Here we go...
I am thinking of building a JLH Class A or one of it's derivatives. Graham's 2SC5200-based schematics look interesting.
So, to make it more interesting, I am thinking about building a water-cooled system. I would base it on this unit from Zalman:
http://www.zalmanusa.com/usa/product/view.asp?idx=63&code=021
I would need to machine some aluminum blocks to mount the transistors on with water passages. I have some ideas there.
Any thoughts?
I am thinking of building a JLH Class A or one of it's derivatives. Graham's 2SC5200-based schematics look interesting.
So, to make it more interesting, I am thinking about building a water-cooled system. I would base it on this unit from Zalman:
http://www.zalmanusa.com/usa/product/view.asp?idx=63&code=021
I would need to machine some aluminum blocks to mount the transistors on with water passages. I have some ideas there.
Any thoughts?
Interesting idea. That also make "VGA blocks" (http://www.zalmanusa.com/usa/product/view.asp?idx=90&code=013) that might work better for this.
I also use one of their copper CPU coolers in my PC, which is very quiet (http://www.zalmanusa.com/usa/product/view.asp?idx=43&code=005009) but may be too loud with two of them running.
I also use one of their copper CPU coolers in my PC, which is very quiet (http://www.zalmanusa.com/usa/product/view.asp?idx=43&code=005009) but may be too loud with two of them running.
When you say "any thoughts"... "thoughts" are only as far as I've got as yet. I've been doing a bit of reading and I think you'll find that copper has nearly twice the thermal conductivity as aluminium so I figure that when I get around to building something similar I'll look at copper. It's not too hard to machine and scrap metal people get 10mm thicknesses from heavy duty electrical installations. The other "thought" is oil instead of water. Again, just theory as yet but oil takes up thermal energy more readily than water. So in theory it is better at removing heat from small spaces (semiconductor junctions) than H2O. On the other hand I haven't checked any reduced flow issues that might arise with the different viscosity. The other disadvantage is that oil give up its energy more slowly than water but I figure that most of us could knock up a decent size heat exchanger from old car radiators or something. So to summerize "thoughts" I think copper and oil could be an elegant way of achieveing good technical results in this area. (and no doubt the extreme subjectivists will insist on "extra virgin olive oil" to get "smooth, pure highs and more faithfull reproduction"
BTW I think there are a couple of other threads on water cooling in the archives. I've got them in my "favourites" if the search mode doesn't deliver results quickly.
BTW I think there are a couple of other threads on water cooling in the archives. I've got them in my "favourites" if the search mode doesn't deliver results quickly.
Hi!
Oil is fine as Jonathan Bright said but water is easier to deal with in many cases. If you get a leak with oil you'll have less fun cleaning
10 years ago I converted a ceramic transmitting tube called 4CX250b from air to liquid cooled. This tube is still running in a high power VHF amplifier. Started with oil but ended with water because of leak inside the amplifier, it was a h*** to clean it up.
With five liter of water and a small heater radiator from an old car and a small quiet water pump I got rid of all noise from what usually comes from air cooling. The system had to take care of more than 400 watts of plate dissipation.
I would go for water
By the way I made an high current psu water cooled also but that was a little over kill I think
Good luck!
Bjorn
Oil is fine as Jonathan Bright said but water is easier to deal with in many cases. If you get a leak with oil you'll have less fun cleaning
10 years ago I converted a ceramic transmitting tube called 4CX250b from air to liquid cooled. This tube is still running in a high power VHF amplifier. Started with oil but ended with water because of leak inside the amplifier, it was a h*** to clean it up.
With five liter of water and a small heater radiator from an old car and a small quiet water pump I got rid of all noise from what usually comes from air cooling. The system had to take care of more than 400 watts of plate dissipation.
I would go for water
By the way I made an high current psu water cooled also but that was a little over kill I think
Good luck!
Bjorn
Nelson Pass said:
Jonathan Bright said:
I believe water has more than twice thermal conductivity of oil
so it is preferred as the heat transfer medium.
This reminds me years ago when I was a Toolmaker. When
heat treating steels, water was used for a quenching medium
when highest hardness was required as a result of most rapid
cooling. Oil was used when a less hard but tougher treatment
was required as a result of it's slower cooling.
Fun stuff.
Terry
So, was talking to my brother about this (cooling medium) last night. He does a lot of machining work.
His first question was, "How hot do these things get?"
I told him I would need to keep temps below about 60C, so the answer was water. If you go close to 100C or above, oil is the solution (water boils).
Anyway, discussion moved quickly to designing a silent water coling system for amps. I'll post updates when and if they become available.
His first question was, "How hot do these things get?"
I told him I would need to keep temps below about 60C, so the answer was water. If you go close to 100C or above, oil is the solution (water boils).
Anyway, discussion moved quickly to designing a silent water coling system for amps. I'll post updates when and if they become available.
I am building 6 additional Pass amplifiers with the two Aleph's that i already have, together with the XVR1 as an active 4-way.
6 for 3-way loudspeakers, 2 for Susan Parker's Subwoofers.
Given the tremendous heat i am planning to use the water from my swimming pool to cool the amps.
(and heat the pool in winter
)
On board of ships overheated steam or thermal oil is used for heating purposes above 100C/222F.
Steam mostly, because heat transfer with thermal oil requires huge amounts of it.
(think heating 10.000 tons of bitumen to 300 F )
Some oil tankers have coil heaters too, to prevent crude oil from become too thick.
Thermal oil is used for special purposes, when steam is a risk.
The most regular used cooling fluid was Freon, R12, for frigges, aircon, etc.
R12 was banned because of environmental hazard, by the HFC R134(a).
Nowadays there are alternatives for R134, because that too has downsides, it causes cancer.
I was thinking of machined copper ducts, with added alloy Papst vented heat tunnels attached.
The copper ducts would be cooled by water, driven from a floor central heating pump, a regular Grundfoss.
In case of emergency, the Papst vents are relay switched to full capacity for cooling, combined with thermal breakers for the powersupply of the amps.
My initial thought is using compressed air connectors for the hosing.
That way i have plenty cooling capacity, no sounds, and the heat doesnt make me dress as DestroyerX.
6 for 3-way loudspeakers, 2 for Susan Parker's Subwoofers.
Given the tremendous heat i am planning to use the water from my swimming pool to cool the amps.
(and heat the pool in winter
)On board of ships overheated steam or thermal oil is used for heating purposes above 100C/222F.
Steam mostly, because heat transfer with thermal oil requires huge amounts of it.
(think heating 10.000 tons of bitumen to 300 F )
Some oil tankers have coil heaters too, to prevent crude oil from become too thick.
Thermal oil is used for special purposes, when steam is a risk.
The most regular used cooling fluid was Freon, R12, for frigges, aircon, etc.
R12 was banned because of environmental hazard, by the HFC R134(a).
Nowadays there are alternatives for R134, because that too has downsides, it causes cancer.
I was thinking of machined copper ducts, with added alloy Papst vented heat tunnels attached.
The copper ducts would be cooled by water, driven from a floor central heating pump, a regular Grundfoss.
In case of emergency, the Papst vents are relay switched to full capacity for cooling, combined with thermal breakers for the powersupply of the amps.
My initial thought is using compressed air connectors for the hosing.
That way i have plenty cooling capacity, no sounds, and the heat doesnt make me dress as DestroyerX.
So how are you guys going to protect the amplifer in case of reduced or cut off water flow?
At that time I built my water cooled amp I searched for any type of flow interlock switch which would turn of the psu in case of cut off or reduced water flow. I did find but they were expensive.
Instead I used an flow indicator but I wouldn't recommend a solution with only flow indicator tough it's nice to look at with a small lamp behind it.
I did use an ordinary small central heating pump as Jacco vermeulen did, lots more flow than needed but I got it for free
A small heat exchanger from an old crashed SAAB 900 with an quiet fan attached at it did the job. Even tough I ran my amp pretty hard the cooling fan didn't start that often. The temp trip circuit was set to some were in the neighbourhood of 60-70C.
I often see water cooled systems made for computer which looks real overkill thinking of how much heat to remove. I heated my foot bath while running my amp and the high voltage had some cleaning effect as well That was an joke do not try!!!
Bjorn
At that time I built my water cooled amp I searched for any type of flow interlock switch which would turn of the psu in case of cut off or reduced water flow. I did find but they were expensive.
Instead I used an flow indicator but I wouldn't recommend a solution with only flow indicator tough it's nice to look at with a small lamp behind it.
I did use an ordinary small central heating pump as Jacco vermeulen did, lots more flow than needed but I got it for free
A small heat exchanger from an old crashed SAAB 900 with an quiet fan attached at it did the job. Even tough I ran my amp pretty hard the cooling fan didn't start that often. The temp trip circuit was set to some were in the neighbourhood of 60-70C.
I often see water cooled systems made for computer which looks real overkill thinking of how much heat to remove. I heated my foot bath while running my amp and the high voltage had some cleaning effect as well
That was an joke do not try!!!Bjorn
Reduced size would be the goal. Fun too (for me).
I keep seeing these giant amp cases with tiny little gain clones or ClassA mosfet amps in them. Wouldn't it be nice to pack 6 or 7 channels of that in one box? It's either fans or water. I think we could have silent (or very near silent) water cooling.
As for protection in case of flow failure, I think the easiest way to do that would be with a temp sensor used to cut the mains via a relay.
Another interesting thing I heard was the method of fitting aluminum fins to copper pipe: The fins are cut or machined to just barely fit over the pipe. Once in place, the pipe is filled with high-pressure nitrogen, which stetches it just enough to lock the fins.
I keep seeing these giant amp cases with tiny little gain clones or ClassA mosfet amps in them. Wouldn't it be nice to pack 6 or 7 channels of that in one box? It's either fans or water. I think we could have silent (or very near silent) water cooling.
As for protection in case of flow failure, I think the easiest way to do that would be with a temp sensor used to cut the mains via a relay.
Another interesting thing I heard was the method of fitting aluminum fins to copper pipe: The fins are cut or machined to just barely fit over the pipe. Once in place, the pipe is filled with high-pressure nitrogen, which stetches it just enough to lock the fins.
My first thought for protection was using a circuit with temperature regulating transistors glued on the heatsinks, with a preset temperature after which a relay shuts the powerlines.
A more sophisticated one could be an active bias regulation when cooling is not sufficient.
As mentioned in previous posting, i thought of making double cooling units, thermally attached to eachother.
A couple of months ago i bought a box of new Papst 30 dBa vents, 1 Lbs Alloy models that will outlive me.
I have enough of them to put several vents on each of the 8 amps.
If water cooling would fail, the temperature controlled transistors on the heatsink would change the bias of the amps, in something like 2 steps, with increasing transistor current activating a Schmitt trigger that shuts the amps down if necessary.
That would enable the amplifiers to be still functional, not only when the pump blocks, my swimmingpool dries out, but also during maintenance, even enables the amps to be separated and go solo.
For the watercooling sink i was considering Alloy, that is not more difficult to machine than copper, but in the size required more easily available, threading holes in alloy makes no difference.
Another option is to use double waterpumps, put a Siemens Micromaster inverter on the waterpump for speed control, temperature switching the second pump.
That would enable the amps to reach equilibrium faster, and have a backup pump and hoses.
Digital inverters have added control inputs, and can be programmed.
With the amplifier towers placed to the wall it will probably only take 1 feet of hose to connect to the wall sockets.
I am thinking braided hoses: i use them for connecting the airco cooler to the TH700R4 automatic on my Chevy.
They look great, they can be bought custom sized with connectors pressed-on, withstand +100 Bar, and survive eternity.
A regular Siemens Micromaster that controls 750watts can be had on the web for like $25.
All it takes is replacing the high voltage caps, cleaning the inch of dirt inside, and putting new thermal grease underneith the SemiKron Mosfet output module.
A more sophisticated one could be an active bias regulation when cooling is not sufficient.
As mentioned in previous posting, i thought of making double cooling units, thermally attached to eachother.
A couple of months ago i bought a box of new Papst 30 dBa vents, 1 Lbs Alloy models that will outlive me.
I have enough of them to put several vents on each of the 8 amps.
If water cooling would fail, the temperature controlled transistors on the heatsink would change the bias of the amps, in something like 2 steps, with increasing transistor current activating a Schmitt trigger that shuts the amps down if necessary.
That would enable the amplifiers to be still functional, not only when the pump blocks, my swimmingpool dries out, but also during maintenance, even enables the amps to be separated and go solo.
For the watercooling sink i was considering Alloy, that is not more difficult to machine than copper, but in the size required more easily available, threading holes in alloy makes no difference.
Another option is to use double waterpumps, put a Siemens Micromaster inverter on the waterpump for speed control, temperature switching the second pump.
That would enable the amps to reach equilibrium faster, and have a backup pump and hoses.
Digital inverters have added control inputs, and can be programmed.
With the amplifier towers placed to the wall it will probably only take 1 feet of hose to connect to the wall sockets.
I am thinking braided hoses: i use them for connecting the airco cooler to the TH700R4 automatic on my Chevy.
They look great, they can be bought custom sized with connectors pressed-on, withstand +100 Bar, and survive eternity.
A regular Siemens Micromaster that controls 750watts can be had on the web for like $25.
All it takes is replacing the high voltage caps, cleaning the inch of dirt inside, and putting new thermal grease underneith the SemiKron Mosfet output module.
WorkingAtHome said:
Why not the other way around?
Size the pipe slightly larger than the alloy fin, put the pipe in cooled nitrogen, slide the fins over it.
When the pipe heats up again, it shrinks the fins on it.
I did that more than once, my dad even tought me the trick when i was 10.
Including hardening metal by cooling in water or oil, i made Jungle Machete's and hardened arrow tips as a youngster.
He started as a toolmaker too, they called it tool master then,like in Germany.
Hate to be a nay sayer, but water and electronics dont mix well too often. The Zalman solution looks elegent, yet I have a feeling it won't run reliably for long and in a few years it will be collecting dust after a small spill.
Having said that I have no empirical data supporting my paranoia
, but if you are doing it for the fun of it, sure go for it, I am sure it will work nicely for a given time.
Having said that I have no empirical data supporting my paranoia
, but if you are doing it for the fun of it, sure go for it, I am sure it will work nicely for a given time.
In my previous house i had floorheating installed at newbuild.
At the time small airco units came on the market overhere.
I installed a radiator with a vent on the outside wall, connected by two hoses to the floorheating installation with an added GrundFoss waterpump.
In summertime the floorheating pipes cooled my livingroom, temperature switched.
When i sold the house it had not misfunctioned a single time, those pumps are made for +100.000 Hrs.
An idea would be to seal the amplifier housing, but if the watercooling unit is sealed properly and connections are below the amp chassis there is no risk in combining the two, i think.
(but dont quote me)
Any idea how to get rid of the heat of 6 Aleph4's and 2 Zen-XA's, K-ampster ?
At the time small airco units came on the market overhere.
I installed a radiator with a vent on the outside wall, connected by two hoses to the floorheating installation with an added GrundFoss waterpump.
In summertime the floorheating pipes cooled my livingroom, temperature switched.
When i sold the house it had not misfunctioned a single time, those pumps are made for +100.000 Hrs.
An idea would be to seal the amplifier housing, but if the watercooling unit is sealed properly and connections are below the amp chassis there is no risk in combining the two, i think.
(but dont quote me)
Any idea how to get rid of the heat of 6 Aleph4's and 2 Zen-XA's, K-ampster ?
Hi there,
I still have some of these if anyone is interested.
http://www.diyaudio.com/forums/showthread.php?s=&threadid=16822&highlight=
Regards
Anthony
I still have some of these if anyone is interested.
http://www.diyaudio.com/forums/showthread.php?s=&threadid=16822&highlight=
Regards
Anthony
Grey made a couple of Alephs and cooled them with water quite a while ago. He used a heat pump condenser (radiator) and no fan.
There are quite a few discussions of this and a search is essential
For example:
To create the cooling block to mount transistors to, he first got a 1/8" thick copper plate long enough to mount them in a row . It was somewhat taller than the mounting area needed for the transistors.
He then soldered a piece of copper pipe along the face close to one edge for coolant.
viola!:
1. It is easy to drill holes in the plate for the output devices,
2. attach hoses to the pipe for coolant -water is best by far!
3. no machining!
4. All copper -great heat transfer!
5.Cheap!!
6. No excuses!
There are quite a few discussions of this and a search is essential
For example:
To create the cooling block to mount transistors to, he first got a 1/8" thick copper plate long enough to mount them in a row . It was somewhat taller than the mounting area needed for the transistors.
He then soldered a piece of copper pipe along the face close to one edge for coolant.
viola!:
1. It is easy to drill holes in the plate for the output devices,
2. attach hoses to the pipe for coolant -water is best by far!
3. no machining!
4. All copper -great heat transfer!
5.Cheap!!
6. No excuses!
WorkingAtHome said:
I hate to destroy the fun, but it's a fact that you can not reduce the size of the heatsinks because you use water cooling or what ever else for that kind of matter. With water cooling you are only transporting the heat from one place to another, the heatsinks need to be just as big or even bigger.
The reason to use water cooling for a computer is that you have no way to mount adequately big heatsinks where the heat is, so you must transport it to a place with a little more space, as in "outside the box". Such a problem there are many easier and better ways to work around on an amp.
Magura
Magura said:
I hate to destroy the fun, but it's a fact that you can not reduce the size of the heatsinks because you use water cooling or what ever else for that kind of matter. With water cooling you are only transporting the heat from one place to another, the heatsinks need to be just as big or even bigger. But Jacco, having huge heatsinks is what makes class-A amps look so cool!
The reason to use water cooling for a computer is that you have no way to mount adequately big heatsinks where the heat is, so you must transport it to a place with a little more space, as in "outside the box". Such a problem there are many easier and better ways to work around on an amp.
Magura
Yes, the only advantage I see is to have the thermal exchange unit external to your home or listening room, and have a "cool" little box with 4 Aleph's sitting in your living room.
Jacco, heavy duty pumps made by organizations (and designs) that have had decades of experience have higher probability of reliable operation than a new niche product. Given that ZALMAN did their tests well, does not preclude my tendency to run away from what I see as unecessary risk. They may still work, but I'd rather not get carried away by a new toy.... but thats me.
I used to overclock my CPU and it was exciting at the time, all the condensation leaking from the Peltier strips was just precious.
.I now wait and learn from other people's disasters... I mean "experiences"
. I will give you that it will be interesting to work on such a project....- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.