My tec liquid chiller project

I've kept up with computer technology and hardware for about 10 years
but never built my own until February. I went into it whole hog - water cooling
and overclocking. It didn't take me long to start thinking about sub-ambient
cooling. Obviously, the main difficulty in cooling a processor is the fact that
150+ watts are being generated in area less than 1.5 sq inches ( high overclocks ).
Phase change gets the job done, but involves skills that I don't
possess ( I'm an electrician, not a HVAC guy ). So, I gravitated towards TECs.
I decided that the best results would come from building a liquid chiller
using multiple TECs. After deciding on my design, I found this part of the
forums, which has given me additional information :).

Here is my design:

One 12" x 1.75" copper chamber mounted to the cold side of the
TECs that is part of the cooling loop of the processor.

Two 12" x 2" copper chambers mounted on the hot side of the
TECs and part of the cooling loop for the TECs.

Two PA 120.3 with Sanyo Denki fans to cool the TECs.

Blueline 30HD pump for the hotside loop.

PDD-20 pump (rebadged Iwaki MD 20) and 24volt power supply
for the coldside loop.

Two server PSUs for power

14 TEC1-12706 tecs

I was going to use 10 TECs in an attempt to get the best temps and stay
under 600 watts used, but right before assembly I decided to add 4 more
TECs ( I had enough room for them and enough power ).

Here are some pictures:

http://img127.imagevenue.com/loc1086...122_1086lo.JPG

These are the copper bars I used. Four of the bars are 3/8" thick and two are
3/16" thick.

http://img148.imagevenue.com/loc869/..._122_869lo.JPG

I used a hand-me-down Black&Decker router and a two flute spiral cut router bit to make the chambers.

http://img169.imagevenue.com/loc1014...122_1014lo.JPG

I had some trouble at first keeping the depth consistent with that cheap router, but on average,
the dimensions of the middle chamber is 11.5" x 1.5" x 3/8"
deep, and for the outer chambers it's 11.5" x 1.5" x 3/16" deep.




http://img177.imagevenue.com/loc384/..._122_384lo.JPG

To improve the heat transfer coefficient of my chiller, I inserted copper mesh in each chamber.

http://img195.imagevenue.com/loc401/..._122_401lo.JPG

I applied silver-bearing solder paste on the inside of the chambers in order
to improve the heat flow into/out of the copper mesh.

http://img137.imagevenue.com/loc1035...122_1035lo.JPG

I clamped each chamber and applied heat ( propane burner
and a thick aluminum griddle to act as a heat spreader )
and soldered the halves of the chambers together.
It was difficult ( little experience at soldering ) and not completely successful.

http://img129.imagevenue.com/loc869/..._122_869lo.JPG

All three chambers have three 1/4" tubes on each end and with a 1/2" copper stubout soldered on each end.


http://img138.imagevenue.com/loc628/..._122_628lo.JPG

Here is one of the chambers after some sanding ( I had to do much more sanding to get them flat enough ).
The gray stuff is JB weld, which is necessary
to seal the various leaks that appeared when I did a leak test.



http://img207.imagevenue.com/loc397/..._122_397lo.JPG

This is a test run on the hotside loop to check out how much flow
it will have.

http://img103.imagevenue.com/loc993/..._122_993lo.JPG
http://img166.imagevenue.com/loc792/..._122_792lo.JPG

As you can see, the flow is a bit uneven between the two parts of the
loop. Also, the flow is a touch lower then I had hoped for ( I wanted
about 2.0 gallons a minute), but I think it will be enough. According
to Martin's test on the PA120.3 and with the fans I am going to use,
the air in / water out delta should be no more than 7 to 8 deg Celsius.


http://img202.imagevenue.com/loc337/..._122_337lo.JPG

This the test run on the coldside loop. I used an apogee gt and an apogee
gtx to simulate, the best I could, a more restrictive block.

http://img194.imagevenue.com/loc224/..._122_224lo.JPG

By adjusting the potentiometer on the Meanwell 24 volt power supply,
I could vary the flow rate from 1.29 gal/min to 1.72 gal/min.

http://img215.imagevenue.com/loc389/..._122_389lo.JPG
http://img126.imagevenue.com/loc1196...122_1196lo.JPG

Here are a couple of pictures of the chiller assembled. I had to figure out
how to sandwich the TECs and chambers together, and I had these
heatsinks that I got cheap off of eBay, so I decided to use them on
the hot side also. Using the info that Uncle Jimbo gave on how to apply
enough force to the TECs, I think that I have about 250 lbs per sq inch
on the TECs.

I still have several things to do. I have to mod my PSUs, construct my
open test bed, and eventually build a case for the chiller.

I will be testing this on a hot QX6700 (B3) and see how high of a stable
overclock I can get.

wow that is one awesome sandwich. 14 TECs, and both WC and air on the hot side - amazing.

You didn't say how you plan to power the TECs. That will make a difference in the performance, obviously. My guess is your heat sinks have performance of around .3C/W, so 4 on the hot plate gives you .075C/W, not bad at all. That will certainly reduce the load in your hot side loop.

Running those 12706 TECs as 7 series pairs, you will have 6V on each TEC and be at CoP of 1.8 on the 10C dT curve, and 1.3 on the 20C dT curve, pretty decent. That puts you at 2.25A. That's 13.5W in per TEC, or a total of 94.5W per side. You move 170W across dt of 10C, for total heat load of 265W, and 123W across 20C, for total heat load of 217W.

Hard to say what the thermal performance of the hot side loop is, but my guess is the plates themselves have maybe .035 or a little less. From the excellent write-up at http://www.xtremesystems.org/forums/...ad.php?t=77260
we know that the PA120 has performance between .03C/W and .02C/W for higher CFM. So you could get to a total of .055C/W from the hot side of the TEC to ambient. I'm assuming you have a separate pump and rad for each side based on the no holds barred way you did the rest of this.

There is a lot of thermal interface to consider in your sandwich, but if we take the combined cooling of both the air and water, you get to maybe .035 as total thermal resistance to ambient. That should give us some idea of where the hot side will be, and how much heat you move. Looking at the 10C dT TEC numbers, the 265W heat load will put the hot side at 8.6C above ambient, so that's probably close to where you will be when you first start things up.

You will be pulling a total of 340W out of the cold side loop, which will drop water going through the cold side chamber about .7C per second at 2GPM. As long as your total heat load in the cold side is less than 340W, the temp will continue to drop. But as it does, the TEC moves less power, since dT increases.

At dT of 30C, your TECs will move a total of 140W. That puts your hot side about 6C above ambient with your cooling, so if the heat load is less than 140W, you will stabilize at around 1C. If the heat load is higher, it will stabilize a little higher. Still, those are pretty good numbers.

If you are game to do either a chopper supply or a bang-bang controller for the TEC power, you could run off the 12V line and control power based on cold side temp. At startup, with 12V, you have about 390W in per side, and move 290W per side. The total heat load is 680W per side, which gives you a rise of 22C with your cooling. But you are pulling 580W out, which drops the temp at over 1.1C per second. With that much power, you can dial in almost any cold side temperature - you can move 170W at 50C dT, which puts you at -5C or so with 25C ambient.

If you need any help with a power controller, I'll be glad to help. With the cooling power you have available, that seems like the way to go if you want sub-zero.

This is certainly impressive and came out of nowhere! I commend your efforts and hope it works as well as it has potential too.

Quote:

---

Originally Posted by Uncle Jimbo

You didn't say how you plan to power the TECs.

---

He's using 2 server PSUs, sounds like he knows what he's doing in that area.

Quote:

---

Originally Posted by Uncle Jimbo

my guess is the plates themselves have maybe .035 or a little less.
we know that the PA120 has performance between .03C/W and .02C/W for higher CFM. So you could get to a total of .055C/W

---

Did you notice he said he is using 2 PA120.3 radiators? That brings him down to around 0.045-0.050 C/W for that part of your estimation surely?

I really like this set up but I have a question/concern

Quote:

---

One 12" x 1.75" copper chamber mounted to the cold side of the
TECs that is part of the cooling loop of the processor.

---

Seems just a little to narrow seems how a tec is 1.97 That means some of the cold side is not touching and could cause a problem.
Did you use some kind of cold plate when mounting?

TEC1-12706s are generally 40mm (1.57" to you imperial types).

Quote:

---

Originally Posted by Scarlet Infidel

He's using 2 server PSUs, sounds like he knows what he's doing in that area.



Did you notice he said he is using 2 PA120.3 radiators? That brings him down to around 0.045-0.050 C/W for that part of your estimation surely?

---

I did the calculation for each 'side' and then combined to get total cooling, my assumption is that there is one PA120 on each side. His total sandwich has a number of interfaces so this is all guesswork until we see some real numbers.

With the big surface area he has, Rt between the cold chamber and the TEC cold side should be insignificant. TEC hot side to the hot chamber is critical but again, with his care about pressure, I assume that is not significant. So we basically have Rt for the hot chamber plus Rt for the PA120 on the water loop. We can say pretty accurately that with big fans, the PA120 will deliver .02 water to ambient, so the big question is the hot side chamber.

If that was a pin grid block, I would expect something under .015, but with the home-made channel, that's probably optimistic. Only some testing will show what he actually has there. My guess of .035 is based on surface area transfer at 2GPM of .4C/W per inch, but it could be half of that and it would not surprise me. So my number is probably worst case, and his total per side WC Rt could be .03 or .025.

The air cooling is a SWAG too - but those look like standard sinks with 80mm fans, which are usually around .3C/W. Rt per side is a straight up parallel so Rt(tot) is the Rt per sink / 4 or .075.

Calculating unequal parallel equivalent resistance is 1/Req = 1/R1 + 1/R2 + ...
so in this case, for my SWAG, 1/Rtot = 1/.055 + 1/.075. For LCD, we can use 1/Rtot = 1/(11*.005) + 1/(15*.005) or .005/Rtot=1/11+1/15 => 15/165 + 11/165 => 26/165 = .1576. Solving for Rtot yields .031 so I SWAGed .035 to account for everything I don't know.

Total Rt to ambient is half of that, or .0175

Quote:

---

Originally Posted by littleowl

I really like this set up but I have a question/concern

Seems just a little to narrow seems how a tec is 1.97 That means some of the cold side is not touching and could cause a problem.
Did you use some kind of cold plate when mounting?

---

littleowl - I wondered about that too. But then I looked up the 12706, and as S-I says, it is a 40mm so 1.57 inches. Close but should be OK with careful placement.

But I wonder why leuler chose those TECs. Cost does not seem to be much of an issue for him. 12709's are 10 for $50 on eBay and 12710's are about $6 each. Those are also 40mm TECs.

Just shifting to the 12710s changes the heat transfer a lot, basically moves 1.7 times the heat, but at 1.7 times the power. He could move a total of 1000W!!

With power control on the TECs, I think the 12710s would be a better choice, and it should be easy for leuler to swap them out. But he can do that when the design is checked out, or any time really.

Well, here's the thing. Whilst grappling with my TEC simulator program I found that when you are aiming for a moderately high dT with a low heatload and not worrying about COP then going for a larger TEC isnt always best. I wasn't sure if this is because the maths im using is still wrong, but let me give you an example.

Theoretically (numbers chosen to make it easier):
Lets say he only wants to cool 154w load with a dT of 30c. This is 11w per TEC if he uses 14. His cooling is so good that his hot side is 25c.

With 12706 this seems to be 3A at 7.5V * 14 so 315w overall.

With 12710 this is more like 4A at 6.25V * 14 so 350w overall.

Now the numbers I have chosen are a bit silly, but it gives you something to think about. Have I made a mistake? Also, with a real cooling system the cold side on the 12706 will be cooler than with the 12710.

Quote:

---

Originally Posted by Uncle Jimbo

wow that is one awesome sandwich. 14 TECs, and both WC and air on the hot side - amazing.

You didn't say how you plan to power the TECs. That will make a difference in the performance, obviously. My guess is your heat sinks have performance of around .3C/W, so 4 on the hot plate gives you .075C/W, not bad at all. That will certainly reduce the load in your hot side loop.

Running those 12706 TECs as 7 series pairs, you will have 6V on each TEC and be at CoP of 1.8 on the 10C dT curve, and 1.3 on the 20C dT curve, pretty decent. That puts you at 2.25A. That's 13.5W in per TEC, or a total of 94.5W per side. You move 170W across dt of 10C, for total heat load of 265W, and 123W across 20C, for total heat load of 217W.

Hard to say what the thermal performance of the hot side loop is, but my guess is the plates themselves have maybe .035 or a little less. From the excellent write-up at http://www.xtremesystems.org/forums/...ad.php?t=77260
we know that the PA120 has performance between .03C/W and .02C/W for higher CFM. So you could get to a total of .055C/W from the hot side of the TEC to ambient. I'm assuming you have a separate pump and rad for each side based on the no holds barred way you did the rest of this.

There is a lot of thermal interface to consider in your sandwich, but if we take the combined cooling of both the air and water, you get to maybe .035 as total thermal resistance to ambient. That should give us some idea of where the hot side will be, and how much heat you move. Looking at the 10C dT TEC numbers, the 265W heat load will put the hot side at 8.6C above ambient, so that's probably close to where you will be when you first start things up.

You will be pulling a total of 340W out of the cold side loop, which will drop water going through the cold side chamber about .7C per second at 2GPM. As long as your total heat load in the cold side is less than 340W, the temp will continue to drop. But as it does, the TEC moves less power, since dT increases.

At dT of 30C, your TECs will move a total of 140W. That puts your hot side about 6C above ambient with your cooling, so if the heat load is less than 140W, you will stabilize at around 1C. If the heat load is higher, it will stabilize a little higher. Still, those are pretty good numbers.

If you are game to do either a chopper supply or a bang-bang controller for the TEC power, you could run off the 12V line and control power based on cold side temp. At startup, with 12V, you have about 390W in per side, and move 290W per side. The total heat load is 680W per side, which gives you a rise of 22C with your cooling. But you are pulling 580W out, which drops the temp at over 1.1C per second. With that much power, you can dial in almost any cold side temperature - you can move 170W at 50C dT, which puts you at -5C or so with 25C ambient.

If you need any help with a power controller, I'll be glad to help. With the cooling power you have available, that seems like the way to go if you want sub-zero.

---

I plan to run the TECs at 12 volts. I bought some server PSUs from eBay
that have three 12 volt rails rated at 15 amps each.

My worst case calculations are darn close ( about 2 degrees higher ) to
the temps you have given. I've been working out the performance of my
hotside cooling from the theoretical side ( using a heat transfer model
that I think applies in this case ). To have someone with experience
give numbers that are in the neighborhood of what I came up with is
reassuring. However, my calculations don't include the effect of the
copper mesh. My hope is that it improves the heat transfer by factor of
2.5, which would lead to a hotside rise of 14 - 15 degrees above ambient.
In the end, the performance is going to depend on how well I have
constructed the chiller. I lack the tools and experience to get the very
best results. We just have to see if I did things good enough to get
good performance.

I would like to thank you Uncle Jimbo. Your knowledge of TECs has been
valuable to me. I studied for this project for about 3 months, and I
have fairly strong foundation in math and physics, so most of the stuff
I have researched has sunk in. But, a good dose of experience is just
as important. And you have supplied that to us.

As far as your offer for help on a power controller, if this chiller works
pretty well, I will take you up on it. It would take me fair amount of
time figure out how to build one by myself. I have thought about using
a couple of 3 position switches and switch between 12 and 5 volts,
but a power controller would be much more versatile.

Woah, I would never run those TECs at 12v! You'll be using 750w+ of power just to run it. Going down to 8-10v might lose you 5-10c but you'll easily half your power usage if not third it.

Out of interest, so I can run some numbers (I know some of this has been mentioned but I'd like to be clear):

What heat load do you intend to cool?
What water temperatures are you aiming for on the cold side?
What do you estimate as your hot (whole cooling system) and cold side (just between the blocks and the water) C/W?

Quote:

---

Originally Posted by leuler

I plan to run the TECs at 12 volts. I bought some server PSUs from eBay
that have three 12 volt rails rated at 15 amps each.

My worst case calculations are darn close ( about 2 degrees higher ) to
the temps you have given. I've been working out the performance of my
hotside cooling from the theoretical side ( using a heat transfer model
that I think applies in this case ). To have someone with experience
give numbers that are in the neighborhood of what I came up with is
reassuring. However, my calculations don't include the effect of the
copper mesh. My hope is that it improves the heat transfer by factor of
2.5, which would lead to a hotside rise of 14 - 15 degrees above ambient.
In the end, the performance is going to depend on how well I have
constructed the chiller. I lack the tools and experience to get the very
best results. We just have to see if I did things good enough to get
good performance.

I would like to thank you Uncle Jimbo. Your knowledge of TECs has been
valuable to me. I studied for this project for about 3 months, and I
have fairly strong foundation in math and physics, so most of the stuff
I have researched has sunk in. But, a good dose of experience is just
as important. And you have supplied that to us.

As far as your offer for help on a power controller, if this chiller works
pretty well, I will take you up on it. It would take me fair amount of
time figure out how to build one by myself. I have thought about using
a couple of 3 position switches and switch between 12 and 5 volts,
but a power controller would be much more versatile.

---

I have also used server supplies for this work - ProLiant DL380 supplies are 12V at 32A, and you can also buy the redundant power sharing board to make it plug-in. I have gotten those for as low as $15 each.

The risk of failure in doing power control of TECs is from thermal cycling. Since you have both the hot side and cold side on water, they won't have much risk, because the heat won't change quickly between power on and power off.

That means you can use a bang-bang controller. A simple thermostat actually does a good job. I used a $10 mercury thermostat on one test. Because it was designed to work in air, I put it in an insulated box mounted on a piece of copper that was stuck into the reservoir. It didn't really matter that the temp switched a little above the 'real' temperature, as long as it was consistent I could put it where I wanted. I had to bend the sensor a little to get it to read down where I wanted since it was designed for 18C to 40C and I wanted to control around 5C.

When the water temp rose above the setpoint, the thermostat triggered a 12V relay which turned on the TEC power, and when the water got cold enough, it turned off. There is a degree or two of hysterisis in that kind of thermostat, so it doesn't turn on and off very fast. In one setup I tested, the whole body of water in the cold loop was about 2.5G and the cooling power was 250W for a 140W heat load. Once water got to 5C, the relay opened, and when it heated up, it closed. It was on for 2 min and off for 30 sec under load, and at idle it was about 1 min on and 1 min off. Water was controlled to between 4.5C and 6C.

It is fairly easy to construct a PWM control with an IC and a big power MOSFET, and for precise control that's the best way, but for what you are doing, I think the simple thermostat control works just as well.

Quote:

---

Originally Posted by Scarlet Infidel

This is certainly impressive and came out of nowhere! I commend your efforts and hope it works as well as it has potential too.

---

Thank you. I've been reading your posts in this forum and at overclock.net.
I hope that your prototype ends up working well enough that you will
continue to give thought to and design (and build )TEC chillers.

As far as coming out of nowhere, I just decided to wait until I had
something everyone could see before speaking up. I've never learned
or even had access to a CAD program, so being able to present a detailed design is beyond my ability. Also, I wasn't completely certain
that I would be able to assemble my chiller. Beyond a couple of waterlines,
I've never done any soldering and 12" copper bars are not the easiest
things to solder.

Quote:

---

Originally Posted by Uncle Jimbo

I did the calculation for each 'side' and then combined to get total cooling, my assumption is that there is one PA120 on each side. His total sandwich has a number of interfaces so this is all guesswork until we see some real numbers.

With the big surface area he has, Rt between the cold chamber and the TEC cold side should be insignificant. TEC hot side to the hot chamber is critical but again, with his care about pressure, I assume that is not significant. So we basically have Rt for the hot chamber plus Rt for the PA120 on the water loop. We can say pretty accurately that with big fans, the PA120 will deliver .02 water to ambient, so the big question is the hot side chamber.

If that was a pin grid block, I would expect something under .015, but with the home-made channel, that's probably optimistic. Only some testing will show what he actually has there. My guess of .035 is based on surface area transfer at 2GPM of .4C/W per inch, but it could be half of that and it would not surprise me. So my number is probably worst case, and his total per side WC Rt could be .03 or .025.

The air cooling is a SWAG too - but those look like standard sinks with 80mm fans, which are usually around .3C/W. Rt per side is a straight up parallel so Rt(tot) is the Rt per sink / 4 or .075.

Calculating unequal parallel equivalent resistance is 1/Req = 1/R1 + 1/R2 + ...
so in this case, for my SWAG, 1/Rtot = 1/.055 + 1/.075. For LCD, we can use 1/Rtot = 1/(11*.005) + 1/(15*.005) or .005/Rtot=1/11+1/15 => 15/165 + 11/165 => 26/165 = .1576. Solving for Rtot yields .031 so I SWAGed .035 to account for everything I don't know.

Total Rt to ambient is half of that, or .0175

---

I had thought about a pin grid, but with the tools I had on hand, I just
didn't think I could do a decent job of it. Plus, I had no idea what effect
it would have on coolant flow. After reading some research papers on
the use of copper mesh in tube and shell heat exchangers, I decided
to go this route. I had a somewhat better idea of how the mesh would
affect the flow, and I thought it was within my ability to make use of
the mesh.

Like you said though, only testing will show how well it works.

Quote:

---

Originally Posted by leuler

Thank you. I've been reading your posts in this forum and at overclock.net.
I hope that your prototype ends up working well enough that you will
continue to give thought to and design (and build )TEC chillers.

As far as coming out of nowhere, I just decided to wait until I had
something everyone could see before speaking up. I've never learned
or even had access to a CAD program, so being able to present a detailed design is beyond my ability. Also, I wasn't completely certain
that I would be able to assemble my chiller. Beyond a couple of waterlines,
I've never done any soldering and 12" copper bars are not the easiest
things to solder.

---

I understand, I'm always apprehensive to show off what I'm doing until its more or less done (as I'm often delayed by several months and people get impatient).

For the CAD thing I just used google Sketchup, free and easy to use. The model of my chiller was the second thing I ever did on it and it took me about 2 hours (could do it in way less now).

Once I get my TEC Simulator fixed I think you should give it a go (because im trying to get people interested more than anything!) to give an idea of performances at different voltages.

You can try it now if you want, the power requirements it gives you will generally be up to 30% lower than they should be but it seems to be ok at showing trends (ie if you change the C/W or number of peltiers etc).

Quote:

---

Originally Posted by Uncle Jimbo

littleowl - I wondered about that too. But then I looked up the 12706, and as S-I says, it is a 40mm so 1.57 inches. Close but should be OK with careful placement.

But I wonder why leuler chose those TECs. Cost does not seem to be much of an issue for him. 12709's are 10 for $50 on eBay and 12710's are about $6 each. Those are also 40mm TECs.

Just shifting to the 12710s changes the heat transfer a lot, basically moves 1.7 times the heat, but at 1.7 times the power. He could move a total of 1000W!!

With power control on the TECs, I think the 12710s would be a better choice, and it should be easy for leuler to swap them out. But he can do that when the design is checked out, or any time really.

---

As far as the copper bars, it was a case of buying when I saw a good deal
on eBay but before I had a fully coherent plan. Since money does matter
some in this project and the bars are marginally wide enough, I decided
to go with them.

Why did I choose the 12706s? My original plans were to build the most
powerful chiller with the constraint of staying under 600 watts of power.
Most home computers in the US are plugged into a 15 amp circuit. I wanted
the chiller to be such that it could also plug into that circuit and not trip
the circuit breaker ( I am an electrician and that sort of thought came
to my mind ). Ten 12706s fit that scheme. However, at the last minute
I decided to add 4 more TECs in order to be sure that I could handle
whatever amount of heat my QX6700 could dish out.

Also, I didn't want to spend too much money at this point without having
any testing. I could have used larger, undervolted TECs and wider copper
bars, but that would have bumped the costs up a fair amount. Every thing
I have used in this project I either had, bought off of eBay, or bought at
Home Depot. Except for one thing. Instead of getting two MCR320s for
the radiators, I decided to splurge for a PA120.3 to match the one I had.
So, to be honest, I didn't stick to my plan.

I do have a grandiose plan. If this version of the chiller works as planned,
I am going to move to more powerful TECs. :hehe::hehe::hehe:

I think it will work very well! the only problem is the amount of electric being used VS what is being cooled.

Quote:

---

Originally Posted by Scarlet Infidel

Woah, I would never run those TECs at 12v! You'll be using 750w+ of power just to run it. Going down to 8-10v might lose you 5-10c but you'll easily half your power usage if not third it.

Out of interest, so I can run some numbers (I know some of this has been mentioned but I'd like to be clear):

What heat load do you intend to cool?
What water temperatures are you aiming for on the cold side?
What do you estimate as your hot (whole cooling system) and cold side (just between the blocks and the water) C/W?

---

I agree that 750+ watts is crazy to run full time. That's why a power controller would make so much sense. Especially if it could be adjusted
to provide as much cooling power as needed at the time. My DFI
motherboard allows to choose from 4 different BIOS settings at start up.
If I was just doing some surfing, I can choose the non-overclock settings,
and if I want to benchmark, I can choose my highest settings. A power
controller that could adjust to the cooling needs of the processor would
be great.

My heat load? Well, I found conflicting information on how many watts
an overclocked QX6700 produces. I do know that it depends on how
much Vcore is needed to get a stable overclock. All I can say is that the
load maybe between 250 watts and 300 watts. Later on, I will use an E8400
and also cool the northbridge and mosfets with the chiller ( should be
around the same total load).

My water temps? At lower loads, I should be below 0 deg C, as you know.
At the highest loads, I just don't really know, to be honest. There are so
many variables ( heat transfer reduces with lower temps ) that I can
only be confident that it will be sub-ambient.

I will give you what my highest hopes are for temps ( which somebody will
point out that I have forgot something and the temps can't possibly
get that low). Water on the coldside at 0 deg C. CPU temp of 10 to 15 deg.
This is at an ambient temperature of 20 deg C. :shrug:

I will do some more research and calculating to give you a more definite
estimate.

Quote:

---

Originally Posted by Uncle Jimbo

I have also used server supplies for this work - ProLiant DL380 supplies are 12V at 32A, and you can also buy the redundant power sharing board to make it plug-in. I have gotten those for as low as $15 each.

The risk of failure in doing power control of TECs is from thermal cycling. Since you have both the hot side and cold side on water, they won't have much risk, because the heat won't change quickly between power on and power off.

That means you can use a bang-bang controller. A simple thermostat actually does a good job. I used a $10 mercury thermostat on one test. Because it was designed to work in air, I put it in an insulated box mounted on a piece of copper that was stuck into the reservoir. It didn't really matter that the temp switched a little above the 'real' temperature, as long as it was consistent I could put it where I wanted. I had to bend the sensor a little to get it to read down where I wanted since it was designed for 18C to 40C and I wanted to control around 5C.

When the water temp rose above the setpoint, the thermostat triggered a 12V relay which turned on the TEC power, and when the water got cold enough, it turned off. There is a degree or two of hysterisis in that kind of thermostat, so it doesn't turn on and off very fast. In one setup I tested, the whole body of water in the cold loop was about 2.5G and the cooling power was 250W for a 140W heat load. Once water got to 5C, the relay opened, and when it heated up, it closed. It was on for 2 min and off for 30 sec under load, and at idle it was about 1 min on and 1 min off. Water was controlled to between 4.5C and 6C.

It is fairly easy to construct a PWM control with an IC and a big power MOSFET, and for precise control that's the best way, but for what you are doing, I think the simple thermostat control works just as well.

---

If this works well, I'll ask you for a design and the part numbers :D

Quote:

---

Originally Posted by Scarlet Infidel

I understand, I'm always apprehensive to show off what I'm doing until its more or less done (as I'm often delayed by several months and people get impatient).

---

I've seen quite a bit of this on the forums.

Quote:

---

For the CAD thing I just used google Sketchup, free and easy to use. The model of my chiller was the second thing I ever did on it and it took me about 2 hours (could do it in way less now).

Once I get my TEC Simulator fixed I think you should give it a go (because im trying to get people interested more than anything!) to give an idea of performances at different voltages.

You can try it now if you want, the power requirements it gives you will generally be up to 30% lower than they should be but it seems to be ok at showing trends (ie if you change the C/W or number of peltiers etc).

---

Thanks for the tip and I will try out the calculator.

Quote:

---

Originally Posted by littleowl

I think it will work very well! the only problem is the amount of electric being used VS what is being cooled.

---

I agree. I forget what the COP is but I think it is going to be around 0.5.
It will be wasteful to use 24/7.

Quote:

---

Originally Posted by leuler

As far as the copper bars, it was a case of buying when I saw a good deal

---

Your bars will be fine as long as you stay with 40mm TECs and place them carefully.

Quote:

---

Originally Posted by leuler

Why did I choose the 12706s? My original plans were to build the most powerful chiller with the constraint of staying under 600 watts of power.

---

You will use less power per watt of cooling with bigger TECs, but what you have will do a nice job. If efficiency is an issue, stick with voltage control to stay at or below 6V per TEC most of the time - efficiency is more than twice as good as at 12V. With the TECs you have, total draw at 6V is less than 200W for 340W cooling.

Quote:

---

Originally Posted by leuler

Ten 12706s fit that scheme. However, at the last minute I decided to add 4 more TECs in order to be sure that I could handle
whatever amount of heat my QX6700 could dish out.

---

I don't have detail on QX6700 OC power, but it could be over 300W depending on the voltage and frequency you pick. Also you need good cooling for a big OC. I think what you have will work fine, but I've got a couple of suggestions for a 'turbocharger' setup that will help give power when you need it and preserve efficiency when you don't.

Quote:

---

Originally Posted by leuler

I do have a grandiose plan. If this version of the chiller works as planned,
I am going to move to more powerful TECs. :hehe::hehe::hehe:

---

To maintain the best efficiency, you want to use PCM control driven by the cold side. That way you will only apply the voltage you need to maintain your cooling. If the resulting efficiency is not good enough, you can go to bigger TECs.

I'll have to dust off some of my PCM control designs and see if the components are easily available. What I have used is based on a thermally controlled PCM fan IC, driving a big power MOSFET. The output goes into a low pass filter so the TEC sees a fairly smooth voltage. The circuits use only a few components and are easy to build. The MOSFETs don't need much heat sinking at these power levels, so the whole thing is compact and can be mounted anywhere convenient. If you have an old blown up power supply you probably have almost everything you need.

You should be able to achieve better than 1.0 CoP even with fairly big dT.

Quote:

---

I will give you what my highest hopes are for temps ( which somebody will
point out that I have forgot something and the temps can't possibly
get that low). Water on the coldside at 0 deg C. CPU temp of 10 to 15 deg.
This is at an ambient temperature of 20 deg C. :shrug:

I will do some more research and calculating to give you a more definite
estimate.

---

According to Swiftech's site, if I manage to get the coolant on the coldside
to be 0 deg C and I use an Apogee GTX ( currently the best block I own )
on my QX6700 and the load is 300 watts, the CPU temp will be around
30 deg C. Not what I am hoping for.

But, the cpu may get to higher speeds at lower voltage with lower temps.
So, I might get the QX6700 a fair amount over 4GHz without approaching
300 watts.

Still, I'll need a block that is better than .1 C/w to get below ambient
at highest loads.

This type of threads are wonderful.. thx to OP and welcome to XS !!!

uncle J.. you knowledge makes me get lost after reading the first line of you comments

will go to sleep as I have German classes tomorrow ( my wanna be 3rd lang :P)

Quote:

---

Originally Posted by Uncle Jimbo

I'll have to dust off some of my PCM control designs and see if the components are easily available. What I have used is based on a thermally controlled PCM fan IC, driving a big power MOSFET. The output goes into a low pass filter so the TEC sees a fairly smooth voltage. The circuits use only a few components and are easy to build. The MOSFETs don't need much heat sinking at these power levels, so the whole thing is compact and can be mounted anywhere convenient. If you have an old blown up power supply you probably have almost everything you need.

You should be able to achieve better than 1.0 CoP even with fairly big dT.

---

I think I can follow any design that you show me. And I have more of those
server PSUs ( got them really cheap ). I'll probably will need help in figuring
out which parts I'll need to rob from the PSU. Outside of resistors,
capacitors, and inductors, I get lost. What knowledge I have about
electronics comes from reading and E/M physics classes I took almost
20 years ago. The only bit of actual experience I have is that I once
replaced a failed resistor on a stereo circuit board.

Quote:

---

Originally Posted by leuler

I think I can follow any design that you show me.

---

OK - here's a simple design, and I'll give you some guidelines to make it work for what you are doing.

This is a simple fan controller using an 8 pin IC, the MIC520.
http://www.chipcatalog.com/Micrel/MIC502.htm
This is a low-frequency (30 Hz as shown) PWM device with the speed set by C1 (100n is .1 uF). You can breadboard this without much trouble. R1 is a 10K thermistor. C2 in the diagram cleans up the 12V so that the IC doesn't see other system noise - it really should go right next to the IC. The IRF530 is a small MOSFET which will do fine on a fan, but to run the TECs you need something bigger. A LOT bigger.

IR makes a HexFet N-Channel which is rated at 177A at 25C, 125A at 100C. It is in a TO-247AC package which can dissipate 300W. The spec is https://ec.irf.com/v6/en/US/adirect/...uctID=IRFP2907

Put that on a good sized heat sink (or strap it to your hpt side plate somewhere) in place of the IRF530 and you are basically there.

The last step is to improve the low pass filtering. Almost any old power supply has a big toroidal choke or two. They are just some big wire wrapped around a ferrite donut and are usually an inch or two in size. You don't want the ones with lots of different windings - just a simple single layer of 12GA wire or something like that. Almost any size will do. Usually the big power toroid has two windings, each going halfway around the donut. Just use one side.

That toroid is being used as an inductor, and goes between the high side of Q1 and the load. On the other side of the choke, you want a big cap tied to ground, basically putting the + lead on the - fan in the drawing. This will produce a second order filter. The values are not critical, especially since we probably have no idea what the toroid impedance is.

A 4700 uF 35V cap should do the job (Radio Shack about $6) - but if there is significant AC on the output when you test it, just go to a bigger cap or add more in parallel.

With an inductor, you also need to 'snub' transients. Put a 6A 50V diode ($2 at Radio shack I think) between ground and the high side of the MOSFET, with the 'line' on the diode attached to the high side. That will keep the inductor from blowing up your MOSFET.

The system turns full on for 2 seconds at power-up, then adjusts to the thermistor value. You adjust VR1 to set the desired level for a given temperature of thermistor R1. With the component values shown (thermistor with 10K at 25C), and VR1 at the bottom of the range, the circuit will be off at 25C, and all the way at the top, it will be 100% power at 25C. The 'off' level is .4 of the supply, and 100% is .7 of supply.

But you probably don't want a 25C design center. You want a thermistor with 10K resistance at 0C. At that temperature, a Sensor Scientific KW302C thermistor has 10K resistance. Let's say you want to maintain water at 1C. You adjust VR1 to get whatever level of power gets you to 1C, at the load you care about.

So lets assume that your load is 300W, and you are willing to wait 10 minutes or so to let things stabilize, tweaking the control a bit. You will probably end up with about 7V on the TECs to get that cooling, based on our other assumptions, which puts VR1 near the middle of its range - 20K below, 15K above. That's just a guess, but OK for this example.

If for whatever reason the cold side starts to rise, R1 value will start to decrease, and more juice will go to the TECs.

The 'slope' of the thermistor determines the sensitivity to temperature change. R1 is a negative temperature coefficient (NTC) thermistor. So when temp gets to 5C, it has 8.4K resistance. That takes the TECs to 8V. At 10C, they are at 9.2V, 15C at 10.5V, and at 20C you are at the full 12V.

Clearly, the performance of your setup and the load also determine the way this works. If your cooling is better than we think, then all the currents will be lower. But let's see what happens if the load goes down - your hard calculation ends and the load drops to 140W.

Now the temp will start to drop in the cold loop. at -5C, you are at 4V. But you will never get there, because the TEC dT combined with even a modest drop in voltage will drop the cooling power to 140W with a shift of even a few degrees.

With this setup, and your current TECs, you would use only about 250W to maintain 0C with a 300W load. If you set the target at 5C or so, you would use less than 200W to do it. And if you go out of OC mode and back to normal, the TECs will drop to almost nothing - maybe 100W in to cool 140W. So you could run the setup 24x7 and it would adjust itself to your cooling needs without any changes.

If the system has been off for a while, and everything is at room temp, the TECs will go full on and stay there for a while. But even with a 300W load, you are pulling .5C per second out at full power, so in a minute or two, they will start to power back, and you should stabilize in 5 minutes or less at your target temperature.

Quote:

---

Originally Posted by leuler

According to Swiftech's site, if I manage to get the coolant on the coldside
to be 0 deg C and I use an Apogee GTX ( currently the best block I own )
on my QX6700 and the load is 300 watts, the CPU temp will be around
30 deg C. Not what I am hoping for.

But, the cpu may get to higher speeds at lower voltage with lower temps.
So, I might get the QX6700 a fair amount over 4GHz without approaching
300 watts.

Still, I'll need a block that is better than .1 C/w to get below ambient
at highest loads.

---

You have plenty of cooling power. You could go sub-zero on the cold side. A really good block at 2.5 GPM can do .06 or better. Going to -10C would get your CPU down around 8C with that combo.:yepp:

Or you could continue in crazy Xtreme and put a 437 TEC between the CPU and the block. That kind of defeats the efficiency of the rest of your design, since everything would be full bore all the time, but you could drop 10C or so more. Not recommended :down:

This is an awesome setup and great work! I just keep wondering what will be done about condensation? Seems like if you're running 0c liquid through that cpu block there will be plenty of condensate all over the place. How are you getting around this small problem design wise?

Quote:

---

Originally Posted by Vinas

This is an awesome setup and great work! I just keep wondering what will be done about condensation? Seems like if you're running 0c liquid through that cpu block there will be plenty of condensate all over the place. How are you getting around this small problem design wise?

---

Vinas - for sure, for any of the temps we have been talking about, Leuler will need to do the whole seal it up thing. Insulating the tubing to the block will be a must also. When he gets that far we can point to the good references on this forum and elsewhere... Keeping the coolant from freezing is another thing to think about if he really wants those temps at the CPU.

Quote:

---

Originally Posted by Uncle Jimbo

You have plenty of cooling power. You could go sub-zero on the cold side. A really good block at 2.5 GPM can do .06 or better. Going to -10C would get your CPU down around 8C with that combo.:yepp:

Or you could continue in crazy Xtreme and put a 437 TEC between the CPU and the block. That kind of defeats the efficiency of the rest of your design, since everything would be full bore all the time, but you could drop 10C or so more. Not recommended :down:

---

The whole project seems crazy Xtreme to everybody already :D

I don't think a 437 TEC would add much to an overclock attempt. I've read
about people using LN on QX6700s and not getting more than 4.5 to 4.7 GHz.

I do have to admit that I started thinking about a chiller when I was trying
to figure out a way to keep the hotside of a big TEC cool. But it is not worth
adding a big TEC at the cpu, power consumption wise. I wouldn't even get
the pleasure of seeing how fast I could make my house meter turn - the
power company changed it to a digital meter last year.

Quote:

---

Originally Posted by Vinas

This is an awesome setup and great work! I just keep wondering what will be done about condensation? Seems like if you're running 0c liquid through that cpu block there will be plenty of condensate all over the place. How are you getting around this small problem design wise?

---

Thank you Vinas. I'm going to treat this like it is a phase change set up.
Everything on the coldside loop will be insulated, including the water block
and cpu, as well as using dielectric grease in the socket and some conformal
spray on the motherboard. Besides protecting against condensation, it will
help against losing cooling power to the air.

It should be a fancy idea to use Cascade TECs:

CPU Surface>WATER LOOP>(Coldest Side)TEC-80Watts>TEC-160Watts(Hottest Side)>WATER+TEC Cooler

CPU Surface>WATER LOOP>(Coldest Side)TEC-80Watts>TEC-160Watts(Hottest Side)>Fan+TEC Cooler

Anyway it's not easy to have some subzero stuff with TECs :confused:

Thank you for the design Uncle Jimbo :up: :clap:

I do understand how it functions.

I'm thinking about making a test run with the chiller this Saturday, provided
that I have the time. I need to see if it will work as well as we have
speculated. If not, then the primary culprit would probably be my application
of the Arctic Ceramique and I will have to unbolt the chiller and reapply.

If the chiller does live up to expectations, then combining it with the
power controller would make it into somewhat of a Coolit system on
steriods. That would exceed my expectations for this project.

Good luck buddy...

http://www.melcor.com/msseries.html

I think I worked out some numbers for all sorts of combinations of stacking/cascading peltiers. You would only ever do it to achieve higher dT than a single layer can provide.

I'm pretty sure you are always better off just raising the voltage of a single layer towards its maximum to achieve the higher dT. Only once you reach this maximum should you try more than one layer (I can't see why it would ever be worth it for our applications).

Edit: On second thoughts maybe there are situations where this is viable (just not how I'd use them). I'll get back on this tomorrow with an example if I can think of one.

Quote:

---

Originally Posted by Scarlet Infidel

I think I worked out some numbers for all sorts of combinations of stacking/cascading peltiers. You would only ever do it to achieve higher dT than a single layer can provide.

I'm pretty sure you are always better off just raising the voltage of a single layer towards its maximum to achieve the higher dT. Only once you reach this maximum should you try more than one layer (I can't see why it would ever be worth it for our applications).

Edit: On second thoughts maybe there are situations where this is viable (just not how I'd use them). I'll get back on this tomorrow with an example if I can think of one.

---

Cascades or multistage peltiers are used in situations where precisely
controlled low temperatures are needed and the heat load is low. Situations
that call for temperatures lower than a single peltier can achieve.
One situation I can think of is the use of a photomultplier, which can
detect a low flux of photons. I know that it is used in physics and astronomy.
It needs to be kept very cool because thermal noise can be misinterpreted
as a photon striking the sensor ( the speckles you see when using a digital
camera in very low light situations is caused by thermal noise ). Also,
cascades are used when cooling small samples ( biology, chemistry,etc)
to, let's say, -30 to -50 deg C.

For PCs, I can't think of a low heat producing part that needs to be cooled
to that low of a temp.

Quote:

---

Originally Posted by Scarlet Infidel

I think I worked out some numbers for all sorts of combinations of stacking/cascading peltiers. You would only ever do it to achieve higher dT than a single layer can provide.

I'm pretty sure you are always better off just raising the voltage of a single layer towards its maximum to achieve the higher dT. Only once you reach this maximum should you try more than one layer (I can't see why it would ever be worth it for our applications).

Edit: On second thoughts maybe there are situations where this is viable (just not how I'd use them). I'll get back on this tomorrow with an example if I can think of one.

---

As you say, it's done for very small TECs to create big dT - usually for instrument control. The problem with moving any power is that the second TEC in the stack is at a bad operating point because it has to move the heat load of the first, so the CoP of the whole stack is not good. But done carefully it can have some benefits.

Let's say you want to move 20W heat load and have the cold side at -50C in a 25C ambient. No TEC in the world will do that in one stage. so for the first stage, target a 40C dT, and we need to have Th at -10C. Max CoP on the 40C CoP curve is .25 or so at .625 of Imax, and pretty flat through 80% Imax. Using a 12715, we need 10A to move 20W at that Th. The 12715 needs 9V to get there, so total power on the hot side is 90W power in + 20W load heat for a total of 110W. If we put a 19933 on top of that, we need to move that load heat plus whatever power we put in. If we assume enough cooling to keep the hot side at 30C, then we are at the 50% Imax point, 16.5A at 12V for power in of 200W and total heat load of 310W.

That's obviously not real efficient, but if what you want is a -50C cold side, and don't need to move a lot of heat, that would get you there.

There are a number of three stage modules for precise temperature control at low temperatures, but they don't move much power. One I saw had power control of each stage to maintain the interface temp, and could control the load temp to .1C at -80C from 25C ambient. But the heat moved was less than a watt for 300W in.

Yeah, I know those kinds of uses, my dad uses multistage peltiers for cooling CCD chips (low heat output and require very low temperatures).

I think we could find some uses for this in PC cooling, I'd give an example but I find the calculations with existing software to be a complete pain. I could do it on my software in seconds, but then my software doesn't give the right numbers yet.

I tried out the chiller over the weekend. I'll try tonight to post some pics
and go into some detail, but for now I'll just give some numbers.


20 watt load ( pump only )

Ambient - approx 29 deg C
Hotside Loop - approx 35 deg C
Cold side Loop - approx - 6.5 deg C

85 watt load ( pump and E8400 at stock and full load)

Ambient - approx 29 deg C
Hotside Loop - approx 35 deg C
Coldside Loop - approx -3 deg C

140 watt load ( pump and E8400 at 4.0 GHz and 1.35 V )

Ambient - approx 29 deg C
Hotside Loop - approx 35 degree C
Coldside Loop - approx -1.5 degree C

These numbers are from memory and are not quite right but not to far off.
Anyway, the Air/Hotside delta stayed at about 6 deg C.

More later.

I used a Gigabyte DS3 motherboard and E8400 processor for the testing.
My DFI motherboard isn't working, and the Gigabyte motherboard can't
handle the load of a quad at high frequencies. I used Prime95 version 25.5
running small FFTs for an hour for each set of temperatures.


http://img129.imagevenue.com/loc1166...122_1166lo.JPG


With no load except for the heat dump from the pump ( approx 20 watts ),
I got these temps:

AMBIENT (29.4 deg C)
http://img108.imagevenue.com/loc1008...122_1008lo.JPG

HOTSIDE WATER (34.6 deg C)
http://img128.imagevenue.com/loc1066...122_1066lo.JPG

COLDSIDE WATER (-6.4 deg C)
http://img151.imagevenue.com/loc1009...122_1009lo.JPG


Then I tried it out with the E8400 at 1.15 volts and 3.0 GHz. I can't find
a consensus on the heat produced by the processor. I've seen a range
of 30 to 65 watts. So, the total heat load at these settings ( counting
pump) is 50 to 85 watts.

AMBIENT (29.5 deg C)
http://img43.imagevenue.com/loc879/t..._122_879lo.JPG

HOTSIDE WATER (35.2 deg C)
http://img186.imagevenue.com/loc170/..._122_170lo.JPG

COLDSIDE WATER (-4.7 deg C)
http://img180.imagevenue.com/loc120/..._122_120lo.JPG


Next, I overclocked to 4.0 GHz and 1.35 volts ( 100 to 140 watts load)

http://img186.imagevenue.com/loc341/..._122_341lo.JPG

AMBIENT (30.6 deg C)
http://img128.imagevenue.com/loc904/..._122_904lo.JPG

HOTSIDE WATER (35.9 deg C)
http://img129.imagevenue.com/loc833/..._122_833lo.JPG

COLDSIDE WATER (-1.6 deg C)
http://img185.imagevenue.com/loc544/..._122_544lo.JPG


I ran one more test this afternoon ( no pics for this one). I'm unfamiliar
with the Gigabyte motherboard, so the highest overclock I've gotten
so far is 4.32 GHz at 1.45 volts. It was prime stable for 7 minutes, and these
are the temps I recorded:

Load - 140 to 160 watts
Ambient - 31.7 deg C
Hotside - 37.4 deg C
Coldside - 0.9 deg C

I tried to use RealTemp and CoreTemp to record the CPU temp, but
neither program could read the temp at 3.0 GHz ( the temp would not change
between idle and full load ), and I don't trust what they reported at
4.0 GHz. At 4.32 GHz, CoreTemp showed 32 C and 27 C, and RealTemp
showed 23 C and 17 C ( the waterblock mount is definitely not good,
but that doesn't matter at this time - this is a temporary set up).

Aren't you worried that the shelf you have that set up on is too wimpy?

Those temps look good!

Those are some nice temps. It looks like you have an effective Rt of .06 or better which is pretty good all by itself - except you also have a -40C dT from ambient in the mix. Really nice work.

I don't know if you saw this slightly cleaner write-up of the pwm controller -

Attachment 80058

what voltage were you running on your TECs?

Quote:

---

Originally Posted by leuler

Thank you for the design Uncle Jimbo :up: :clap:

If the chiller does live up to expectations, then combining it with the
power controller would make it into somewhat of a Coolit system on
steriods. That would exceed my expectations for this project.

---

The TC648VPA from Microchip Technology is almost identical to the MIC502, except it is a 5V max Vdd and has slightly different cap values (also the second input on pin 5 is not there). It is a little cheaper than the MIC502. C1 is a 1uF for 30Hz, Z1 is 4.7V, R1 is 680 Ohm. If anyone is interested I could do some testing and give actual circuit values.

Also someone suggested that with either IC, use a .1 uF from pin 1 to ground, to dampen any noise - in some cases people have heard humming or other audible noise without that cap.

Both parts are available in the US at http://www.cavalierstock.com - other places too I'm sure.

I'm worried about the condensation on the CPU block because it doesn't look insulated... If the board isn't prepared for sub ambient temps I am afraid you'll find water in the CPU socket, or worse just a bunch of corrosion on the CPU pins resulting in a dead board. My last chill water setup had the later, unfortunately. I had insulation on the CPU block but no dielectric grease in the socket -- I learned the hard way. =\ After that setup I made a program to throttle the temp to 18C by controlling the COM port. Just a simple relay off the COM port to turn on 120vac to the compressor. My program would read the coretemp log and turn on or off the temp according to that. Fun stuff.
http://i279.photobucket.com/albums/k...oto/cpu-pp.gif

I'm excited about this rig so please don't think I'm trying to be pessimistic -- just excited to see it work well for you! Thanks for sharing the pictures, post more soon! =)

Quote:

---

Originally Posted by Clue69Less

Aren't you worried that the shelf you have that set up on is too wimpy?

Those temps look good!

---

Heck No!!! There's too much crap underneath the shelf for it to collapse :rofl:

I have to do all my work on the chiller at my warehouse because the room
that I would use at my house is being used as an extra large closet at the
moment :shakes:

I am pretty happy with the temps, and I think they can be made a touch
better.

Quote:

---

Originally Posted by leuler

I tried to use RealTemp and CoreTemp to record the CPU temp, but
neither program could read the temp at 3.0 GHz ( the temp would not change
between idle and full load ), and I don't trust what they reported at
4.0 GHz. At 4.32 GHz, CoreTemp showed 32 C and 27 C, and RealTemp
showed 23 C and 17 C ( the waterblock mount is definitely not good,
but that doesn't matter at this time - this is a temporary set up).

---

Have you tried Smart Guardian for CPU temperature? It was the only program I found that showed a difference between idle and load temps for the e8400 below freezing. CPU temps need to be converted from 255C = -1C, 254C = -2C and so on. Also, you might need to adjust the cpu temp in the BIOS.


In any case, great water temps! :up: I look forward to seeing the permanent setup.

Quote:

---

Originally Posted by Uncle Jimbo

Those are some nice temps. It looks like you have an effective Rt of .06 or better which is pretty good all by itself - except you also have a -40C dT from ambient in the mix. Really nice work.

I don't know if you saw this slightly cleaner write-up of the pwm controller -



what voltage were you running on your TECs?

---

I did see a revised drawing, but the added notes on the drawing are a
welcomed addition.

I am ashamed to admit, though I had access to at least a half a dozen
multimeters, I forgot to use one to see how much voltage the server PSUs
were actually putting out. I was using the 12 volt rails.

Quote:

---

Originally Posted by Vinas

I'm worried about the condensation on the CPU block because it doesn't look insulated... If the board isn't prepared for sub ambient temps I am afraid you'll find water in the CPU socket, or worse just a bunch of corrosion on the CPU pins resulting in a dead board. My last chill water setup had the later, unfortunately. I had insulation on the CPU block but no dielectric grease in the socket -- I learned the hard way. =\ After that setup I made a program to throttle the temp to 18C by controlling the COM port. Just a simple relay off the COM port to turn on 120vac to the compressor. My program would read the coretemp log and turn on or off the temp according to that. Fun stuff.
http://i279.photobucket.com/albums/k...oto/cpu-pp.gif

I'm excited about this rig so please don't think I'm trying to be pessimistic -- just excited to see it work well for you! Thanks for sharing the pictures, post more soon! =)

---

It is insulated better than it appears in the picture, and I did use dielectric
grease in the socket. I was careful and actually over did it with the RTV -
all because I wasn't careful enough when I first set this up and killed my
primary board :down: . I hadn't had much time to work on my set up the
previous two weeks and I was eager to test how well the chiller would
work. So, one afternoon I got back a little early to the shop and rushed
to get the water block mounted and insulated so that I could do my testing
last Saturday. Big mistake. I didn't get it completely sealed. To compound
things, I let the chiller run for a while without starting up the computer and
didn't check the motherboard over before starting up. Needless to say,
it wouldn't boot :doh: I don't think water got into the socket. Instead,
I believe water got into circuitry around the northbridge. Hard lessons
are usually the most effective lessons.

Anyway, this is a temporary set up. I'm going to break it back down,
tweak it a little, build Uncle Jimbo's controller, shoehorn all the compenents
into a case I'm going to build, and most importantly, properly insulate
the coldside loop.

By the way, I like your way of controlling your temps. Chillers should be
a versatile cooling method. Keep the temps around ambient for normal use,
and extra cooling power when you want to go for extra speed.

Quote:

---

Originally Posted by scifikg

Have you tried Smart Guardian for CPU temperature? It was the only program I found that showed a difference between idle and load temps for the e8400 below freezing. CPU temps need to be converted from 255C = -1C, 254C = -2C and so on. Also, you might need to adjust the cpu temp in the BIOS.


In any case, great water temps! :up: I look forward to seeing the permanent setup.

---

I actually have Smart Guardian, and I did use it. It wasn't reading the
Gigabyte motherboard correctly - the voltage readings were way off and
the CPU temp was being reported as the PWM temp. However, I did forget
that the temps have to be converted when they are below 0C. The board
came with EasyPro 5, which also reports -1C as 255C and so on.

I can confirm a similar problem with my e8400 and CoreTemp/Everest/RealTemp... The temperatures with these programs get stuck at 30C for CoreTemp and 20C for RealTemp. Smart Guardian reports idle temps at 242C which should translate to -13C. I had to calibrate and offset the BIOS temp using the RealTemp method... After that I undervolted and underclocked my CPU to 1.8GHz just to make certain, and the cores were still stuck on 20C/30C... The temps eventually go up as CPU load gets over 80% but never below 20C/30C. :confused:

I have some power consumption numbers.

Chiller ( PSUs, TECs, fans, and pumps ) - 925 watts at 30C ambient

This is just a little over the max wattage that I had planned for the
chiller. However, the ambient temperature at my house will be around 20C,
so the power consumption will be higher ( at peak loads ) when all is said
and done.

Here is how it breaks down :

Pumps - 104 watts
Fans - 54 watts ( approx)
PSUs and TECs - 767 watts

I forgot to try to see how many watts the PSUs use with no load. Anyway,
the calculator at http://www.ferrotec.com/products/the...tail.php?id=70
shows that ( assuming my Chinese TECs are similiar enough to the Ferrotec
TECs) the TECs are pulling about 4 amps at 35C hotside, for a total of
about 650 watts ( at 11.6 volts ).

Would a PSU consume 55 - 56 watts with no load?

Quote:

---

Originally Posted by leuler

I have some power consumption numbers.

Chiller ( PSUs, TECs, fans, and pumps ) - 925 watts at 30C ambient

This is just a little over the max wattage that I had planned for the
chiller. However, the ambient temperature at my house will be around 20C,
so the power consumption will be higher ( at peak loads ) when all is said
and done.

Here is how it breaks down :

Pumps - 104 watts
Fans - 54 watts ( approx)
PSUs and TECs - 767 watts

I forgot to try to see how many watts the PSUs use with no load. Anyway,
the calculator at http://www.hebeiltd.com.cn/peltier.d...TEC1-12706.pdf
shows that ( assuming my Chinese TECs are similiar enough to the Ferrotec
TECs) the TECs are pulling about 4 amps at 35C hotside, for a total of
about 650 watts ( at 11.6 volts ).

Would a PSU consume 55 - 56 watts with no load?

---

It is amazing how close your actual numbers came to the original design point calculations. If you tell me what gauge wir you are using between the PSU and the TEC, I can tell you how to measure current directly - well pretty close - from the voltage drop in the wire. Based on the performance you are getting, I suspect you are actuallly drawing a little less power than you think.

PSU efficiency varies quite a bit. 90% is pretty good, and if you are looking for the difference between calculated and measured power, then expect the PSU to use 70W or so to deliver your 675W to the TECs.

Just going to the 12710 TEC and using PWM to control the power, and making NO other changes, you get exactly the same cooling power for only 550W in.

The biggest 40mm that is easily available is the Kryotherm Drift-0.6, which is a 19915. Using those units, you will get the same cooling power for about 450W in.

Your power consumption will go down with a lower ambient, assuming the same cold side temps. You are also operating at a lower efficiency point but you also have a smaller dT requirement. If the hot side remains at 6C or so above ambient, and ambient drops to 20C, your current 12706 rig will drive the cold side to 10C lower than it is now. If you drop power to maintain your -1C load temp, you would need only 7.5V in, and total power in is only 260W.

This is really a killer design, and you have a lot of good options to either reduce the power in or get even lower temps.

noticed a problem with the circuit while testing a prototype. Pin 1 (the control) sets 100% duty cycle when the control is at .7 of Vdd. If the value becomes greater than .74 of Vdd, the MIC502 senses this as an overtemp fault, and shuts down the output. That will happen at 25C if the pot is set above 50%. To avoid this problem, put an 8.2V Zener Diode between pin 1 and ground. I will put up a revised schematic in a bit.

Quote:

---

Originally Posted by Uncle Jimbo

It is amazing how close your actual numbers came to the original design point calculations. If you tell me what gauge wir you are using between the PSU and the TEC, I can tell you how to measure current directly - well pretty close - from the voltage drop in the wire. Based on the performance you are getting, I suspect you are actuallly drawing a little less power than you think.

PSU efficiency varies quite a bit. 90% is pretty good, and if you are looking for the difference between calculated and measured power, then expect the PSU to use 70W or so to deliver your 675W to the TECs.

Just going to the 12710 TEC and using PWM to control the power, and making NO other changes, you get exactly the same cooling power for only 550W in.

The biggest 40mm that is easily available is the Kryotherm Drift-0.6, which is a 19915. Using those units, you will get the same cooling power for about 450W in.

Your power consumption will go down with a lower ambient, assuming the same cold side temps. You are also operating at a lower efficiency point but you also have a smaller dT requirement. If the hot side remains at 6C or so above ambient, and ambient drops to 20C, your current 12706 rig will drive the cold side to 10C lower than it is now. If you drop power to maintain your -1C load temp, you would need only 7.5V in, and total power in is only 260W.

This is really a killer design, and you have a lot of good options to either reduce the power in or get even lower temps.

---

Thanks to my math and physics background and the number of hours I spent
researching, I pretty much knew what to do to meet my design parameters.
However, constructing the chiller and it actually coming close to my
calculations surprises me; I thought it would take more than one try to get
this close, given that some of the construction involved doing things that
I have never done before.
--------------------------------------------------------------------------
The wiring is temporary; the final set up will decrease voltage drop ( which
is already pretty low ). But let me describe the PSUs and how everything
is wired :

Two 600w PSUs with four 12v rails each with double 18 awg leads for each rail.
12v rail1 = 18 amps
12v rail2 = 18 amps
12v rail3 = 15 amps
12v rail4 = 12 amps

rail1 + rail2 + rail3 rated at 45 amps total.

Both rail1s are each connected to three TECs. The 18 awg leads are approx
12" long. The 16 awg leads from the TECs are approx 3" long.

Ditto for the rail2s.

The rail3s are connected to one TEC and three heatsink fans each ( 0.3 amps
each ). Lead lengths are the same as above.

The rail4s are connected to three San Ace fans ( 0.48 amps each ). The
18 awg leads are approx 18" long.
--------------------------------------------------------------------------
I really believe that this design can be improved. PWM power control alone
will make this chiller much more versatile, and as you pointed out, allow
for more cooling power ( when needed ) and more efficiency ( for every day
use ) to be built in. Also, I think I can improve heat transfer efficiency.
Maybe not with this chiller ( though I've got some ideas ), but definitely
with my next one ( sometime down the road ).

Quote:

---

Originally Posted by Uncle Jimbo

noticed a problem with the circuit while testing a prototype. Pin 1 (the control) sets 100% duty cycle when the control is at .7 of Vdd. If the value becomes greater than .74 of Vdd, the MIC502 senses this as an overtemp fault, and shuts down the output. That will happen at 25C if the pot is set above 50%. To avoid this problem, put an 8.2V Zener Diode between pin 1 and ground. I will put up a revised schematic in a bit.

---

here's a revised schematic that fixes the problem

Attachment 80651

:up: Thank you very much.

Quote:

---

Originally Posted by leuler

:up: Thank you very much.

---

This is not actually my original suggestion - this is a better design because it also puts the thermistor at a constant voltage which gives tighter temp control. It also allows pretty much any supply voltage - I have tested it to 30V.

Leuler ! where do you buy copper ? How much is that ?

Quote:

---

Originally Posted by boy_gameover

Leuler ! where do you buy copper ? How much is that ?

---

I bought the copper bars on eBay. Counting shipping, I paid $85 to $90
for it.

There are at least three suppliers online in the US that sell small amounts
of copper for hobbyist. If I bought the copper from them, it would have cost
$150 to $180, counting shipping.

I can't seem to find a thermistor probe that will have a resistance of
10k ohms when the water temp is 0 C to 10 C. All that I can find are
at least 10k at 25 C ; I need one that has a resistance of 3.5k to 5k
at 25 C ( assuming Rt/Rt(25 C) is around 2 at 10 C ).

Jimbo, what could I expect from my chiller using your controller design
and a 10k (25 C) thermistor probe? Could I use two 10k thermistor probes
in parallel?

By the way, I've ordered all of the parts for the controller, minus the
thermistor probe. And I've designed my enclosure for the chiller, rads,
pumps, and PSUs. So, in the next couple of months I should have a
water chiller that looks and performs halfway decent.

killer project luelur :up: well done

Quote:

---

Originally Posted by leuler

I can't seem to find a thermistor probe that will have a resistance of
10k ohms when the water temp is 0 C to 10 C. All that I can find are
at least 10k at 25 C ; I need one that has a resistance of 3.5k to 5k
at 25 C ( assuming Rt/Rt(25 C) is around 2 at 10 C ).

Jimbo, what could I expect from my chiller using your controller design
and a 10k (25 C) thermistor probe? Could I use two 10k thermistor probes
in parallel?

---

You just need to change the other resistor values to match. A NTC probe that has 10K at 25C will have a much higher value at 0C - if you don't have the table, the easiest thing is just put it in an ice water bath and measure the resistance. Then adjust the other resistors in the bridge to match.

So let's say you measure 30K at 0C. Then use something close to that for VR1, and about 1.5 times the value for R2. They don't have to be exact - if you use 27K for VR1 and 47K for R2 that will work fine.

Quote:

---

Originally Posted by Uncle Jimbo

You just need to change the other resistor values to match. A NTC probe that has 10K at 25C will have a much higher value at 0C - if you don't have the table, the easiest thing is just put it in an ice water bath and measure the resistance. Then adjust the other resistors in the bridge to match.

So let's say you measure 30K at 0C. Then use something close to that for VR1, and about 1.5 times the value for R2. They don't have to be exact - if you use 27K for VR1 and 47K for R2 that will work fine.

---

Ah, so it is the relative resistances of the probe, VR1, and R2 that matter,
not the total resistance of that branch.

Okay, I'll get the sample thermistor probe from Ametherm and also get a
trimpot and resistor to match with it.

Thank you Jimbo :up:

Well, here is my version of the controller, just lacking the thermistor probe.
My very first time constructing an electronic circuit, not counting working
on a breadboard in college.

I made the positive and negative rails by splitting in half a flat copper bus
from a meter base. I thought that it was a good idea until I tried soldering
to them. That's a lot of copper to heat up, and I'm inexperienced at soldering
anyway. So that explains the scorch marks.

Anyway, if I luck, the controller will work. If not, I will blame the assemblier,
not the designer.

I haven't finished the controller nor have built the case yet.

However, I have been tweaking the chiller and will be testing again this
weekend. I've exchanged the TEC1-12706 pelts for TEC1-12709s. I've
also improved the insulation, upgraded the wiring, and improved the
cooling for the hotside loop ( shrouds and push/pull fans on the rads ).
I'm also cooling the northbridge with a TEC ( direct die with a SanAce
Liquid Cooler to cool the hotside plate ).

I'm not sure if I will have a camera on Saturday to take pics, but I will
report the results.

I'm having some problems to get the right parts shipped to the netherlands. I just cant get them here!

You said it would only cost like $20 of parts.

Is there a SLIGHT possibilty that someone could sent me parts?

I would be very VERY gratefull! :clap:

I love to get my pelts started, with 100% on both pelts the water just freezes, I need to get a grip on these pelts.

The tricky part, is the PWM-module in the middle, its just out of my reach.

That in advance!

Quote:

---

Originally Posted by jens21

I'm having some problems to get the right parts shipped to the netherlands. I just cant get them here!

You said it would only cost like $20 of parts.

Is there a SLIGHT possibilty that someone could sent me parts?

I would be very VERY gratefull! :clap:

I love to get my pelts started, with 100% on both pelts the water just freezes, I need to get a grip on these pelts.

The tricky part, is the PWM-module in the middle, its just out of my reach.

That in advance!

---

See your pm's...