Overclock.net › Forums › Cooling › Specialized Cooling › Peltiers / TEC › Peltier/TEC Hybrid Water Cooling vs 2.0
New Posts  All Forums:Forum Nav:

Peltier/TEC Hybrid Water Cooling vs 2.0 - Page 13

post #121 of 392
Quote:
Originally Posted by toeCUTTER78 View Post

If you look close i also have one evo, the other three are mx.
I didnt want to "destroy" them before i knew this was gonna work, as its not an exact replica of your system.

I am using two 360 rads to cool the hot side. Unfortunatly i found out that my 2nd peltier only drew 100w and now 0w .. So i have a 2nd on order.
The hot side Water barely warms up.

It is correct that the peltier is a little bigger than the blocks. However i thought that "gasket" from the "potted peltier" is a bit withdrawn from the edge and thus the actual cooling / heating surface roughly matches the blocks. I thought of cold / heat plates but questioned their thermal resistance as to further layers of thermal comound asf. I might be wrong. Mind you only been running this week.
Being married with child, sets its limits on the time left for tinkering :-)

The TEC is 62x62 yes, the specs are from eBay, so i doubt it is even a 12730 tec.
I can tell you it draws around 235w at 12v after running for some time. Thats roughly 20 amps.

Hope i got it all answered.

I an building a filtered PWM controller for the one TEC. This will get temp sensor feedback from the coolant. When it reaches a certain point a relay will trigger the 2nd peltier, as when I'm gaming.
I hope this will keep the electrical bill somewhat lower ;-) .. Plus its fun to build.

/Toe

I just notice this your running two 360mm radiators for the hot side of two TEC's. don't get me wrong I am not knocking it. but the water temperatures on the hot side water loop should be around 50C-55C this is normal for that many radiators on that strong of TEC's.
a easy way to drop 10C on the water temperatures would be to add two more 360mm radiators or better yet add two more 480mm radiators to the hot side water loop.

word of caution before doing this kind of upgrade, that would bring your water temperatures into the realm of condensation, a PWM controller for the TEC's would be a good idea to prevent damage to the PC system. thumb.gif
post #122 of 392
@toeCUTTER78
That power draw is massive, and is probably dumping loads of heat for those 2x 360 rads. Do you know what your hot loop coolant temp is?

The build looks great so far, besides the lack of cold and hot plates (which may explain the short lifespan of one of the TECs...they need even cooling across the whole surface). Will be great for a proof-of-concept, but I would look into some better quality TECs once you get it dialed in. For example, a CTE400 at 12v pulls under 13A vs your 20A TECs and can still control ~150w of a heat source while dumping almost half as much into your hot loop. Since TECs work off a delta, by having a lower temp hot side you will have colder coolant temps even if your TECs each are smaller and lower wattage. Alternately, you can just add more rad to handle the heat load and get the hot side as close to ambient as you can. If I had to guess, your hot side loop coolant is probably 30-40c? You got 10+c of potential cooling ability just stuck in there wink.gif

You are currently asking 2x360 rads to move probably over 600w of heat down to as close to ambient as you can (235wx2+unknown CPU), which is a lot for them smile.gif.
Behemoth
(25 items)
 
  
CPUMotherboardGraphicsGraphics
Intel i7 4790K "Relidded" @ 5ghz <50c Asus Sabertooth Z97 MK1 Custom WC'd 7970 3GB @ 1100/1500 stock volts Custom WC'd 7970 3GB @ 1100/1500 stock volts 
RAMHard DriveCoolingCooling
2X4GB Patriot Viper Extreme 1TB Velociraptor 10k RPM w/ 64MB Cache Two Arqtik TEC Chillers in Series Archimedes TEC block w/TN660 TEC 
CoolingCoolingCoolingCooling
Two Ported/tapped MCP655 Pumps 2x Black Ice GTX 480 Gen2 Rads DD dual drive bay Res with LCD water temp monit... GPU Loop: Modded Enzotech Sapphire CPU blocks, ... 
CoolingOSOSMonitor
36x Globe 120mm 80cfm server fans Win 10 Pro Mint 15(KDE) 4K 28" Samsung U590D 
KeyboardPowerCaseMouse
Logitech G710+ Mechanical CM Silent Power M 1000watt Main, OCZ Z850W Slave Custom Mountain Mods Extended-Extended Ascension Logitech G502 Proteus 
Mouse PadAudioAudioAudio
Homemade 16ga 304 Stainless pad, Neoprene backed Aune T1 DAC Various Tubes: Mostly switching from 60's GE "S... AKG K7XX Headphones 
  hide details  
Reply
Behemoth
(25 items)
 
  
CPUMotherboardGraphicsGraphics
Intel i7 4790K "Relidded" @ 5ghz <50c Asus Sabertooth Z97 MK1 Custom WC'd 7970 3GB @ 1100/1500 stock volts Custom WC'd 7970 3GB @ 1100/1500 stock volts 
RAMHard DriveCoolingCooling
2X4GB Patriot Viper Extreme 1TB Velociraptor 10k RPM w/ 64MB Cache Two Arqtik TEC Chillers in Series Archimedes TEC block w/TN660 TEC 
CoolingCoolingCoolingCooling
Two Ported/tapped MCP655 Pumps 2x Black Ice GTX 480 Gen2 Rads DD dual drive bay Res with LCD water temp monit... GPU Loop: Modded Enzotech Sapphire CPU blocks, ... 
CoolingOSOSMonitor
36x Globe 120mm 80cfm server fans Win 10 Pro Mint 15(KDE) 4K 28" Samsung U590D 
KeyboardPowerCaseMouse
Logitech G710+ Mechanical CM Silent Power M 1000watt Main, OCZ Z850W Slave Custom Mountain Mods Extended-Extended Ascension Logitech G502 Proteus 
Mouse PadAudioAudioAudio
Homemade 16ga 304 Stainless pad, Neoprene backed Aune T1 DAC Various Tubes: Mostly switching from 60's GE "S... AKG K7XX Headphones 
  hide details  
Reply
post #123 of 392
CTE400's are cheaper than rads, just saying ;-) 2 x 360mm rads is enough for 2 x CTE400's at 12v + CPU heat load, more rads is always better but 2 x 360mm is fine.

Your water blocks are holding you back toecutter, both in size and internal design. Phanteks Glacier C350i is the most suitable off the shelf CPU block currently available for TEC usage in terms of internal design of the block, the base plate is most likely only suitable for 40mm TEC's though, possibly 50mm, can always add a 62x62mm cold plate or just mill the base plate flat to maximize the usable surface area.

I'd be tempted to build a chiller with 4 x C350i blocks and 2 x CTE225w TEC's. Unfortunately the phanteks water blocks seem to be unreasonably expensive in Australia compared to EK blocks etc.

CTE225 19915 http://customthermoelectric.com/tecs/pdf/19911-5L31-15CQ_spec_sht.pdf

CTE400 19928 http://customthermoelectric.com/tecs/pdf/19911-5M31-28CZ_spec_sht.pdf

C350i internals



C350i inlet/outlet, much better for TEC's than almost every other off the shelf block out there


Edited by LiamG6 - 4/8/17 at 10:22pm
post #124 of 392
Thread Starter 
Quote:
Originally Posted by b0oMeR View Post

The contraptions going on here

I think this thread is going to start looking like a bomb making class...
thumb.gif to NSA analyst that placed this thread on notice.

ROFL It does kinda look like that in some ways. smile.gif
post #125 of 392
Thread Starter 
Quote:
Originally Posted by toeCUTTER78 View Post

Thx for all the tips, they are appreciated, and yes tinkering budget shrinks with family added ;-)

Regarding the screws on EKWB.
I sanded them down to match the cooper base. Then just poured thermal Grease in there :-)


I know you don't want to hear it but you should have lapped those blocks flat as the EVO for sure is bowed outward in the center, which is great for a CPU heat spreader but not for flat peltiers.

Also you still need to cover the uncovered peltier edges with a cold plate.

Quote:
Originally Posted by toeCUTTER78 View Post

Reservoir made up of sewer tube, in a spirotube(used for venting) and then PU foam inside (used to insulated Windows frames and such)

That sounds great should really hold in the cold.

Quote:
Originally Posted by toeCUTTER78 View Post

So the only difference between our systems is basically Water vs air cooling the hot side. Mind you i did initially get a TT TRUE, but found that i had problems fittings two inside my case vs Water loop.

You're right the heat pipe air coolers do take up a lot of space because they have to be mounted vertically to successfully cool peltiers.


Quote:
Originally Posted by toeCUTTER78 View Post

Regarding pump position in the loop, does it matter ? I thought it would reach some sort of temperature equilibrium no matter where in the loop ?

Anyway my loops are

Hot:
Res - pump - rad - rad - TEC - TEC

Cold:
Insulated res - temp probe - pump - CPU - TEC - TEC - flow meter

Temperature equilibrium is basically friendly to a CPU, GPU, water loop, but a peltier constantly generates its heat load and is going to be extremely unfriendly to a water pump, so for longevity of the pump give it the coolest coolant you can and that is after the radiators.

When I experimented using water cooling to cool the hot side I was only using 1 peltier and destroyed a pump, so I suggest having the pump in the loop after the radiators have cooled the coolant, the cold side doesn't matter as the cold doesn't seem to affect longevity of the pump in a below ambient chilled water cooling setup.


Quote:
Originally Posted by toeCUTTER78 View Post

I know Arthur H build one just like Yours Ryan, but who else ?

/Toe

Arthur H was the only one that duplicated the setup exactly as far as the components used, he arranged the components in the loop differently to test if the end results were the same and they were, he ran the same air coolers on the hot side that I did and the same exact peltiers and supplied voltage.

Arthur ran 2 TEC assemblies because at the time that was all I was running, I had discovered that when I ran Intel Burn Test with my 3770K overclock that the coolant temperature increased.

I added a 3rd TEC assembly which was temperature controlled to maintain coolant temperature in a 8.9c ~ 10c window, and after that not only did it hold the coolant temperature with Hyper Threading Enabled, but with Hyper Threading Disabled and IBT running the coolant temperature continued to drop with the 3rd TEC assembly powered.

Arthur H is still running his setup to this day which should put him over 3 years now, but he is also still running a 2700K Sandy Bridge CPU which did not get as hot as the 3770K I was running.

As far as others that have duplicated the exact concept, I don't know, there have been many experimenters over these past 4 years changing various aspects of the idea but I don't know of any that were successful enough to post back their results.

After saying all of that I want to say to you that I am excited for you!, and I hope you achieve what you want to achieve and that your cooling is successful.
Edited by OCAddict - 4/9/17 at 7:03am
post #126 of 392
Quote:
Originally Posted by OCAddict View Post


I don't know, there have been many experimenters over these past 4 years changing various aspects of the idea but I don't know of any that were successful enough to post back their results.

yes, well I am an ********** so I get extremely curious, so I go find there PM address, and start asking questions about what happened. tongue.gif

that's why I refuse to watch pilot shows, if they don't make it into a series, the questions will drive me crazy for years. lmaosmiley.gif
post #127 of 392
Thread Starter 
Quote:
Originally Posted by toolmaker03 View Post

yes, well I am an ********** so I get extremely curious, so I go find there PM address, and start asking questions about what happened. tongue.gif

that's why I refuse to watch pilot shows, if they don't make it into a series, the questions will drive me crazy for years. lmaosmiley.gif

Personally I can only comment on what I know that absolutely 100% positively works!

I have had numerous PMs regarding whether I thought a change here, or a change there would work, and my response is always you will just have to try it and see what happens.

Many do not like getting that answer, but otherwise it is just an assumption.

If we all share what actually does work, then maybe one day some little genius will come up with a major breakthrough in cooling, that is an idea that none of us even thought of.
post #128 of 392
flux capacitance cooling it drains the heat away to another time. lachen.gif
post #129 of 392
Quote:
Originally Posted by OCAddict View Post

Personally I can only comment on what I know that absolutely 100% positively works!

I have had numerous PMs regarding whether I thought a change here, or a change there would work, and my response is always you will just have to try it and see what happens.

Many do not like getting that answer, but otherwise it is just an assumption.

If we all share what actually does work, then maybe one day some little genius will come up with a major breakthrough in cooling, that is an idea that none of us even thought of.

in all reality there seems to be a good group here that are willing to post what worked, and what did not. I have looked over most of the posts made by the regulars over the years. this has been very helpful to me as well as many of the suggestions made. thumb.gif
Edited by toolmaker03 - 4/9/17 at 9:31am
post #130 of 392
Hi guys

Let me start by informing you what it is I am cooling.
It's my old rig consisting of and I7 860 (Lynnfield), currently OC'ed to 4 gHz @1,386v (if i remember correctly)
Currently my coolant temp will continue to drop while idling in Windows, but when playing games or doing prime95 test
it will rise to ambient/above.
Currently HT is off, and energy saving features an enabled.

So to sum up - I will eventually be trying to cool 2 x tec1-12730 using around 250 - 260w each - approx 500w total of heat.
I dont have a clue what wattage my cpu is consuming when OC'ed.


@Tool
Yes, i have 2x360 rads - not monsta models.
Unfortunately i did not think of adding a temp sensor probe for the hot side water, but i might want to retrofit one.
My current hot side cooling temp is 37 C measured on the radiator inlet, just after the tec.
This is with two TEC's enabled - one not as spec'ed tho, and working periodically - see answer to Puck.

@Puck
You are pretty spot on - temp is 37 C as mentioned above.
This is however not with the full wattage I expected as the one TEC was delivered defective, so not result of excess temp.
TEC 1 draws around 260w
TEC 2 (defective) draws around 114w
They are the same China tec1-12730 knock-off model .. totally and utterly unspected rolleyes.gif
The CTE400 sounds interesting, it's 24v tec i see ... the numbers you mention are they from personal experience ?
Unfortunately custom thermo electric is located in the States, and since I live in Denmark, I might have to sell a kidney to import one of those ... then again, I can live with one kidney wink.gif
I think I could fit in another 360 rad without changing my layout completely - Would that be sufficient do you think ?
A 360 rad will cool around 300w of heat, yes ?

@Liam
I'm kindda stuck with the ekwb block now, as I'm not able to fund further major investments redface.gif
I guess I'm gonna have to "make do" with these, and maybe modify them a bit.

@Ryan
Uh oh - no I did not want to hear that regarding the convex copper base biggrin.gif
Somehow I knew blocks were shaped like that, I just didnt occur to me doing all the other stuff.
Oh well, when my new China TEC arrives I'm gonna have to take it all apart anyhow, that will be a good opportunity to sand them down flat and to "lap" them afterwards.
How much are we talking mm -wise in regard to "convex-ness" ?
Regarding the hot/cold plates - I will keep that in mind, but will continue on as now for the moment, as I'm almost certain that the inner of the tec is not have a bigger than the waterblock - it's only the exterior ceramic plates and the "potting / gasket" which is out over the side of the waterblock.
The less thermal resistance the better - so I'd like to avoid extra assemblies ... I might come around at some point, but this is my belief atm wink.gif



Hope i did not forget anything or anyone

Cheers Toe
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Peltiers / TEC
Overclock.net › Forums › Cooling › Specialized Cooling › Peltiers / TEC › Peltier/TEC Hybrid Water Cooling vs 2.0