water cooled zen amp.

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Disabled Account
Joined 2007
Magura said:


I greatly respect your input, but in this case you need to do the math....it's a plain matter of radiating surface area.


Hi,
How much surface area do you think a radiator has? You have several tubes, surrounded by fins. The water contacts the inner walls of the tubes, heat radiates out.
It doesn't take a very big radiator to out do the biggest heatsink.
No math required
 
MJL21193 said:



Hi,
How much surface area do you think a radiator has? You have several tubes, surrounded by fins. The water contacts the inner walls of the tubes, heat radiates out.
It doesn't take a very big radiator to out do the biggest heatsink.
No math required

If you compare apples to apples, you will see that if a radiator is not force cooled by a fan, it hardlt works, and is actually outnumbered by a regular heatsink.

Magura :)
 
Magura said:
Now all you need is to tell us what the actual difference is between a radiator and a heatsink when it comes to dissipating heat?

As a heat exchanger, a radiator is more effective because the temperature is close to constant at every location.
Which is not the case for a heatsink, the reason why thermal efficiency drops for a bigger heatsink.

Most heat in a car engine is generated in the heads, easier to take away the heat with watercooling, unless you make really big heads.
 
jacco vermeulen said:


As a heat exchanger, a radiator is more effective because the temperature is close to constant at every location.
Which is not the case for a heatsink, the reason why thermal efficiency drops for a bigger heatsink.

Most heat in a car engine is generated in the heads, easier to take away the heat with watercooling, unless you make really big heads.


I have actually spent a little time on this stuff a while ago, and if the heatsink has a reasonably massive backbone, the temperature differences for heatsinks in a size of interest to us, is neglectible. We are talking less than 1 degree C for a 180*180*60 heatsink, measured anywhere on the sink.

Again compare apples to apples, your theory only holds true if forced cooling is used. I dare to assume you don't want a fan in the size of the radiator running in your living room.

As for the car engine cooling, yes, watercooling is the way to go, due to a number of reasons, one is size of heatsinks required on the cylinder heads.......but we are not building car engines ;)

Magura :)
 
Whether you are using radiators or heat sinks, the use of slow running fans is probably desirable.

The PC world is full of slow running 120mm DC fans that are 15-20dba.

Under-volting fans from 12v to 7v is commonplace in trying to achieve a "quiet PC".

Another tactic I've considered is placing the heat sinks/radiator inside a pipe of say 10-20" long. My guess is that the fan noise would be reduced significantly and lowered in frequency.

From the PC world you can get four 120mm radiator systems pretty easily http://www.dangerden.com/store/product.php?productid=261&cat=5&page=1

The thing I wonder about is how you'd mate the water cooling system to your transistors? For PC's they make copper water blocks for easy connection to CPU's, GPU's, even hard drives.

Could you use a CPU water block?
 
jacco vermeulen said:


Which most do not have.


Now that is a fact :)

But that is just a matter of making people understand that it's actually an important factor.

Back when I desigened the heatsinks for the round boxes, I first tried myself to make a design, but quite soon I found out that the way to go was to simply adapt an existing design, or it would take a lot of math and most likely suffer in one way or the other anyway.

My conclusion back then was that this design (see image) is among the most versatile.

Magura :)
 

Attachments

  • heat1.jpg
    heat1.jpg
    93 KB · Views: 685
When I was monkeying around with the alephX a couple of years ago. I had a heat sink about 6" high by 9" long by 1 1/2" deep that I wanted to use to test the board with after assembly. I was building an AlephX-100 so I knew the heat sink wasn't nearly enough. I found a surplus Cool plate on ebay. I took alot of thermal transfer goo and smeared it on the plate and bolted it to the heatsink. I then mounted the board to the cool plate. A good friend offered me a Zalman external water cooling system to use for 2 weeks (his son was being punished with no gaming for two weeks and this was one way to be sure his son didn't cheat). I filled the system with ethylene glycol and water. I was very surprised to find out that the combo of water cooler/heatsink worked well to keep the output transistors cool.

Here is what the cool plate I used looked like:

An externally hosted image should be here but it was not working when we last tested it.



I never used the aleph x as I sold the loaded boards here on the forum. I instead built an aleph5 using a conventional heatsinks. The solution for water cooling worked but too expensive for my tastes.
 
Groovy...

I'm a DIY kind of guy, so...

I'm going the water route because I like to experiment. I've gone the diy aluminum heatsink route, which is also a blast, but want something more streamlined.

I have a 3/4 ton Ford F250 radiator that I saved for a project like this. I'll mount it outside, in the shade, on the back porch, just outside my brand-new hobby room. Eventually, I'll couple a fan to the radiator. I'm tired of worrying about heatsinks, capacity, etc. Water works well, and will relieve me of the hefty labor involved in fabricating my own heatsinks.

You guys sure know how to work up a discussion...I like it!:D
 
carpenter said:
Groovy...

I'm a DIY kind of guy, so...

I'm going the water route because I like to experiment. I've gone the diy aluminum heatsink route, which is also a blast, but want something more streamlined.

I have a 3/4 ton Ford F250 radiator that I saved for a project like this. I'll mount it outside, in the shade, on the back porch, just outside my brand-new hobby room. Eventually, I'll couple a fan to the radiator. I'm tired of worrying about heatsinks, capacity, etc. Water works well, and will relieve me of the hefty labor involved in fabricating my own heatsinks.

You guys sure know how to work up a discussion...I like it!:D

If watercooling must make any sense, this is sure that way to go about it.
One thing you should keep in mind is that you need to be able to control the flow in order to get the MOSFETs up to working temperature without having to heat half of your backyard up to the desired working temperature. If made reasonably simple with a thermostatic valve from a domestic radiator, you can even just set the working temp with the thermostat, and just leave it at that, amp working or not, rain or shine.

Magura :)
 
Magura said:


you can even just set the working temp with the thermostat, and just leave it at that, amp working or not, rain or shine.

Magura :)


Excellent idea, Magura!:) Do you think a snap-switch would work?


Snap-Action Thermostat

The 36T series of 1/2'' bimetal thermostats from Therm-O-Disc® offers proven reliability in a compact, versatile, cost-effective design. The snap action of the bimetal disc provides high-speed contact separation resulting in excellent life cycle characteristics at electrical loads up to 15 amps at 120VAC and 10 amps at 250VAC. A wide variety of terminal and mounting configurations are available for maximum design flexibility. The quality, reliability, affordability, versatility and world-wide agency approvals of the 36T Series makes it the snap-action thermostat of choice for a wide variety of temperature control applications.

Switch Actions:

The 36T is available in three single-pole, single-throw (SPST) switch actions:

Automatic Reset – Can be built to either open or close its contacts on temperature rise.

Manual Reset – Available with contacts that open on temperature rise. The contacts can be reset by depressing the button after the control has cooled down.

Single Operation Fuse Disc – Available with contacts that open on temperature rise, and which never reclose unless the ambient temperature drops below +32°F (0°C) or below -31°F (-35°C).

Typical Applications:

The 36T is applied to a wide variety of applications as either a regulating control or a safety limit. Examples of applications include:
• Automatic drip coffeemakers
• Water heaters
• Sandwich toasters
• Furnaces
• Dishwashers
• Boilers
• Dryers
• Electric heaters
• Washing machines
• Office equipment
• Refrigerators
• Automotive car seat heaters
• Microwave ovens


36T Terminals:

Most 36Ts are supplied with 1/4” x .032” (6.3mm x .8mm) quick connect blade terminals that are available in unplated brass, tin-plated brass and nickel-plated steel. We also have 3/16” (4.8mm) blade terminals available in both .020” (0.5mm) and .032” (0.8mm) thicknesses. Most of these blade terminals are available in 0, 45 and 90 degree angles. The 36T can also be supplied with various weld tab and crimp/solder terminals. Other configurations/platings may be available if our standard offerings do not satisfy your application.
 

Attachments

  • p-36t-fig-2.jpg
    p-36t-fig-2.jpg
    2.5 KB · Views: 512
Member
Joined 2005
Paid Member
Try a better heat transfer fluid like antifreeze.

Remember the difference between Ethylene Glycol and Propylene Glycol as a coolant media - One will completely destroy your kidneys if ingested (Ethylene Glycol), and the other is a standard food additive.

Both are very good heat transfer fluids.

They also put lots of Propylene Glycol in KY Jelly. Keeps "stuff" from drying out....
 
john65b said:
Try a better heat transfer fluid like antifreeze...
I beleive we would not find an affordable or readily availble "better heat transfer fluid" than pure water. Every other additive degrades it's ability but controls boiling/freezing point etc.
What we want is to have a power stage submerged in Flourinert. A fluid that is as it's name describes, inert. The thermal transfer from the hot devices is carried away easily. A low boiling point flourinert can be made to regulate the device temp closely, with a wide range of power dissipation. Cray computors were once totally filled with the stuff.
As ussual, there is a downside. When I was ussing the stuff it was $350-$550 a gallon :xeye:
 
Pure water would be an ideal coolant.

But it needs anti-microbial agents like antifreeze.

So just use plain old green or orange car antifreeze.

And change it every 30,000 miles or so.

Seriously, PC water cooling guys use watered down car antifreeze. It doesn't gunk up and runs for a year or more before flushing.

I'd just use distilled water and car antifreeze mixed 3-4 parts water to 1 part antifreeze.

The antifreeze isn't for increased temperature range, but for the anti-fungus/bacteria agents they put in it.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.