heat sink for UCD700? - diyAudio
Go Back   Home > Forums > Amplifiers > Class D
Home Forums Rules Articles diyAudio Store Gallery Wiki Blogs Register Donations FAQ Calendar Search Today's Posts Mark Forums Read

Class D Switching Power Amplifiers and Power D/A conversion

Please consider donating to help us continue to serve you.

Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
Thread Tools Search this Thread
Old 11th April 2006, 10:03 PM   #1
pburke is offline pburke  Germany
diyAudio Member
Join Date: Nov 2003
Location: Madison Wisconsin
Default heat sink for UCD700?

I'm looking for some additional heat sinking to use on the back of my amp chassis. I'm going to bolt the module against the back face plate of my chassis, which is all aluminum, but the material back there is pretty darn thin. I was thinking of just sandwidching a 2mm machined plate from Frontpanel express on the back of the cassis, which will also give me nicely machined mounting holes for everything I need back there. However, would that be enough for a UCD700AD module? Just one per chassis, about 100mm tall, 250mm wide in back. The entire rest of the chassis is aluminum, so the heat has somewhere to travel to. I'm just converned about the local area where the module is mounted possibly getting too warm.

As an alternate, I had something like a Pentium2 processor heat sink in mind, adding that to the back of the chassis with arctic Silver thermal epoxy, but if I don't need it, I would rather not do that.

So would the chassis on its own be adequate?

  Reply With Quote
Old 11th April 2006, 10:38 PM   #2
diyAudio Member
Join Date: Mar 2002
Location: Soquel, CA
Yes, the chassis is plenty good by itself, however for best sound you should damp the module. By using a dissimilar material layer in conjuction with the chassis you will reduce the ringing and audible effects of the chassis. In my commercial UCD amps I use one quarter inch thick brass inside the chassis and then the alluminum chassis outside layer. You could experiment with brass screws as well but these are very hard to find in metric sizes here in the US.

Regular heatsinks suck big time. Resonating machines. Take any amp that has external heatsinks and damp the heatsinks with damping material along the edges of the fins so they no longer ring when you run your fingernail along them.....WAY better sound.
  Reply With Quote
Old 12th April 2006, 12:53 AM   #3
Account Disabled
Join Date: Feb 2004
Send a message via AIM to classd4sure Send a message via MSN to classd4sure

hahahaaa You're too much! really..
  Reply With Quote
Old 12th April 2006, 05:42 AM   #4
MOER is offline MOER  United States
diyAudio Member
Join Date: Aug 2003
Default Ringing heatsinks

I would dearly love to know how a heatsink can affect the sound of an amplifier. If the amplifier circuits themselves do not cause any mechanical vibration, how does the heatsink affect the circuits?

It is kind of like that French manufacturer, FM Acoustics who put their power amplifier drive circuit in a plastic box and filled it with some fancy beach sand form a Pacific South Sea Island and told the public that this "Improved" the sound. What a lot of junk. I had the opportunity to repair one of these amplifiers and it sounded awful even with their beach sand.
  Reply With Quote
Old 12th April 2006, 06:49 AM   #5
diyAudio Member
Join Date: Mar 2002
Location: Soquel, CA
Trouble is. most engineering types use their brain instead of their ears. You cannot think good sound, you have to listen for it.

The analog signal goes through the components, the microvibrations of the componets gets transfered to the signal...much the same way that vibrating cabinets effect the sound of a driver....of course, it is much more subtle a vibration but just as audible. You must have heard about different kinds of feet for components...different types of racks, etc. These are all ways to deal with microvibrations that effect the sound. Many, many people put damping material on the chassis, circuit boards, etc. to good effect. The sound you hear when you take your fingernail to a component and flick it is the exact sound that will be transfered to the sound coming from your speakers. If you hear a tink when you do this to a undamped transistor mounted in space, then when you damp it the tink sound will go away. This really is science...at some level...whether or not it is measureable it is certainly audible. Peter Moncrief of IAR used to take a recording that contained a lot of percussion and especially a bell ting and would send the signal through a preamp and look at the output on a super fast digital scope....he would then compare the bell ting trace to another preamp....what he found is that the more revealing preamp showed more subtle variations and dynamic swings on the wave form than the less audibly revealing component. Now this guy makes wonder caps, wonder solder, etc...so he knows about how a single component has a sound....so I asked him if he could measure the difference using a bell ting when he tests a single piece of wire, resistor, cap or whatever....he said....no, he could not. The only way he could measure it is to run the signal through a preamp or amp that has many parts in the signal path and power supplies, etc. All brands of caps/resistors/solders/wires etc. sound different and we can hear it, but at this time we have no measurement that can tell us which one will sound better or purer, execpt straight wire bypass tests.

You can argue with your brain...but you listen with your ears....have an open mind...try it and see if you can hear a difference...if you cannot then, so be it. Because I do not come from an engineering background and because I have an open mind, I try everything that I read about....this is why I come to these forums because someone here will no doubt tell me something I don't know. Always searching for more. And occasionally I give out some information that I hope will be of value (sonically).

There are many manufacturers that damp circuit boards, parts, heatsinks, etc. This is not new news. I did not invent it! When someone says they hear a difference doing something, I put that info into my biocomputer and at some point try it and see if I hear the same thing...if I do and I like what it does then I put that in my tweak ammo box. It is very pragmatic to me....my goal is to make a component disappear ..to help make the connection between the heart of the musician, the music and the listener.
  Reply With Quote
Old 12th April 2006, 07:37 AM   #6
SSassen is offline SSassen  Netherlands
diyAudio Member
SSassen's Avatar
Join Date: Mar 2005
Location: Edam, The Netherlands

You made me spill my morning coffee, now I have to go home and change shirts Microphonic phenomena do exist, but what you're proposing is utter nonsense, but a good read nonetheless (chuckles).

Best regards,

Sander Sassen
Mediocre is not my middle name.
  Reply With Quote
Old 12th April 2006, 08:15 AM   #7
Eva is offline Eva  Spain
diyAudio Member
Eva's Avatar
Join Date: Oct 2003
Location: Near the sea
Send a message via MSN to Eva
Trouble is the terrible advice I'm reading in that thread.

Every fast-switching high-frequency circuit requires the heatsink to have a very short return path to circuit (switching) ground, and some means to reduce the capacitance between the metal tabs of the power transistors and the heatsink is also very desirable. Otherwise, the heatsink will follow the switching voltage waveform with all its high dV/dt switching transients and will radiate EMI as a big antenna. Attaching directly the heatsink or the power transistors of a class-D module to a big metal case is a very very bad idea.

Furthermore, as the insulators mounted between the transistors and the heatsink are sandwiched between two conductive surfaces whose relative voltage is changing very quickly and by very big amounts, the insulators will exhibit piezoelectric effects and will expand and contract, thus transmitting mechanical energy to the heatsink and causing audible noise (now some naive audiophile may pretend to fix the whole problem by stuffing the fins of the heatsink, without being aware of the complex phenomena behind).

Actually there is not an easy fix for that. Grounding the heatsinks (to some internal switching ground or supply rail in the circuit following a very short path, never to signal ground or using a long wire) gets rid of most of the EMI issues (while attaching the heatsinks to a rack case produces a nice FM transmitter), but the piezoelectric effects remain unsolved (if you ever care about that). Using plastic-cased power transistors is quite effective as it gets rid of most of the capacitance to the heatsink, the real root of the problem. Using thicker insulators or several standard ones stacked may be useful too. Also, there are commercially available aluminium oxide ceramic insulators (they may be 3 milimeters thick while keeping low thermal resistance) that are routinely employed to attach switching transistors to big metal cases without capacitive issues.

Check that link: http://www.warth.co.uk/site2003/blue...ucts_fast.html

p.s. Does anybody know where aluminium oxide insulators are available in small quantities?
  Reply With Quote
Old 12th April 2006, 09:18 AM   #8
Pierre is offline Pierre  France
Join Date: Nov 2004
Location: Paris
Thanks, Eva, your comments are always wise.

My self-made Class-D amplifiers have a 100nF capacitor from one of the chassis screws to the power supply GND input faston, (the screw is at a couple of centimeters from where the output mosfets are) in order to short that capacitively coupled HF noise to GND.
Besides that, the heatsink is bolted to the case, and the case is earthed via the IEC mains input.
Is this correct or a very bad idea in terms of EMI emissions?

(BTW: no piezoelectric or microphonic effects noticed )
  Reply With Quote
Old 12th April 2006, 09:19 AM   #9
Account Disabled
Join Date: Feb 2004
Send a message via AIM to classd4sure Send a message via MSN to classd4sure

"p.s. Does anybody know where aluminium oxide insulators are available in small quantities?"

I'm fairly certain you'll find two on the UCD400, the 180 uses plastic transistor cases.

Also, the T sink is well decoupled for EMI. These are very well engineered modules.

It would seem there's good reason to stick to engineering after all.

I don't doubt the existance of the piezoelectric effect, but it's already mechanically damped by bolting the mosfet to the heatsink. I do have serious doubts even in worse case this would translate into audible artifacts through the speakers, measurable or otherwise. I can't really see this happening to any worrysome degree with a product such as thermasil pads either?

You just have to draw the line somewhere I guess. Is it still OK to listen to an amp that's placed in light intensive environments... or can I find some audiophile window tint?

Also, keep in mind the filter core itself is magneto-restrictive, I could believe that translating itself into audible artifacts loooong before flicking my finger over a heatsink, but I wouldn't even go so far as to say that "that's" why an aircore sounds better.

Excellent post btw.
  Reply With Quote
Old 12th April 2006, 10:39 AM   #10
Pierre is offline Pierre  France
Join Date: Nov 2004
Location: Paris
ClassD4sure is right in that the piezoelectric effect of the core is muuuch bigger than in the heatsink, so I think it makes no sense worrying about that. EMI is another thing, for sure.
  Reply With Quote


Hide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off

Similar Threads
Thread Thread Starter Forum Replies Last Post
Heat sink schematics Llafriel Pass Labs 10 18th June 2004 07:05 AM
How much heat sink is necessary? rosss Pass Labs 11 23rd February 2003 05:23 AM
New heat sink group buy Brian Donaldson Pass Labs 5 25th January 2003 10:13 AM

New To Site? Need Help?

All times are GMT. The time now is 02:48 PM.

vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2017 DragonByte Technologies Ltd.
Copyright 1999-2017 diyAudio

Content Relevant URLs by vBSEO 3.3.2