Kilowatt - you could use a nylon spring stretched out along the straight part of your block. This will cause vortices and accomplish the goal of turbulence. Also ruffing up the inside surfaces works well [scratching, gashing, etc.]
IMHO ,
the main limiting factor when using a radiator is the thermal conductivity air-radiator, and not water-radiator-
It is not possible by simple designs and reasonable air velocity to get higher than alphaair= 50W/m2C, as compared to the thermal conductivity on the waterside which starts at about alphawater= 200-250 W/m2C with very low flowrates and reaching up to 5000 W/m2C at very high flowrates.
At a water velocity (in the radiator) of 0,1 m/s the alphawater is rougly 600 W/m2C...
the thermal resistance of the "radiatorwings" , alpharadiator , itself is also very small compared to the others
total thermal resistance k can be calculated from :
1/k = 1/alphaair+ 1/alphawater + 1/ alpharadiator
which leads to the conclusion that the main thermal resistance is on the air side. A 10% improvment gives almost 10 % more
cooling capacity.
But increased air velocity leads to more noise 🙁
How would DI water promote corrosion more than ordinary water which is more impure?
It is not unwise to use DI water if you have water with high calcium content , which would deposit on the hottest parts of the cooling circuit, lowering cooling capacity.
But generally it is not a problem in a closed circuit where the water isnt replaced regulary.
/rickard
the main limiting factor when using a radiator is the thermal conductivity air-radiator, and not water-radiator-
It is not possible by simple designs and reasonable air velocity to get higher than alphaair= 50W/m2C, as compared to the thermal conductivity on the waterside which starts at about alphawater= 200-250 W/m2C with very low flowrates and reaching up to 5000 W/m2C at very high flowrates.
At a water velocity (in the radiator) of 0,1 m/s the alphawater is rougly 600 W/m2C...
the thermal resistance of the "radiatorwings" , alpharadiator , itself is also very small compared to the others
total thermal resistance k can be calculated from :
1/k = 1/alphaair+ 1/alphawater + 1/ alpharadiator
which leads to the conclusion that the main thermal resistance is on the air side. A 10% improvment gives almost 10 % more
cooling capacity.
But increased air velocity leads to more noise 🙁
How would DI water promote corrosion more than ordinary water which is more impure?
It is not unwise to use DI water if you have water with high calcium content , which would deposit on the hottest parts of the cooling circuit, lowering cooling capacity.
But generally it is not a problem in a closed circuit where the water isnt replaced regulary.
/rickard
DI is more likely to become ionized by the impurities of the system.... DI with a small amount of antifreeze for corrosion and a wetting agent to reduce surface tension would be ideal.
I also agree that the limiting factor of the cooling system is the air. It is not a very good conductor of heat so the surface area of the air side has to be much larger. Through careful design you can have a very high rate of air flow with minimal noise. Just get a high quality fan [only the air moving should make noise, not the motor.] Some fan designs also minimize vorteces and cavitation. Cavitation is where the majority of the noise comes from [put your finger very close to a fan blade to hear cavitation, try not to lose the tip of your finger either 🙂] Squirrel cage fans are the best fans to use. Low noise, very high flow rates, minimal vorteces. The air comming from one of these fans is like a solid block and doesn't spread out much. Normal fans have air flow that spreads out and goes everywhere in a spiral pattern where squirrel cage fans pushes forward [mostly]
Also, why have you [Kilowatt and others] decided to use water cooling?
I also agree that the limiting factor of the cooling system is the air. It is not a very good conductor of heat so the surface area of the air side has to be much larger. Through careful design you can have a very high rate of air flow with minimal noise. Just get a high quality fan [only the air moving should make noise, not the motor.] Some fan designs also minimize vorteces and cavitation. Cavitation is where the majority of the noise comes from [put your finger very close to a fan blade to hear cavitation, try not to lose the tip of your finger either 🙂] Squirrel cage fans are the best fans to use. Low noise, very high flow rates, minimal vorteces. The air comming from one of these fans is like a solid block and doesn't spread out much. Normal fans have air flow that spreads out and goes everywhere in a spiral pattern where squirrel cage fans pushes forward [mostly]
Also, why have you [Kilowatt and others] decided to use water cooling?
Water Cool audio circuity
Grey,
I am impressed with you cooling ideas.
But do you have a fail safe system in case the pump stops for any reason.
I say this as Hugh Dean of ASKA amps has also made a water cooling system for his Glass Harmony amplifiers.
I think it was a total loss system not recirculating and it worked quite well until one day when he was out shopping one of his darling daughters turned off the water and the whole thing boiled dry and died! (this actually happened twice).
regards
macka
Grey,
I am impressed with you cooling ideas.
But do you have a fail safe system in case the pump stops for any reason.
I say this as Hugh Dean of ASKA amps has also made a water cooling system for his Glass Harmony amplifiers.
I think it was a total loss system not recirculating and it worked quite well until one day when he was out shopping one of his darling daughters turned off the water and the whole thing boiled dry and died! (this actually happened twice).
regards
macka
Phage, It's a little late to rough up the inside of my blocks, but a spring in there should create turbulence well. Thanks. And the reason I'm using liquid cooling is to avoid buying big heatsinks. The only cost of my liquid cooling system so far has been the tubing, connectors, and the belt and pulley for the motor.
I got the pump running, but I had to pull start it because the motor doesn't have enough starting torque. I'll have to get a motor with MUCH higher starting torque.
I don't have a heat source (like a heatsink-mount power resistor) to test the block-air thermal resistance right now.
The motor by itself is quiet, but the sound of rushing air and vibration measures over 80dB@1m. Obviously, that won't do for hifi except at really high volume levels, which is fortunately just what this amp will do. Maybe I can get a 2 speed motor so it will be quieter at medium volume levels.
Also, I can't seem to get all the air out of the system, which isn't good, but at least it's turbulent in there, the clear tubes turn white with bubbles all around when it's running.
I've definately have some kinks to work out, but it should work quite well when it's done.
P.S. What % antifreeze should be in there to prevent corrosion?
I got the pump running, but I had to pull start it because the motor doesn't have enough starting torque. I'll have to get a motor with MUCH higher starting torque.
I don't have a heat source (like a heatsink-mount power resistor) to test the block-air thermal resistance right now.
The motor by itself is quiet, but the sound of rushing air and vibration measures over 80dB@1m. Obviously, that won't do for hifi except at really high volume levels, which is fortunately just what this amp will do. Maybe I can get a 2 speed motor so it will be quieter at medium volume levels.
Also, I can't seem to get all the air out of the system, which isn't good, but at least it's turbulent in there, the clear tubes turn white with bubbles all around when it's running.
I've definately have some kinks to work out, but it should work quite well when it's done.
P.S. What % antifreeze should be in there to prevent corrosion?
I finally got around to testing my cooler. Not very accurately, because I don't have any big power resistors with known thermal resistance right now, but here's what I did:
I got a heating element, about 1kW, and bent to fit on top of one of my blocks. I then put the element in a mold and filled the mold with lead, leaving me with a lead block with a heating element embedded in it that fits on one of the blocks. I then greased it up with heatsink grease and clamped it tightly to the block. After well over an hour of running, the top of the lead block did not rise above about 90 degrees C. The cooling block, tubes, and radiator only went up a few degrees. My conclusion is that since lead is far less than perfect as a heat conductor, much heat stayed in the lead block, rather than being conducted to the aluminum, and therefore screwing up the accuracy of the test a lot. The temperature of the rest of the system stayed quite low, so I assume that the rest of the heat, probably well over 500W was dissapated very well by the radiator. It appears to be a remarkable cooling system, and made mostly from scrap.
I got a heating element, about 1kW, and bent to fit on top of one of my blocks. I then put the element in a mold and filled the mold with lead, leaving me with a lead block with a heating element embedded in it that fits on one of the blocks. I then greased it up with heatsink grease and clamped it tightly to the block. After well over an hour of running, the top of the lead block did not rise above about 90 degrees C. The cooling block, tubes, and radiator only went up a few degrees. My conclusion is that since lead is far less than perfect as a heat conductor, much heat stayed in the lead block, rather than being conducted to the aluminum, and therefore screwing up the accuracy of the test a lot. The temperature of the rest of the system stayed quite low, so I assume that the rest of the heat, probably well over 500W was dissapated very well by the radiator. It appears to be a remarkable cooling system, and made mostly from scrap.
kilowatt - everyone reccomends a different mixture ratio. 2 oz of watter wetter, keep cool, or jet dry to the gallon and about 6 oz all weather antifreeze to the gallon. If you don't want to buy a wetting agent like water wetter, keep cool, or jet dry then just add dish detergent [not too much!] it will accomplish the same goal. As long as there is no water trapped inside the block you should be ok. The water wetter should also help these bubbles go away.
Dissipating 500W is a nice feat kilowatt. You could have drilled a hole in the lead to put the heat sensor closer. What kind/size radiator are you using?
PS to those using evaporative cooling systems never put the above chemicals in them. The water wetter will make soap suds come out of the cooler for a very long time and you really don't want to breathe antifreeze
Dissipating 500W is a nice feat kilowatt. You could have drilled a hole in the lead to put the heat sensor closer. What kind/size radiator are you using?
PS to those using evaporative cooling systems never put the above chemicals in them. The water wetter will make soap suds come out of the cooler for a very long time and you really don't want to breathe antifreeze
I guess I didn't mention it, but I did drill a hole about 1/4" into the lead to take the measurement.
The radiator is from a little Toyota with a 4 cylinder engine. I don't know the model. The radiator is about 20" square and maybe 1 1/2" thick. I got the pump and fan from the same car. I had the pump running at about 3450rpm for my testing. The temperature of the aluminum blocks only rose a tiny amount, so I would think this system is quite capable of cooling things with much more heat dissapation. The amp That this thing will be for is a 7200W class AB amp, I figure maybe 4000W heat dissapation under a worst case scenerio. The system might be able to handle that already, but if not, I can further improve the heat dissapation by adding a turbulator in the blocks, bypassing the viscous clutch on the fan to get more airflow (louder though), and getting all the bubbles out of the system.
The radiator is from a little Toyota with a 4 cylinder engine. I don't know the model. The radiator is about 20" square and maybe 1 1/2" thick. I got the pump and fan from the same car. I had the pump running at about 3450rpm for my testing. The temperature of the aluminum blocks only rose a tiny amount, so I would think this system is quite capable of cooling things with much more heat dissapation. The amp That this thing will be for is a 7200W class AB amp, I figure maybe 4000W heat dissapation under a worst case scenerio. The system might be able to handle that already, but if not, I can further improve the heat dissapation by adding a turbulator in the blocks, bypassing the viscous clutch on the fan to get more airflow (louder though), and getting all the bubbles out of the system.
Hello All,
Thought this might be relevant (pertinant).
In the Feb 2002 edition of Electronics World mag is a photo of 6 guys standing infront of a massive new supercomputer.
The caption reads :
Maybe this machine can be used to help find soloutions to our greenhouse warming problems.
Regards, Eric.
Thought this might be relevant (pertinant).
In the Feb 2002 edition of Electronics World mag is a photo of 6 guys standing infront of a massive new supercomputer.
The caption reads :
Maybe this machine can be used to help find soloutions to our greenhouse warming problems.
Regards, Eric.
I'm going to use water cooling for my 4 channel Leach Amp too. I finished one of the boards the other day, and should be done with one channel real soon. As it neared completion, I came to the dramatic realization that I don't have any large TO-3 heatsinks, and impatient as I am, (you'd never know it from my rate of progress) I didn't want to look for any or buy any, and I'd hate to pay for new heatsinks anyway. So the only thing I can do is make my own heatsinks and make them water-cooled.
With a well design ventilation in the chassis. You don't have to rely on water cool systems.
Using water to cool down solid state devices is great if you like the noise. The disadvantage of using water is that it can destroy the devices if the water cool system leaks. To correct this problem is finding a liquid that does not harm the electronic components. One liquid that is good for this is HFE or hydrofluoroether. This liquid is a dielectric liquid that does not harm electrical devices but it is much much better of consuming the heat from devices. A few computer overclockers are using this liquid to push their processors to extreme limits. One problem of using this liquid is that it cost a lot to use and replace.
Using water to cool down solid state devices is great if you like the noise. The disadvantage of using water is that it can destroy the devices if the water cool system leaks. To correct this problem is finding a liquid that does not harm the electronic components. One liquid that is good for this is HFE or hydrofluoroether. This liquid is a dielectric liquid that does not harm electrical devices but it is much much better of consuming the heat from devices. A few computer overclockers are using this liquid to push their processors to extreme limits. One problem of using this liquid is that it cost a lot to use and replace.
Yes, but heatsinks big enough for convection cooling or even forced air are very expensive and bulky. A water cooling system can be made from scrap aluminum and an old pump and heat exchanger, and some tubing, all of which can be obtained for very cheap or even free. A well-designed water-cooling system can be almost silent, it's just that my big one is noisy because it has a 12" fan spinning at 3450rpm in order to be compact and remove huge amounts of heat from the output devices. My little one should be pretty quiet.
Electro,
As I believe has been covered several times earlier in the thread, a water-cooled system is completely silent once it's been bled of air.
Water is a great deal more benign as regards to the environment than any of the refrigerants. In addition, it's inexpensive and readily available. Not a small benefit if you decide to tear down your system and rebuild it late on a Saturday night.
The electrical conductivity of water isn't as great as you suppose. I've watched bands play live in the rain, and though I thought they were pretty foolish for doing so (in terms of danger of electrocution, at least from tube amps), the amps survived quite nicely. As a kid, I used to tape small, battery-powered motors to the bottom of toy boats to create my own self-propelled craft. The motors were meant for open air, but performed quite well underwater, completely wetted; they did not short out, nor did they fail to work.
No one contests that air-cooled heatsinks work. It's not a question of 'well designed ventillation.' It's a question of availability of lots of large heatsinks for a reasonable sum of money. I'm running two pairs of Aleph 2s off of one water-cooled system, and have no problems...and it was cheap...far, far, cheaper than buying a bunch of the usual aluminum heatsinks.
As I said at the beginning of the thread, it's not for everyone, but never doubt that it works and works well.
Grey
As I believe has been covered several times earlier in the thread, a water-cooled system is completely silent once it's been bled of air.
Water is a great deal more benign as regards to the environment than any of the refrigerants. In addition, it's inexpensive and readily available. Not a small benefit if you decide to tear down your system and rebuild it late on a Saturday night.
The electrical conductivity of water isn't as great as you suppose. I've watched bands play live in the rain, and though I thought they were pretty foolish for doing so (in terms of danger of electrocution, at least from tube amps), the amps survived quite nicely. As a kid, I used to tape small, battery-powered motors to the bottom of toy boats to create my own self-propelled craft. The motors were meant for open air, but performed quite well underwater, completely wetted; they did not short out, nor did they fail to work.
No one contests that air-cooled heatsinks work. It's not a question of 'well designed ventillation.' It's a question of availability of lots of large heatsinks for a reasonable sum of money. I'm running two pairs of Aleph 2s off of one water-cooled system, and have no problems...and it was cheap...far, far, cheaper than buying a bunch of the usual aluminum heatsinks.
As I said at the beginning of the thread, it's not for everyone, but never doubt that it works and works well.
Grey
Water is not a good conductor for heat because of this you need to make a huge radiator. Water is a poor substance for radiating heat.
Amplifers playing in the rain. It must have been a drizzle. Also the amplifier might be completely enclosed or the ventillation is only on the bottom.
I did put a motor in the water several years ago or when I was a kid. The motor did stop because of corrosin. Using AC decreases the chances of corrosin.
My Technics amplifier has a well design ventillation. It only needs one fan when the amplifier is at max power. Yes, there is a 340 watt transformer in it as well as six 100 watt amplifiers. When it is amplifying sound, it does breath. I wonder whats the volecity of the air that it is sucking through the vents. When my amplifier is outputing about a tenth of the power. The transformer heats the chassis of my amplifier. When I turn the amplifier at 12 o'clock, the amplifiers starts cooling itself down. Alot of heat came out of the fan but the chassis cooler than it was.
Amplifers playing in the rain. It must have been a drizzle. Also the amplifier might be completely enclosed or the ventillation is only on the bottom.
I did put a motor in the water several years ago or when I was a kid. The motor did stop because of corrosin. Using AC decreases the chances of corrosin.
My Technics amplifier has a well design ventillation. It only needs one fan when the amplifier is at max power. Yes, there is a 340 watt transformer in it as well as six 100 watt amplifiers. When it is amplifying sound, it does breath. I wonder whats the volecity of the air that it is sucking through the vents. When my amplifier is outputing about a tenth of the power. The transformer heats the chassis of my amplifier. When I turn the amplifier at 12 o'clock, the amplifiers starts cooling itself down. Alot of heat came out of the fan but the chassis cooler than it was.
Somehow, I'm getting the impression that you haven't read the thread...
Water not a good conductor of heat? Compared to air? Hmmm...sounds like you need to get out some physics books.
I'm using the heat exchanger from a heat pump. It's comparatively large, but no larger than two commercial Aleph 2s. Even with two Aleph 2s, I still haven't had to use a fan; convection moves plenty of air through it. To say that water isn't a good radiator of heat is equivalent to saying that aluminum isn't a good radiator of heat. True as far as it goes. But I'm not using a bucket of water any more than a passive heatsink is made of a solid block of aluminum. It's all about increasing the surface area. That's what the fins are for.
A 340W transformer for 600W output? (I'll leave the obvious discrepancy in power capability as an exercise for the reader.) The output stage will be running class B. Very little heat to dissipate, since the bias is so light. Fans do increase the efficiency of passive heatsinks (but they're noisy--something you don't want in a high end system), but your Technics isn't in the same ballpark as an Aleph, as far as heat dissipation goes. Once you build an amplifier with more heat dissipation requirements and see what large heatsinks cost (when you can find them at all), you'll find out why people are always groaning about the price of heatsinks...and why I'm using water.
Grey
Water not a good conductor of heat? Compared to air? Hmmm...sounds like you need to get out some physics books.
I'm using the heat exchanger from a heat pump. It's comparatively large, but no larger than two commercial Aleph 2s. Even with two Aleph 2s, I still haven't had to use a fan; convection moves plenty of air through it. To say that water isn't a good radiator of heat is equivalent to saying that aluminum isn't a good radiator of heat. True as far as it goes. But I'm not using a bucket of water any more than a passive heatsink is made of a solid block of aluminum. It's all about increasing the surface area. That's what the fins are for.
A 340W transformer for 600W output? (I'll leave the obvious discrepancy in power capability as an exercise for the reader.) The output stage will be running class B. Very little heat to dissipate, since the bias is so light. Fans do increase the efficiency of passive heatsinks (but they're noisy--something you don't want in a high end system), but your Technics isn't in the same ballpark as an Aleph, as far as heat dissipation goes. Once you build an amplifier with more heat dissipation requirements and see what large heatsinks cost (when you can find them at all), you'll find out why people are always groaning about the price of heatsinks...and why I'm using water.
Grey
In my opinion, water cooling makes perfect sense- it's an easy way to remove a lot of heat with commonly available and cheap parts. The only thing I don't understand is how to mount the output devices to the system. Grey, how are you doing this? Forgive me if you have posted this- it's been a while since I read through this thread completely.
I am thinking of putting my amps in the basement of my house and going to water cooling- hopefully I may be able to pump (pun intended) up the bias on my A75's (when they're done).
Steve
I am thinking of putting my amps in the basement of my house and going to water cooling- hopefully I may be able to pump (pun intended) up the bias on my A75's (when they're done).
Steve
Steve,
I'm using 1/2" copper pipe. To this I solder a piece of 1/4" thick 1" wide copper bar, like this O- so that the copper bar creates a nice mounting flange for the output devices.
Once you've got that together, you can arrange the rest of your plumbing in any manner that's convenient.
Grey
I'm using 1/2" copper pipe. To this I solder a piece of 1/4" thick 1" wide copper bar, like this O- so that the copper bar creates a nice mounting flange for the output devices.
Once you've got that together, you can arrange the rest of your plumbing in any manner that's convenient.
Grey
And I'm running the coolant through aluminum blocks, as you can see in the picture of the complete cooling system for my big amp. MOSFETs in TO-247 cases will be mounted on all sides of the blocks. There will be 160 such MOSFETs mounted on the completed blocks. For the Leach amps, which use TO-3's, I have the transistors mounted as in the picture, (I only have enough for one channel mounted now) and I will mount 1 1/4" square aluminum tubes between the transistors.
Attachments
- Status
- Not open for further replies.