 ohm resistors that are tied together, they fit VERY nicely on the AMP15:
http://nl.farnell.com/productimages/farnell/standard/42724135.jpg
http://nl.farnell.com/caddock/mp915-20-0-1/resistor-20-ohm/dp/1612033
They're pretty expensive though...
------X-max. can be several meters on any driver, too bad it can be done only once...------
How many of these are needed for one AMP15 ? only two right ?
at any rate these devices can dissipate only 1.25W without heatsink.. is it enough?
Hi,
yes only two... without heatsink just 1.25W indeed, but you can mount them on the bottom to go on the same heatsink as the rest :)
------X-max. can be several meters on any driver, too bad it can be done only once...------
Are these the Zebel network resistors? Sorry, I don't have an amp15, but I do like to keep up with tweaks as they appear to be applicable to other models as well.
Yep they are part of the Zobel network... I just wanted to try a metal film resistor for once. In some amps they are carbon and in some they are wirewound (so pretty inductive)... And the construction is nice and small which leaves more space for different output filter capacitors...
------X-max. can be several meters on any driver, too bad it can be done only once...------
quote:Originally posted by V-bro
Yep they are part of the Zobel network... I just wanted to try a metal film resistor for once. In some amps they are carbon and in some they are wirewound (so pretty inductive)... And the construction is nice and small which leaves more space for different output filter capacitors...
------X-max. can be several meters on any driver, too bad it can be done only once...------
Oh ok, very interesting I've played around with a few resistors in the output filter, and I find slight differences compared to having no zobel (speakers attached of course). I plan to use some Mills non-inductive wirewounds in the AMP7, as they seem to be the most popular crossover resistors.
Oh yes they use bifilar winding techniques, but that doesn't mean they are completely non inductive... Did you measure this?
I can tell you the caddock metal film resistors are 99,99999% non inductive by nature....
------X-max. can be several meters on any driver, too bad it can be done only once...------
quote:Originally posted by V-bro
Oh yes they use bifilar winding techniques, but that doesn't mean they are completely non inductive... Did you measure this?
I can tell you the caddock metal film resistors are 99,99999% non inductive by nature....
------X-max. can be several meters on any driver, too bad it can be done only once...------
I have not measured the inductance. I will when I get some resistors on hand since I don't want to take them out of some other things I have.
On an interesting note about the Caddock 900 series, it was indictaed that they were magnetic, which put me off from them. I only read good things about them, but I haven't had a chance to try them.
> it was indictaed that they were magnetic
magnetic? Strange.. documentation is speaking about allumina compund so it shouldn't be magnetic..
they're also recommended for high frequencies..
V-Bro your idea of putting them on the bottom side of the pcb and use the main heatsink is very neat and very cool.. good point !
Only one question, the ones you indicated (915-20) are in a TO126 container, in terms of distance between the holes, diameter of the terminals and ruggedness to screw the part to the heatsink wouldn't be better to use MP820 instead (TO220)?
What do you think is the best solution?
Very interesting, as to my ears the output filter parts having quite some impact on the sound.
I think these here:http://nl.farnell.com/caddock/mp930-20-0-1/resistor-20-ohm/dp/1612023 are TO220 and capable of 2,25 Watts without heatsink, so these could be mounted on top as usual.
But I think even more important are the 22K input resistors,as they are directly in the signal path. I took some Vishay cmf55 and also a very good 0,1 µF film/foil cap instead of the small yellow MKT.
There is also the 0,1µf bias cap (C22 on the board), which is quite critical from my experience with the small Tripath amps.
I often read about the advantages of small parts because of less EMI, did anyone actually made a comparison? How does EMI sound?
In my experience the good caps, which are significantly bigger, always sound WAY better than the small SMT or MKT.
So maybe the advantages of good parts are more significant than the disadvantage of the big size.
Cheers,
Juergen
Here are some pics of a prototype board:
The top (as you can see I've used different caps as well)
http://www.snapdrive.net/files/361770/AMP15/SNC10175k.JPG
The goo you see here and there is silicone glue I've added to make sure the washers stay in place...
http://www.snapdrive.net/files/361770/AMP15/SNC10172k.JPG
http://www.snapdrive.net/files/361770/AMP15/SNC10168k.JPG
http://www.snapdrive.net/files/361770/AMP15/SNC10166k.JPG
http://www.snapdrive.net/files/361770/AMP15/SNC10165k.JPG
As you can see there's no problem at all to make them go on the same surface as the rest, just add a blob of paste and you're done...no need to bolt them down any further... now they're a massive 15Watt capable!!![}:)] I've placed the same type of nylon washer under it to make sure the pressure of mounting the board doesn't bend the pins and to make sure they have the same profile height as the rest....
I've also used a ruler to make sure all the devices will be nicely flat on the heatsinking surface...
------X-max. can be several meters on any driver, too bad it can be done only once...------
Just to recap so, how many of these resistors would be needed for each AMP15 and for each AMP7 .. two each?
what is exactly the distance between the holes in the pcb ?
The distance is 5mm both on the AMP7 and AMP15, you need only two resistors per amplifier.
------X-max. can be several meters on any driver, too bad it can be done only once...------
Now the big question, does this mod make it sound better :)
I think all mods that get into the project make the builder feel better :)
I,m not sure I will be able to hear it as I don't have the best ears...
My girlfriend might...
Theoretically it should make some difference as there's no inductance parallel to the resistor in the zobel network any more. So it should have some consequences to the load behaviour...
------X-max. can be several meters on any driver, too bad it can be done only once...------
Yes, I was thinking the same thing. I'm not so fond of the doubled up resistor arrangement myself, this is much more elegant. My plan is to wait until v-bro has done all the hard work optimising the amp15 and then copy him :)
At any rate I guess the best benefit is to avoid the two carbon resistor dissipating the heat.. from another 3d I heard they get pretty hot.. so they will degrade with time.. placing these two resistors connected to the heatsink will avoid this to happen..
quote:Originally posted by wadoka
Yes, I was thinking the same thing. I'm not so fond of the doubled up resistor arrangement myself, this is much more elegant. My plan is to wait until v-bro has done all the hard work optimising the amp15 and then copy him :)
Hahaha! Sounds like a plan! :D
Guys!!?! I'm sweating my balls off here!![xx(][xx(]
:D
No seriously, I'm glad it's appreciated! Glad I'm not doing it just for my own sake...
I completely forgot to post this picture:
http://www.snapdrive.net/files/361770/AMP15/SNC10166k.JPG
------X-max. can be several meters on any driver, too bad it can be done only once...------
Another good mod I find is using SMD ceramic caps instead of all the glossy little yellow ceramic caps..
In fact there is and SMD cap inside the glossy yellow part, when you heat up that part the paint comes off like an orange peel and you can solder the wires off and use it as SMD capacitor :)
That way you decrease lead length even more and especially the ones next to the fets on AMP15 will fit better.
The pitch distance on the PCB footprint in any case is perfect for any 0805 size SMD cap and they come cheap these days so when you build a lot of amps (like me :)) you can stock a few pieces of the most used values (like 0,1uF)
Here's an example of an AMP9-B where I first did the mod:
http://www.snapdrive.net/files/361770/AMP9%20and%209B/SNC11318.JPG
------X-max. can be several meters on any driver, too bad it can be done only once...------
SMD caps are a great idea, I'll definitely have to copy that.
back on topic... what do you think of these thick film planar resistors for this purpose?
http://au.farnell.com/tyco-electronics/mpc52200j/resistor-5w-20r/dp/1174267
there are 5w and 10w non-inductive versions
quote:Originally posted by col_s
back on topic... what do you think of these thick film planar resistors for this purpose?
http://au.farnell.com/tyco-electronics/mpc52200j/resistor-5w-20r/dp/1174267
there are 5w and 10w non-inductive versions
Yes, they were the others that popped up in my search at Farnell.com when I chose for the Caddock resistors. They sure look interesting, but also a bit vulnerable... Maybe I'm wrong...
------X-max. can be several meters on any driver, too bad it can be done only once...------
Powered by vBulletin® Version 4.2.0 Copyright © 2013 vBulletin Solutions, Inc. All rights reserved.