Zen Mod said:
This is what I thought and figured that at 37C this would bring it to 57C(room 20C). A little hot, but does it not sound better hot? I could also build the chassis as a tubular stand that would provide additional heatsink or a really small fan.
http://www.aavidthermalloy.com/cgi-...M&CType=Natural&ExLength=6.000&LengthUnits=in
As long as each output device is isolated on it's own heatsink. ie you would have two isolated heatsinks per channel. Though it seems like a lot to go through just to eliminate the insulators.
I would question possiblilty of rail voltages live on the heatsinks acting like an antenna to pick up noise.
I would question possiblilty of rail voltages live on the heatsinks acting like an antenna to pick up noise.
I am thinking that a 2"x 2" piece of copper 1/8" or a 1/4" thick spreader to mount the FET directly with no insulation, then insulating the copper spreader to the heatsink. This should result in lower temps at the FETs, no? Also, a small thermoeletric plate at each copper spreader could totally eliminate the threat of cooking them at all. A 12V power supply and some sort of thermistor controller to regulate the FET spreader temp to any setting. Would this not allow for more bias with the same heatsink with only a rise in temp on the sink"
I don't really understand why would you insist on cooling the heatsinks down by reinventing the wheel... Wouldn't it be much simpler doing it the proper way, and increasing the size of the heatsinks? Or even simpler, as Krell KSA50/100 was made, add a noise making fan that will gather the dust for you... 
I'm sure that would work alright, but it just seems to me that it would be a lot easier to insulate the fets first and then do whatever with the spreader.
Yes the spreader should help (not from personal experience but from this thread). Probably bigger heatsink would be easier. Fans will work too, but the noise is an issue. Use Nelson's controller circuit to run them at a lower voltage if you want to do that.
The only real objection I would have is the extra circuitry - I am a bit of a believer in using the simplest way out if I can. It might just work well though.
******************************
On a side note, what are the reports of higher bias levels effects on sonics? How high a bias have people gone... eg those who are water cooling? Has anyone compared higher than normal bias levels (eg 0.8V per side or even 1V per side)?
Fran
Yes the spreader should help (not from personal experience but from this thread). Probably bigger heatsink would be easier. Fans will work too, but the noise is an issue. Use Nelson's controller circuit to run them at a lower voltage if you want to do that.
The only real objection I would have is the extra circuitry - I am a bit of a believer in using the simplest way out if I can. It might just work well though.
******************************
On a side note, what are the reports of higher bias levels effects on sonics? How high a bias have people gone... eg those who are water cooling? Has anyone compared higher than normal bias levels (eg 0.8V per side or even 1V per side)?
Fran
stein2 said:I don't really understand why would you insist on cooling the heatsinks down by reinventing the wheel... Wouldn't it be much simpler doing it the proper way, and increasing the size of the heatsinks? Or even simpler, as Krell KSA50/100 was made, add a noise making fan that will gather the dust for you...
Ok, I will explain why I am interested in reinventing the wheel. What if I wanted one that was more portable? What if I wanted one for a boat? What if I did not want 400W of heat in my already bloody hot tropical bedroom? What if I want to better cool the FET by increasing the direct contact and heat transfer in utilizing copper spreader plates, then insulate, giving a more stable FET temp by increasing its mass? Could be a little copper clamp that allows maximum contact with the FET instead of a cm sq. You should know that the worst juction for heat transfer can be and usually is right at the device itself. If the FET was inserted directly into a heavy copper hat of some type it would likely greatly aid in heat transfer and stability. What if no one ever asked what if? You would be carving pictures of buffalo in a cave somewhere. I am not building your amp. I do not have to conform to the same restraints as you do. If I want to cut a hole in the side of my house to stick a heat pipe and external sink, I can. I may turn my F4 monos into a hot water heater for my shower, get the point? Why not start a line of class A shower heaters for sail enthusiasts?
Here you can see one way of doing it.
http://www.diyaudio.com/forums/showthread.php?s=&threadid=139253
Magura
http://www.diyaudio.com/forums/showthread.php?s=&threadid=139253
Magura
khundude:
I like your copper spreader idea, and I have no doubt it would probably drop your die temps by a few degrees.
The TEC idea, however, is not going to work well at all. They're extremely inefficient, and you'd end up dumping three times the heat into the surroundings in order to keep your die at a slightly cooler temp. You'd need a bunch of 12V supplies, which would make the amp large, heavy, and expensive to run.
I think the point people are trying to make is "why bother". My F5 measures 12" x 12" x 5" and is easily portable. It runs at around 48C and at that temp the fets will probably outlive me. Why spend twice as much, or mess around twice as long, for essentially the same result?
If you want some really interesting and unique cooling solutions, check out the people who have built SOZ amps on the Passdiy website. I think one guy used switch-mode telecomm supplies, and a fish pond as giant water-cooling reservoir. When you need to get rid of kilowatts of heat, then you can start getting creative. For the 64 watts a channel this thing dumps off, it's just not worth it.
Cheers,
Owen
I like your copper spreader idea, and I have no doubt it would probably drop your die temps by a few degrees.
The TEC idea, however, is not going to work well at all. They're extremely inefficient, and you'd end up dumping three times the heat into the surroundings in order to keep your die at a slightly cooler temp. You'd need a bunch of 12V supplies, which would make the amp large, heavy, and expensive to run.
I think the point people are trying to make is "why bother". My F5 measures 12" x 12" x 5" and is easily portable. It runs at around 48C and at that temp the fets will probably outlive me. Why spend twice as much, or mess around twice as long, for essentially the same result?
If you want some really interesting and unique cooling solutions, check out the people who have built SOZ amps on the Passdiy website. I think one guy used switch-mode telecomm supplies, and a fish pond as giant water-cooling reservoir. When you need to get rid of kilowatts of heat, then you can start getting creative. For the 64 watts a channel this thing dumps off, it's just not worth it.
Cheers,
Owen
Sorry, didn't mean no offense...
If you want to, you can do it any way it makes it ok for you and your needs. Especially if you do want to use it under odd conditions. I have no intentions arguing, and placing copper between the OD and the heatsink might be a solution to test and measure if you need to increase the insulation between the copper and aluminum. The results might be pleasing but then, they might be not. Fortunately, there are cheap and reliable multimeters to measure the difference...
If you want to, you can do it any way it makes it ok for you and your needs. Especially if you do want to use it under odd conditions. I have no intentions arguing, and placing copper between the OD and the heatsink might be a solution to test and measure if you need to increase the insulation between the copper and aluminum. The results might be pleasing but then, they might be not. Fortunately, there are cheap and reliable multimeters to measure the difference...
khundude: as stein2 said, absolutely no offence meant... you are of course dead right, if no-one pushed the boat (!) out a bit, no progress would be made.
Be sure to post pics & details of whatever you do - it will serve as a template for others.
I suppose I would like to find out what the benefit is of running at higher bias levels in general - jus tthinking in my own case, would it be worth going to some extra effort to dump more heat and run them harder is all.....
Fran
Be sure to post pics & details of whatever you do - it will serve as a template for others.
I suppose I would like to find out what the benefit is of running at higher bias levels in general - jus tthinking in my own case, would it be worth going to some extra effort to dump more heat and run them harder is all.....
Fran
My main issue is that my ambient room temp is around 32C. This is a high starting point and you can imagine that I do not want more heat inside. Water cooling makes sense to me.
The TEC plates are in hand and won't hurt to try. I can whip up a 12V PS. My problem hear is how to control it? I am seriously thinking of turning my F4 balanced monos into a hot water heater for my shower. I am trying to make my house as simple as possible. It makes sense to me to use the heat for something and the only thing I need it for is the shower water and it is the easiest way to get it out of the room. I would heatsink the power supplies of all audio, TEC, and the PC to the shower as well. My thought for the TEC and spreader plate is not to be more efficient, but to try and lower the die temps. I think that if I sandwich the FET between a couple of water cooled cold plates with a couple of TEC in there too I should be able to greatly reduce the die temp. My thoughts are that if I can signifantly lower the die temps, I can vary the bias and bias them more!
The TEC plates are in hand and won't hurt to try. I can whip up a 12V PS. My problem hear is how to control it? I am seriously thinking of turning my F4 balanced monos into a hot water heater for my shower. I am trying to make my house as simple as possible. It makes sense to me to use the heat for something and the only thing I need it for is the shower water and it is the easiest way to get it out of the room. I would heatsink the power supplies of all audio, TEC, and the PC to the shower as well. My thought for the TEC and spreader plate is not to be more efficient, but to try and lower the die temps. I think that if I sandwich the FET between a couple of water cooled cold plates with a couple of TEC in there too I should be able to greatly reduce the die temp. My thoughts are that if I can signifantly lower the die temps, I can vary the bias and bias them more!
Magura said:Here you can see one way of doing it.
http://www.diyaudio.com/forums/showthread.php?s=&threadid=139253
Magura
That would do exactly what I mean, just excecuted differently. I would use two flat peices of copper instead of one and screw them together to sandwich it. How did it work out?
khundude said:My main issue is that my ambient room temp is around 32C. This is a high starting point and you can imagine that I do not want more heat inside. Water cooling makes sense to me.
The TEC plates are in hand and won't hurt to try. I can whip up a 12V PS. My problem hear is how to control it? I am seriously thinking of turning my F4 balanced monos into a hot water heater for my shower. I am trying to make my house as simple as possible. It makes sense to me to use the heat for something and the only thing I need it for is the shower water and it is the easiest way to get it out of the room. I would heatsink the power supplies of all audio, TEC, and the PC to the shower as well. My thought for the TEC and spreader plate is not to be more efficient, but to try and lower the die temps. I think that if I sandwich the FET between a couple of water cooled cold plates with a couple of TEC in there too I should be able to greatly reduce the die temp. My thoughts are that if I can signifantly lower the die temps, I can vary the bias and bias them more!
When the water cooler breaks down you'll have a China Syndrome in your living room.
Why would it do that and wouldn't a thermistor prevent it? Nobody listens... I am not going to have a bloody 500W space heater in my tropical home where night time ambient is around 30C! I am also not going to leave the whole amp outside. What scares me is when my best friend, a top engineer, looks at the FET/sink combo and laughed, "that sure is primitive." I think in most situation a big chunk of heater in a North American homes is a good thing. It is never a good thing in the jungle. I was even looking as sealed waterproof aluminum project boxes to hold the boards and mounting the coldplate to the box leaving a humidity free chamber.
A little perspective on the environment called the rainforest... It's hot and wet! If I already have the amp sealed as tight as a drum I could float the whole think in my aquarium and it would make more sense that a big heatsink in the room. I beleive Mr. Pass has purchase water cooling systems before, must not have been to dumb an idea. Porsche finally switched the 911 to water cooled in 1993 because the power output was too great for an air-cooled engine. When you cut metal with a power saw you use water or oil to cool it, not air. There is a reason for that, water removes heat 20 times as well as air.
A little perspective on the environment called the rainforest... It's hot and wet! If I already have the amp sealed as tight as a drum I could float the whole think in my aquarium and it would make more sense that a big heatsink in the room. I beleive Mr. Pass has purchase water cooling systems before, must not have been to dumb an idea. Porsche finally switched the 911 to water cooled in 1993 because the power output was too great for an air-cooled engine. When you cut metal with a power saw you use water or oil to cool it, not air. There is a reason for that, water removes heat 20 times as well as air.
I wonder how complicated it would be to fit one CM Gemini II S per output device... Would surely reduce the free space within the housing but then... I bought one for my PC proc and removed all the ventilators except the one in the PSU which I replaced with Gemini's... Cold, cool and above all, silent. ... or to try a similar solution... but don't mind me, I just doubled the required amount of heatsinks on my F5s and keep it at a reasonable 45 degrees celsius... 
Ask him about the efficiency improvement of your heatsinks due to the extreme humidity in the air
Khundude:
You seem to be all over the place with your requirements. Do you want your room to be cooler, or your dies to be cooler? You keep insisting that your ambient temp in the room is too high, but dissipating another 50W per channel to run a TEC is alright?
The heat has to go somewhere, and unless the amp is outside, then it's going to end up in the room where you keep the amp. It doesn't matter if you put it in an aquarium, or use it to heat shower water, you're still going to end up with more heat, unless you come up with a way of moving the heat outside. As for the shower idea, what happens for the other 23.75 hours a day that you're not running the shower? You'd need a large, well insulated reservoir, along with a pump and waterblocks for the fets. If you have all that indoors, then you're still going to find the room warmer, unless you happen to run across some perfect insulation somewhere.
From my experience with PC cooling, water really is great, and yes, I suppose a transistor mounted to a block of aluminum is probably a little old fashioned when you look at it next to the more sophisticated systems in PC's. In the end though, computers are all about high density, and that's why water cooling makes sense in that application. Hi-fi amplifiers on the other hand, don't really need that density, or that level of cooling.
I've attached a picture of the water loop in my PC, so you know I'm not talking out my ****. The PC draws a constant 355 watts from the wall, and closer to 650W under full load. It keeps my office nice and toasty all year round, but at least it's quiet, which was the point of the water cooling.
Tossing your computer in a large vat of oil is all the rage right now in the PC cooling world. Maybe you could toss your F5 in an aquarium full of oil, seal it up, and toss the whole thing outside. I think connectivity could become a little bit of an issue though. Maybe you could try different oils and see how it changes the sound? I bet EVOO would sound best, but only cold pressed.
Cheers,
Owen
You seem to be all over the place with your requirements. Do you want your room to be cooler, or your dies to be cooler? You keep insisting that your ambient temp in the room is too high, but dissipating another 50W per channel to run a TEC is alright?
The heat has to go somewhere, and unless the amp is outside, then it's going to end up in the room where you keep the amp. It doesn't matter if you put it in an aquarium, or use it to heat shower water, you're still going to end up with more heat, unless you come up with a way of moving the heat outside. As for the shower idea, what happens for the other 23.75 hours a day that you're not running the shower? You'd need a large, well insulated reservoir, along with a pump and waterblocks for the fets. If you have all that indoors, then you're still going to find the room warmer, unless you happen to run across some perfect insulation somewhere.
From my experience with PC cooling, water really is great, and yes, I suppose a transistor mounted to a block of aluminum is probably a little old fashioned when you look at it next to the more sophisticated systems in PC's. In the end though, computers are all about high density, and that's why water cooling makes sense in that application. Hi-fi amplifiers on the other hand, don't really need that density, or that level of cooling.
I've attached a picture of the water loop in my PC, so you know I'm not talking out my ****. The PC draws a constant 355 watts from the wall, and closer to 650W under full load. It keeps my office nice and toasty all year round, but at least it's quiet, which was the point of the water cooling.
Tossing your computer in a large vat of oil is all the rage right now in the PC cooling world. Maybe you could toss your F5 in an aquarium full of oil, seal it up, and toss the whole thing outside. I think connectivity could become a little bit of an issue though. Maybe you could try different oils and see how it changes the sound? I bet EVOO would sound best, but only cold pressed.
Cheers,
Owen
