Different heatsource but similar requirements [Wcing udio chips]

it has been some time since I posted in Xtremesystems. Good to be back even though my problem today does not involve computers but rather audio amp devices.

here is the problem. Large class A amplifiers often output 2watts of heat per 1w of musical power. That is to say that a 200w amp can easuly out out 500ws of heat.

Systems are stereo and my currently problem is to deal with 500/600w of heat.

here are the modules...

http://i29.servimg.com/u/f29/15/79/86/51/5455df10.jpg

and here are the devices outputing the heat:

http://www.wholesaleec.com/upload/up...SFET-17832.jpg

I would like to watercool these devices in order to increase efficiency, reduce failure rate and reduce size of the enclosure. Ordinarily this amount of power would require enormous alu heatsinks.

This is a schematic of the "block" I envisioned:
Attachment 121088Attachment 121089Attachment 121090

Basically I need the most silent combo of pump and radiator/fans to deal with 500w of heat given the waterblock designed.

head pressure is less important than flow in my opinion. Silence is paramount.

Thanks for any help
Alex

P.s. I have built a system based on 2 audi tt radiators but obviously this translated into an impossibile to move system.

As even very powerful amps often have only basic heatspreaders, imho no need to go LC this time. Just install few quiet fans in amp's case, should be more then adequate (active air cooling already is several times more effective then passive heat dissipation).

You can send me 200$ saved on just buying 3*5$ fans instead of all the LC loop components :).

thanks for the suggestion. I was exploring this possibility since class A amps can overwhelm a passive radiator and if one can shave something off a 50kg beast (think 3kva toroids and 33kg of casing) it might be helpful.

I like your idea. I would say you could use something like two xspc rx360 rads, and some nice low speed fans.

EDIT: here's a list of fan's with their db/cfm http://www.quietpc.com/gb-en-gbp/products/120mmfans

Quote:

---

Originally Posted by churchy

As even very powerful amps often have only basic heatspreaders, imho no need to go LC this time.

---

Class A amps run very hot, they basically never have a idle state since the mosfets never get to the off state they are always conducting.

If you are going to make that block you have envisioned, I would make the fins run north-south if you want to keep it passive. If you are blowing air over the fins it won't matter. Also. I would try to get whatever is generating the heat as close to the heat spreader as possible.

I am srry for the fin orientation....that was just laziness. First sketchup model I found.

The closes thing possibile are the mosfets directly in contact with the block clamped by plexy, lexan, or other tough thick layer of material. Possibly non doncutvie to avoid short circuits.

What is the absolute quietest pump out there?

Laing D5 Vario or Swiftech MCP655 are very quiet, and have an RPM adjustment that can make the pump even more quiet

Go with a Jingway, low noise and lower power too. Rads are free flowing, so flow rate should not be an issue.

why u redundant?

when have u seen a waterblock with alu heat disappation fins which actually workd?

Your missing the point in water.

Water is not a heat bank, which ur making it as... its a method to move heat from source A to source B... the alu heat sinks on the side are pointless, because of your radiator up front handing all the heat exchange.

Yeah I'm scratching my head why there are passive heatsinks strapped to the water blocks.

Yeah my amps only do about 25w in class A... and they weigh something like 70lbs each, (parasound Jc1)
according to http://www.irf.com/product-info/data...a/irfp150n.pdf
They dissipate about 150w at 25c, your design calls for 6 of them.
Your best bet, is going to be a bunch of 120mm fans at or below 800rpm.
Maybe attach a pair of of thermalright machos on each side, provide venting, and place your trannsformer in a lower area, since it will block air flow.
I have always wondered why there aren't heatpipe designs, with active cooling in audio.
I guess a block of a luminum is cheaper than heatpipes and fins.
I honestly wouldn't be surprised if you could run them passively.

Also... i'm sure your set on your design... but I would imagine that other problems in your audio chain will mask anything crossover distortion may contribute.. so it may be a good idea to look at a high bias class a/b. But if your dead set on going class A... I don't think liquid cooling is the answer. Remember the devices are specced at 175c, and i think there are also some other ir mosfets in different cassing that have much higher heat rating, however your design probably doesn't allow for them.

Btw which design is this?