diyAudio

diyAudio (http://www.diyaudio.com/forums/)
-   Chip Amps (http://www.diyaudio.com/forums/chip-amps/)
-   -   Recommend me an amplifier for this heatsink ;) (http://www.diyaudio.com/forums/chip-amps/130663-recommend-me-amplifier-heatsink.html)

TheSeekerr 1st October 2008 05:12 AM

Recommend me an amplifier for this heatsink ;)
 
2 Attachment(s)
Hi all,

I've got a pair of these rather small, but quite stylish, heatsinks sitting around gathering dust, and I was wondering if anyone could recommend me a chipamp project to use them with? (failing that, they'll probably go into some sort of overkill headphone amp)

The photo shows the heatsink atop a DVD case for scale - roughly 13cm long (I don't have a ruler to hand, though), with fins about 1cm long on one side (6 of those) and 1.5cm long on the other (4 of those).

Thanks

AndrewT 1st October 2008 11:44 AM

that would make a good sink for the drivers in a Krell KSA50 Klone.

gmphadte 1st October 2008 12:25 PM

They are good for 50 Watt class AB

After all, u will not party continuously.
Put a small fan, in case.

Gajanan Phadte

TheSeekerr 1st October 2008 12:48 PM

AndrewT: As much as that's an interesting amplifier, there's rather a lot of other costs to go along with it. Given that it's a small heatsink, I was rather hoping someone might be able to suggest a suitably scaled amplifier to go with it, complete with suitably scaled transformer requirements ;)

So it's your estimate, gmphadte, that it's good for a continuous dissipation of perhaps 50 watts, so I could run a class AB amplifier up to about, say, 30 watts, at continuous full power? (Not that I have a use for such a thing. My main amplifier at the moment is a class T capable of all of about 6 watts per side)

AndrewT 1st October 2008 01:29 PM

Hi,
I'd guess, very roughly, that this sink is about 2 to 3Cdegree/W.

If you try to dissipate 50W through a 3C/Wsink then the Ts will be about 150degC above ambient.
Let's assume your sink is 2C/W and you need to dissipate 30W continuously through it.
Ta=25degC,
Ts = 25 + [2 * 30] = 85degC.
Tc ~= 85 + [2 * 15] = 115degC.
Tj ~=115 + [3 * 15] = 160degC. Equals blown semiconductors if only two devices are fitted each dissipating 15W.
For reliability you could be looking at <=10W continuous dissipation.

Nico Ras 1st October 2008 10:26 PM

You need a thermometer, a power supply and a bolt on 10 or 25 watt resistor. Mount the resistor on the heat sink.

Measure the heat sink temp at ambient (in still air). Adjust the power supply so that the resistor dissipates 1 watt.

Wait 10 minutes or so for the temperature to stabilize and measure the temperature on the heat sink again. Subtract the temperature at ambient. The answer is deg C/w.

Now that you know the spec of the heat sink you can decide what you want to use it for.



Regards

Nico

danielwritesbac 3rd October 2008 01:07 PM

Put the Freeze on it.
 
Hi. There's a couple of tricks to expand the usefulness of a little heatsink.

First, avoid extra large caps at the amplifier board itself--limit to 470uF per rail maximum (Range: 150uF to 470uF). Put any larger capacitance onto a seperate power supply board (not point-blank at the chip amp).

Second, Artic Silver (for "TF" insulated chips) or non-conductive Artic Ceramique (for everything else) can increase the speed of heat transfer, and although the heatsink might still run out of resources, getting that whole heatsink up to the internal temp of a chipamp is far more useful than a "hot chip on a cool heatsink."

Third, snubbing the rectifier with little caps, one per each diode, sized by this chart: http://www.diyaudio.com/forums/showt...19#post1588619
If you have a single bridge rectifier, you'll need four caps from the center tap listing. If you have two bridge rectifiers, you'll need eight caps from the dual secondaries listing. See also "ring around the rectifier" and Mark Houston's project. Avoid botique(ish) diodes.

Fourth, Bleeder Resistors at the power supply. Both "snubbing the rectifier" and bleeder resistors will push down non-useful peaky voltage by giving the rectifier a slight load.

Fifth, as documented on the Carlos FM 2005 power supply, a 100v or higher tolerance capacitor in the range of 2uF to 4.7uF can be installed from V+ to V- at the amplifier board.

Sixth, Attach the heatsink to something else. That makes it a bigger heatsink. Reference: "Heat Spreader" technique where a heatsink is thermally coupled to a larger surface (usually amplifier enclosure, or metal mounting brackets).

Seventh, Cool air into the enclosure via the vent in the bottom of the enclosure, and then allow the now-heated air to escape via the vent in the top (or top of the rear panel) of the amplifier's enclosure.

Eighth, size your DC blocking "input filter capacitor" to the needs of your speakers, and also use the optional "RF block" (usually 220pF) capacitor to prevent amplification if high tones that are above the audio band. Think Audio Only and the amp will run cooler.

Ninth, Use either a dual-bobbin EI core transformer or a high quality Toroid. Size the voltage input for the expected output (use less voltage if using 4 ohm speakers).

And then my answer to your post can be this:
http://www.diyaudio.com/forums/showt...hreadid=123255

Summary: Give the heatsink all the help you can with fast compound, mounting to something larger, and sensible venting. Use a good linear (old fashioned) power supply with a classic snubbed rectifier, bleeder resistors and a 4.7uF "rail to rail" for quiet, level voltage. Think Audio Only because your amplifier doesn't need to be doing anything else. And, size the transformer voltage to the needs of your amplifier rather than (instead of) using a maxium.

Have fun, stay safe (not for use in the bathtub). ;)


All times are GMT. The time now is 02:11 AM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright 1999-2014 diyAudio


Content Relevant URLs by vBSEO 3.3.2