Water cooled audio circuitry

[phase_accurate]

as Charles remarked, I have not found a way to reduce the energy consumption of Class A, dammit, but there you go..... onwards and upwards!

So you didn't succed with sliding-bias class A ?

Regards

Charles

 

[GRollins]

I don't think there will ever be a truly effective way to get the sound quality of class A without paying the power dissipation penalty. It's not for lack of people trying, though. Nelson and scads of others have done their damnedest to bend the laws of physics and sonics in clever ways, but there's always some trade off that pops up.
I'd love to be wrong on this one, but I'm not holding my breath, either. For the time being I just assume from the start that all of my amplifier circuits (at least from subwoofers up) will be class A--it's the only way I've found to get solid state to even approximate the sound of tubes.
Having a water cooling system on hand takes the heat dissipation worry out of the equation. I keep meaning to run piping up from my listening room into the living room above for winter use, but have yet to get it done.

Grey

 

[AKSA]

Charles,

Sheepishly I must admit I've not yet built the sliding Class A on which we collaborated together. I need to get around to it. The problem is that I've gone ahead with Class AB, and developed the sonics to fever pitch, unbelievable, in fact.

Further, the SB ClassA needs an additional voltage amp, and it really should be a tube..... that's more power supply complexity, a pity, and the cost of it has held me back.

But if you'd like to email me, perhaps I could restart the project, or at the least apprise you of what I have done in the interim!

Thanks for your post,

Hugh

 

[ostripper]

FANATICS indeed !!! water and electronics don't mix...
As computer repair is my forte, I've seen my share of
nuked motherboards by watercooled "fanboys".

Now, PC heatsinks and phase change heatpiping is another
story as I have had very good success in using PC
thermal control technology applied to audio amps.

A very good and economical way to cool 100 - 300 watt
BJT and MOSFET amps is by using 2 socket A (0.06 c/w) heatsinks

bolted horizontally to standard 1 1/2" aluminum L brackets.
When you put
a stereo pair with 4 socket A heatsinks face to face with a
80CM fan blowing through them (tunnel), you have one cool amp in a small package.(only $16.00) A single heatsink will do a
gainclone perfectly.

Just for the fun of it I screwed 4 IRF260's onto this

and ran the amp (Quasi)at 150W for an hour ,only to find a modest
15C temperature rise.

I still like the socket A heatsinks (outdated ,cheap,durable)
and will use them on all my DIY projects.
As far as fan noise goes ,one can have fun with all the PWM
thermal feedback fan circuits available...
OS

 

[GRollins]

Er...perhaps you missed the idea that I was trying to avoid a fan.
And I gather you weren't running class A; certainly not on the chip amp. Class A is a whole 'nother realm of heat dissipation. The power I've been speaking of is not speaker output wattage--it's pure heat. Roughly 1500W for starters. I think by the time you do the math, you'll find that it gets prohibitively expensive to buy multiple CPU cooling units for that amount of heat. Then you've got the logistical/electronic difficulties of stacking them in some manner.
Meanwhile, I got out for less than $100 and have had over seven years of cooling out of the same rig.
Saying that water and electronics don't mix is silly. They have a long, long history together. Besides the mainframe computers I mentioned at the beginning of the thread, there were water-cooled tubes, resistors, and so forth from even earlier. It's all a question of plumbing, and the basic elements of plumbing with copper pipe have been in place for quite some time.
I'm sure that if you take a moment to evaluate the IQ of an average plumber, you'll realize that they rarely moonlight as rocket scientists. Yet they manage--in spite of that--to put together plumbing that doesn't leak. Anyone with at least the IQ of a light bulb can learn to do their own plumbing. I can't tell you how much money I've saved over the years replacing my own water heaters, reservoirs, well pumps, etc. I rank it as a basic survival skill, right in there with the ability to change a tire or paint a wall.
Besides, joining copper to copper with solder is something that anyone who works with electronics should be able to do with one hand tied behind his/her back. Just pretend the pipes are big wires and you'll do fine.

Grey

 

[AKSA]

Grey,

That was grandly amusing, I laughed, appropriately enough, like a drain - if at the expense of the plumbing fraternity!

Hugh

 

[lineup]

Lineup,
The liquid sound of Class A cooled by water, and the hermetic imaging thus derived is a thing of beauty. Nothing like it....
Hugh

Now, PC heatsinks and phase change heatpiping is another story as I have had very good success in using PC thermal control technology applied to audio amps.
A very good and economical way to cool 100 - 300 watt BJT and MOSFET amps is by using 2 socket A (0.06 c/w) heatsinks

has anyone tried the motor bike engine heatsinks to cool thier amps?.
They should easily get rid of heat.

Motor Bike Engine motor coolers .

Have had the thought, yes
Now there are other 'heatsinks' we normally use for other things.
- Like electrical room heaters
- Like car engine coolers
Anything built out of copper / aluminium that is
built for radiate heat
or transport away heat
are possible candidates for cooling audio amplifiers.

And then we have the PC cooling gears.
CPU processors and Graphics processors working at several GHz produce tremendous heat.

The highest speed computers have special arrangements.
If I am not wrong, some use very cold substances, like liquid nitrogene gas,
for cooling, to several 100 degrees below the freezing point

It is just a matter of your idea of how to adjust these implementations for electronics cooling.
And to find a practical solution for this.


May I add, that 39 out 40 of all my Circuit Ideas for Audio Amplifiers
>>>> works Every Stage in 100% or more!!!! Class A
>>>> from the Input stage to the OUTPUT Stage
In this matter me and GRollins think the same.
In order to get best possible performance out of our amplifiers.

Lineup

 

[ostripper]

By grollins:

Er...perhaps you missed the idea that I was trying to avoid a fan.

I don't miss too many points, and I DO plumb houses.

As far as 1500W of dissipation, why not go all the way and
use a mini heat exchanger/heatpump (small personal
powered "beer fridge") to do the cooling.

With any watercooled system you need
a pump/compressor to circulate/condense the coolant which defeats the purpose
of a noiseless system ,so whether it's a fan or a circulating device
you will need some sort of active component in the system.
Some of the new Thermaltake water cooler pumps are
virtually noiseless as are the new heatpiped Zalman exchangers
(120mm fans/phase change).

For 1500W dissipation you would need a fan if you
wanted a standalone system that would not be awkward in size.
So for the sake of simplicity if I really had to have a class A
setup ,4 zalman 50$ coolers all bolted to a block would be the way
I'd go,and I'd port my entertainment center to direct the exhaust
to my feet (winter is a comin') to offset my heating bills and
set up a wind tubine in the yard for "green class A"

I guess it's just personal choice whether the DIY'er wants
awkward high maintainance (love the idea of home plumbing
heat dissipation) or compact equiptment.
(Same as the choice of A vs. AB/D classes.)
BTW , I can't hear the 120mm fan (700RPM) AT ALL in my HTPC..

 

[lineup]

there are different types of fans
and some types are very quiet .. at least at some speeds ..

now, you can actually silence a fan,
like you can make silent PC

Silent PC is so quiet,
that the sound will disappear in room background noise.
Or in the terrible noise of your bad music

 

[Variac]

Auto heater cores are smaller than radiators for smaller needs

HOw about 2 vertical 6' tall lengths of 2" copper pipe connnnected with elbows top and bottom? Put the outputs devices on the base of one and get a serious thermo syphon going.

Trouble is at the current cost of copper you would have to find sime taken out of a building..

Of course for noise and heat, just put a fan cooled amp in the room next to your listening room and poke holes through the Gypsum board for the wires.. No one does this because they want t o look at the amp while listening to t he tunes..

For safety, just mount an AC thermal switch on the output bar.
These switches are cheap and available in various set temps.

Wouldn't these work?

http://dkc3.digikey.com/PDF/T083/P1977.pdf

Put one on each channel and you have redundant protection

 

[lineup]

tlaking about cooling techniques

Any body tried Peltier Effect cooling.
I have a vague remembering it has been mentioned before
here in www.diyaudio.com
... and what have not been mentioned in our topics archives

From wikipedia.
Jean Charles Athanase Peltier (1785 - 1845) was a French physicist.
http://en.wikipedia.org/wiki/Peltier_effect

Peltier effect

When a current I is made to flow through the circuit, heat is evolved at the upper junction (at T2), and absorbed at the lower junction (at T1). The Peltier heat absorbed by the lower junction per unit time, \dot{Q} is equal to

\dot{Q} = \Pi_\mathrm{AB} I = \left( \Pi_\mathrm{B} - \Pi_\mathrm{A} \right) I

Where Π is the Peltier coefficient ΠAB of the entire thermocouple, and ΠA and ΠB are the coefficients of each material. P-type silicon typically has a positive Peltier coefficient (though not above ~550 K), and n-type silicon is typically negative, as the names suggest.

The Peltier coefficients represent how much heat current is carried per unit charge through a given material. Since charge current must be continuous across a junction, the associated heat flow will develop a discontinuity if ΠA and ΠB are different. This causes a non-zero divergence at the junction and so heat must accumulate or deplete there, depending on the sign of the current. Another way to understand how this effect could cool a junction is to note that when electrons flow from a region of high density to a region of low density, they expand (as with an ideal gas) and cool.

The conductors are attempting to return to the electron equilibrium that existed before the current was applied by absorbing energy at one connector and releasing it at the other. The individual couples can be connected in series to enhance the effect.

An interesting consequence of this effect is that the direction of heat transfer is controlled by the polarity of the current; reversing the polarity will change the direction of transfer and thus the sign of the heat absorbed/evolved.

A Peltier cooler/heater or thermoelectric heat pump is a solid-state active heat pump
which transfers heat from one side of the device to the other.
Peltier cooling is also called thermo-electric cooling (TEC).

 

[gaetan8888]

tlaking about cooling techniques

Any body tried Peltier Effect cooling.
I have a vague remembering it has been mentioned before
here in www.diyaudio.com
... and what have not been mentioned in our topics archives

Hello

I did use them for super-cooling of ccd camera for astrophoto (to reduce noise at very long expositions).

You can see Peltier plate in portable 12 volt refrigerator for car and small RV.

A Peltier Effect plate need 12 volt at arround 10 amp.

It can also be use for class A amp, but you still need a BIG heatsink on the hot side of the Peltier plate.

There is also the vortex effect cooling devices, but they need compress air, so it can be very noisy.

Bye

Gaetan

 

[Variac]

Peltier MOVES the heat but doesn't dissipate it. So.... you need as big a heatsink as the amp would require (actually bigger because the peltier device isn't very efficient) on one side of the device, which gets warm.. Make sense from a physics point of view..


Plus you have to power the device which consumes power too..

 

[ostripper]

I used peltier cooling modules but the problem is the same,
you still have to remove the heat from the chassis.

Also with a peltier unit ,IT needs it's own power supply..
(even more heat to remove). The only real application
of a peltier is to insanely overstress/overclock at cold
temperatures (lower than ambient),or for compact cooling apps.
(refrigerators).

One could make a hybrid scheme with a closed loop liquid
or phase change system with an integrated exchanger on the
amp and use flexible ductwork (like dryer exhaust tubing) to
extract from the amp's exchanger remotely (reversed fan
outside or in another room). This would allow for "squirrel
cage" fans or other high CFM devices and eliminate long
runs of coolant..

 

[GRollins]

By grollins:


I don't miss too many points, and I DO plumb houses.

As far as 1500W of dissipation, why not go all the way and
use a mini heat exchanger/heatpump (small personal
powered "beer fridge") to do the cooling.

With any watercooled system you need
a pump/compressor to circulate/condense the coolant which defeats the purpose
of a noiseless system ,so whether it's a fan or a circulating device
you will need some sort of active component in the system.
Some of the new Thermaltake water cooler pumps are
virtually noiseless as are the new heatpiped Zalman exchangers
(120mm fans/phase change).

For 1500W dissipation you would need a fan if you
wanted a standalone system that would not be awkward in size.
So for the sake of simplicity if I really had to have a class A
setup ,4 zalman 50$ coolers all bolted to a block would be the way
I'd go,and I'd port my entertainment center to direct the exhaust
to my feet (winter is a comin') to offset my heating bills and
set up a wind tubine in the yard for "green class A"

I guess it's just personal choice whether the DIY'er wants
awkward high maintainance (love the idea of home plumbing
heat dissipation) or compact equiptment.
(Same as the choice of A vs. AB/D classes.)
BTW , I can't hear the 120mm fan (700RPM) AT ALL in my HTPC..

This is almost physically painful. You appear to have completely, absolutely, totally missed the entire point of the thread. The polite response to your post is to invite you to to read the thread, starting with the first post. All of it. Every word. Pay attention.
Here's the sound bite version: There is no Freon. None. Not a drop. There is no compressor. This is not a refrigeration unit.
The water is the coolant. It flows from the pump, through a braided line, through the copper pipes (where the output devices are mounted), back into braided line, into the heat exchanger (where the heat goes from the copper lines to the aluminum fins, thence into the air), to the reservoir, and from there back to the water pump. That's it. It's that simple.
With the exception of the braided hose, all of this is shown in the pictures for those who have trouble with verbal descriptions. I trust that the--admittedly short--lengths of hose visible in the pictures will suffice for everyone to imagine the rest. If anyone really has to see the hoses, I'll get my wife to take some pictures of them...but they'll be really boring pictures. Don't say I didn't warn you.

Auto heater cores are smaller than radiators for smaller needs

HOw about 2 vertical 6' tall lengths of 2" copper pipe connnnected with elbows top and bottom? Put the outputs devices on the base of one and get a serious thermo syphon going.

Trouble is at the current cost of copper you would have to find sime taken out of a building..

Of course for noise and heat, just put a fan cooled amp in the room next to your listening room and poke holes through the Gypsum board for the wires.. No one does this because they want t o look at the amp while listening to t he tunes..

For safety, just mount an AC thermal switch on the output bar.
These switches are cheap and available in various set temps.

Wouldn't these work?

http://dkc3.digikey.com/PDF/T083/P1977.pdf

Put one on each channel and you have redundant protection

Auto heater cores, motorcycle radiators, oil coolers, etc. would all work, but you're likely to have to go to forced air cooling to move enough air to coll things properly. A car or truck radiator would probably do okay without having to use a fan.
I don't think 12' of copper pipe alone will be sufficient in terms of surface area. Purely convective water flow is attractive--I'll be trying that first if I ever get around to putting a heat exchanger in the living room above my listening room. If I don't get enough flow, I'll use a pump.
Yes, copper is expensive at the moment. There's absolutely no reason not to use 'used' pipe. If my read on the market is correct, we'll see a sharp spike in commodities in the immediate future, followed by a fall in prices.
Most any thermostatic switch will work. The pump doesn't draw a lot of power.

Any body tried Peltier Effect cooling.

Peltier modules have been suggested many times in many threads--perhaps even in this one--I don't remember. In any event, the problem with Peltier modules is that they don't get rid of the heat. The just move it a few millimeters away. Then you've got to find a way to dissipate not only the heat created by the output devices, but the additional heat generated by the modules. So you're back to big passive heat sinks or trying to water cool the Peltier. Things are worse, not better.

Grey

 

[Variac]

Most any thermostatic switch will work. The pump doesn't draw a lot of power.

While that is a good idea- have a thermo switch on the pump, That wasn't what I was saying. I meant to have the switches turn off the AC supply to the amp channels if the temp goes high. This is because the most common objection to cooling systems more complicated than output devices on air cooled fins is that the whole thing will melt down if something goes wrong.

I know people don't want more switches in their circuits, but certainly having it control the AC isn't objectionalble..

 

[ostripper]

Grollins, maybe some of us don't want 5-12 gallons of water
swirling around under natural convection,just trying to
pass some more practical ideas around.

When I worked at GE plastics I had to remove 12 kilowatts of
heat (totally silent instrument shack ,fans external)
from industrial control semiconductors so I've seen all the
options, (except for yours).

Water cooling was never a option (safety concerns) with the
control systems involved ,as we were using non-conductive
coolant ,phase change or forced air thermal solutions
way back in the 20th century.

I would just get a real cold feeling if our fellow DIY,ers had
a horrible accident so I just wanted to "modernize" the
discussion.Even the PC mod group has moved to non-
conductive coolant..

Radio tower transmitters used to use only purified de-ionized
water (for safety)but now even they to have moved on to
liquid gases or non-conductive coolants.

 

[lineup]

another issue with circulating water system
may be:

what if there is some stop in water supply, circulation
then circuit would be overheated

I admit, that most times when I have no water pressure
in my flat (one apartment among many)
is when the electrical pump in our house have stopped
by mains electrical drop out
... and so the amplifier would be off on these occations

but suppose we get a leakage
somewhere in those pipes

 

[Variac]

I just mentioned a temp cutoff switch..

 

[AKSA]

Hi OS,

I don't think anyone doubts your obvious experience, from this or other threads. You are a worthy contributor, and for one I appreciate it.

Couple of things to note about water cooling electronics. First, we are not dealing here with vacuum tube voltages, or industrial control currents. There are fuses on the domestic switchboard, and in the amp at mains level and as well on the rails. Grey's installation is clearly soldered elbow joints on copper, so would not leak, or if it ever did, give a pin prick leak only. Witness the hydraulics on modern dump trucks; it almost never bursts or leaks and they function at thousands of kilopascals. Sealing is easy. Pressures and flow for water cooling domestic amps are very low, probably no more than 2 psi and a couple of gallons a minute. There is no reason why this is more unsafe than say connecting up a transformer to the mains, and no DIYer thinks twice about the dangers when he cranks up another project.

I think also that until you have done it on a high dissipation Class A amp, and seen just how effective it is, you would shy away from water cooling. Moreover, the PC cooling sinks are now very cheap, thanks to mass production, and due to design iteration they are now very effective. If you DIY, particularly with 120 x 120 fans at half speed, decent airflow rates can easily be achieved at very low noise, less than 20 dbA is easily possible. I have done experiments with these fans on an MF18-200 heatsink from Conrad, that's 5" high, 8" long with fins around 2" long, and it easily dissipated 175W with 27C rise above ambient over a few hours. This is practical for a large Class A amp, but I suspect you'd get even better performance from the liquid cooled Zalman PC sinks, which are very impressive.

Different folks, different strokes. At the time Grey and I built our systems, there were no suitable PC sinks available. Now the situation has changed. I do agree that the tubes, pumps, reservoirs and safety switches are a PITA, however. I even set up mine with a level sensor and automatic solenoid to admit more water should the level drop. It worked like a charm, but removing it from the house was like brain surgery.....

Sincerely,

Hugh