Overclock.net › Forums › Cooling › Specialized Cooling › Peltiers / TEC › New cooling, some questions I need to ask first
New Posts  All Forums:Forum Nav:

New cooling, some questions I need to ask first - Page 2

post #11 of 19
I do forget that you guys in the US only use 120v AC, wheras we have 240v AC.

Also, regarding series or parallel at 24v, those 2x 331w CTE tecs will still try and draw ~14A and you'll still only end up with a -10C coldside and still be needing a 750W PSU. And at the end of the day, 2x TECs in parallel from a 24V PSU is going to achieve the same as 2x TECs in series from a 48v PSU - so what it really comes down to is cost of the PSU.

I generally suggest 12V and wiring in parallel because an ~85A 1000w PC PSU is almost always cheaper then a 1000W meanwell at the right amperage.

The other benefit of using a PC PSU is that they can be used safely inside a PC case, without any exposed 120v/240v AC terminals that can kill you if you touch them when the meanwell is powered on.

If it was me trying to get to -20C with a 250w heat load, I'd most likely use 4x 245W CTE 50mm TECs and a ~1000W PC PSU with either a single 12v rail rated at 85A or 4x 12v rails rated at 25A and then wire all 4x TECs to the single 12v rail rated at 85A or each TEC to it's own 12v rails if it had 4 12v rails rated at 25A.

I'd also use at least 2x decent copper finned 320mm radiators with 6x fans and ducts on each rad, but I'd almost likely need more then that to keep the 1250w heatload from the 4x TECs hotside under control.

But honestly, I would even try it, because using 1000w of electricity in a daily rig just for cooling makes no sense to me - that's 5kW of electricy if your PC is on for 5 hours a day, or an average of around 152kW of electricity per month and over 450kW of electricity a quarter - which is a HUGE amount of electricity being used.

That's 1/4 the amount of electricity I currently use in my whole house per quarter - just for cooling a PC, which is crazy.

I think the best approach is what Mindchill did a while - use a PWM controller to keep the keep temps at a specific temp. Then you'd keep temps below ambient and above the dew point to stop condensation.

That would use keep C.O.P high and use much less electricity, and be much more practical for a daily rig imo smile.gif
Edited by un-nefer - 3/23/11 at 11:24pm
Crimson Basilisk
(25 items)
 
  
CPUMotherboardGraphicsGraphics
Intel i7 2600K @ 5.1Ghz AsRock Extreme4 Gen3 Palit GTX 680 Jetstream 2GB Palit GTX 680 Jetstream 2GB 
RAMRAMHard DriveHard Drive
G.Skill F3-12800CL9D-8GBXL G.Skill F3-12800CL9D-8GBXL Sandisk Extreme 120GB SSD Sandisk Extreme 120GB SSD 
Hard DriveOptical DriveCoolingCooling
WD Caviar "Black" 1TB SATAII LG WH14NS40 Blu-ray Burner XSPC Rasa CPU (Acetal) Black RX360 Triple Fan Radiator 
CoolingCoolingOSMonitor
XSPC DDC Two BayRes BMaverick Laing DDC (MCP350) Windows 7 Ultimate x64 Philips 201B 21" CRT 
MonitorKeyboardPowerCase
ViewSonic VX922 19" LCD Diamond Digital Black Antec True Power 750w Antec LanBoy Air Red 
MouseMouse PadAudioOther
Epic Gear Meduza (By GEIL) Steel Series QCK+ XL Asus Xonar DX ROCCAT Kave 5.1 Surround Sound Gaming Headset 
Other
Sennheiser hd555 
  hide details  
Reply
Crimson Basilisk
(25 items)
 
  
CPUMotherboardGraphicsGraphics
Intel i7 2600K @ 5.1Ghz AsRock Extreme4 Gen3 Palit GTX 680 Jetstream 2GB Palit GTX 680 Jetstream 2GB 
RAMRAMHard DriveHard Drive
G.Skill F3-12800CL9D-8GBXL G.Skill F3-12800CL9D-8GBXL Sandisk Extreme 120GB SSD Sandisk Extreme 120GB SSD 
Hard DriveOptical DriveCoolingCooling
WD Caviar "Black" 1TB SATAII LG WH14NS40 Blu-ray Burner XSPC Rasa CPU (Acetal) Black RX360 Triple Fan Radiator 
CoolingCoolingOSMonitor
XSPC DDC Two BayRes BMaverick Laing DDC (MCP350) Windows 7 Ultimate x64 Philips 201B 21" CRT 
MonitorKeyboardPowerCase
ViewSonic VX922 19" LCD Diamond Digital Black Antec True Power 750w Antec LanBoy Air Red 
MouseMouse PadAudioOther
Epic Gear Meduza (By GEIL) Steel Series QCK+ XL Asus Xonar DX ROCCAT Kave 5.1 Surround Sound Gaming Headset 
Other
Sennheiser hd555 
  hide details  
Reply
post #12 of 19
Elloquin is right about 67.77 % power being the best point for cooling. At this point using any more power gets very little gain. It has been quoted variosly in the past as 70-75% but Elloquins testing has obviously been more accurate.
While this is the point of best "efficiency" it certainly is not the best point of COP. The best COP is achieved at 35-40% power. This is the best point of efficiency. It will enable you to keep power usage to the minimum but it will require more TECs to get the cooling you require but with careful use of series and parallel connections, and mindful application of mountings and TIMs will get a better system, but it needs careful planning and the knowledge to pull it off.

Crabid wrote"I have a planned custom block being imagineered at the moment, I do realise that there are a huge number of variabes involved in the calculations but at the very least you mut be able to calculate Amps, Volts & produce estimates via a heat ballance."

The point is you dont calculate - you use charts.
TECs are semiconductors.... any regular calcs such as Ohm's law will not be accurate. If you take the trouble to look at the charts and understand them you will see this. There are no calculations for TECs you can do that are regular. Do them if you insist..... you will get an answer ....but it will be wrong. You should insist on getting the charts when you buy TECs or at least look them up on the internet. Obviously charts also have there inherent faults some are small and difficult to read accurately but you will often be far better using them than doing any calculations, and once you learn how to read charts it is much faster BUT you must remember you still only have ballpark figures. The charts are hotside temperature related so unless your hotside is exactly at the temp of the chart rating the results you get wont be the same. If you think about it this is the very reason why your calcs will never be right ! They are not temp related.
Edited by zipdogso - 3/24/11 at 3:07am
post #13 of 19
Thread Starter 
Quote:
Originally Posted by zipdogso View Post
Elloquin is right about 67.77 % power being the best point for cooling. At this point using any more power gets very little gain. It has been quoted variosly in the past as 70-75% but Elloquins testing has obviously been more accurate.
While this is the point of best "efficiency" it certainly is not the best point of COP. The best COP is achieved at 35-40% power. This is the best point of efficiency. It will enable you to keep power usage to the minimum but it will require more TECs to get the cooling you require but with careful use of series and parallel connections, and mindful application of mountings and TIMs will get a better system, but it needs careful planning and the knowledge to pull it off.

Crabid wrote"I have a planned custom block being imagineered at the moment, I do realise that there are a huge number of variabes involved in the calculations but at the very least you mut be able to calculate Amps, Volts & produce estimates via a heat ballance."

The point is you dont calculate - you use charts.
TECs are semiconductors.... any regular calcs such as Ohm's law will not be accurate. If you take the trouble to look at the charts and understand them you will see this. There are no calculations for TECs you can do that are regular. Do them if you insist..... you will get an answer ....but it will be wrong. You should insist on getting the charts when you buy TECs or at least look them up on the internet. Obviously charts also have there inherent faults some are small and difficult to read accurately but you will often be far better using them than doing any calculations, and once you learn how to read charts it is much faster BUT you must remember you still only have ballpark figures. The charts are hotside temperature related so unless your hotside is exactly at the temp of the chart rating the results you get wont be the same. If you think about it this is the very reason why your calcs will never be right ! They are not temp related.
Calm down man! Why shouldn't the calculations be temp related?

You don't need to tell me that calculations wont be exact, I know, I have done engineering mathematics before. But you can calculate to margins of error by making assumptions.

I'm not looking to do stupid calculations that don't apply to TECs, I want to know how to do TEC mathematics, if it's chart based that's fine, but I'm not going to spend ****loads of money on setting a system up without at the very least ballpark figures.




Anyway, thank you again un-nerfer, dw about getting to -20oC, -9 unloaded on the cold plate is fine, I wanted -20 on the cold plate to be able to get the CPU temp as close to 0 as possible. I would like to know where you learned so much about TECs though, do you know of any online resources for it? or is your kowledge on it academic/experience based?

Also PWM control would be awesome, but do TECs work well with PWM? I thought that was best for momentum based applications, also I'm guessing you'd need some hefty transistors to carry that current.
    
CPUMotherboardGraphicsRAM
e5645 @ 4.2Ghz P6X58D-E 5450 x2 6Gb Corsair Dominator 1600 
CoolingOSPowerCase
ALPENFOHN K2 Windows 7 Pro Antec Truepower Trio 650W Dell Optiplex 
  hide details  
Reply
    
CPUMotherboardGraphicsRAM
e5645 @ 4.2Ghz P6X58D-E 5450 x2 6Gb Corsair Dominator 1600 
CoolingOSPowerCase
ALPENFOHN K2 Windows 7 Pro Antec Truepower Trio 650W Dell Optiplex 
  hide details  
Reply
post #14 of 19
Quote:
Originally Posted by Crabid View Post
Calm down man! Why shouldn't the calculations be temp related?
I am not in the least bit heated !
Ohms Law for instance is not temp related plus is a linear calculation but as far as I know there is no other way to doing the trianglar calculation of amp, Ohms and volts. There are no specific TEC calcs useful to anyone but the manufacturers thats why they produce charts because TECs are totally non-linear and you cant use the standard calcs which are all linear.

Quote:
Originally Posted by Crabid View Post
I'm not looking to do stupid calculations that don't apply to TECs, I want to know how to do TEC mathematics, if it's chart based that's fine, but I'm not going to spend ****loads of money on setting a system up without at the very least ballpark figures.
What I trying to impress is that maths and TECs dont really mix get the charts for your TECs.One often uses Ohms Law can at times, dependant on the TEC, be a long way off but if your using multiple TECs even a small deviation can end up miles out.
The was a constantly quoted formula for the delta t but that assumed Dt was linear across the range of input power and that the hotside temp at full power was the same whatever the input - it isnt. Some of the possible Dt calcs using that formula were totally ridiculous but that didnt stop it going worldwide to all and sundry.

Quote:
Originally Posted by Crabid View Post
Anyway, thank you again un-nerfer, dw about getting to -20oC, -9 unloaded on the cold plate is fine, I wanted -20 on the cold plate to be able to get the CPU temp as close to 0 as possible. I would like to know where you learned so much about TECs though, do you know of any online resources for it? or is your kowledge on it academic/experience based?
One of the manufacturers had a very good guide online free but it disappeared I have just done a quick google take a look at this ;- http://www.tetech.com/Peltier-Thermo...r-Modules.html
It gives instructions for reading charts properly, the correct mounting procedure and even talks about PWM !
The only other way to learn about TECs like I did and un-nefer is to use them. You never know the final outcome till you set the TECs up and run them even if only in a test setup.

Quote:
Originally Posted by Crabid View Post
Also PWM control would be awesome, but do TECs work well with PWM? I thought that was best for momentum based applications, also I'm guessing you'd need some hefty transistors to carry that current.
PWM is actually the recommended way to power TECs by the manufacturers.
Edited by zipdogso - 3/25/11 at 5:46am
post #15 of 19
Are both of you forgetting that ultrasonic has written a calculator, which can make fairly accurate calculations on TEC's. And fortunately does not make calculations based on the Ohm law. In-fact all of un-nefer's calculations are from ultrasonics calculator. I agree with zipdogso, I wouldn't even use the Ohm law. Your calculations will be far from accurate, and personally wouldn't even consider using them for ball park figures.

Basically temperature of TEC's greatly affects their characteristics, and because the ohm law can only make calculations based on resistance, amperage and voltage. Temperature can not be added to the equation, thus the results of the calculations will be highly inaccurate. I would love to know the calculations from ultrasonics calculator, however ultrasonic prefers to keep them a secret. Which is fine, considering he has spent the time researching.

GOOD LUCK
Edited by xtremetechuk - 3/25/11 at 10:08am
post #16 of 19
Thread Starter 
Quote:
Originally Posted by zipdogso View Post
What I trying to impress is that maths and TECs dont really mix get the charts for your TECs.
They must they're used in industry...

Quote:
Originally Posted by zipdogso View Post
The was a constantly quoted formula for the delta t but that assumed Dt was linear across the range of input power and that the hotside temp at full power was the same whatever the input - it isnt. Some of the possible Dt calcs using that formula were totally ridiculous but that didnt stop it going worldwide to all and sundry.
Sounds like a second or third order differential calc (second order put me off engineering at uni).

No wonder charts are used...


Quote:
Originally Posted by zipdogso View Post
One of the manufacturers had a very good guide online free but it disappeared I have just done a quick google take a look at this ;- http://www.tetech.com/Peltier-Thermo...r-Modules.html
It gives instructions for reading charts properly, the correct mounting procedure and even talks about PWM !

You're a legend, thank you very much for the guide.


Quote:
Originally Posted by zipdogso View Post
The only other way to learn about TECs like I did and un-nefer is to use them. You never know the final outcome till you set the TECs up and run them even if only in a test setup.
My main issue is the money involved in setting up a TEC cooling system, I really can't afford to play around with them...

Including the few hundred pounds of liquid cooling gear I'd need to buy on top of making my own cooling blocks, insulating the loop, the power supply, modding my case to fit another 480 rad & PSU...

It's too much to do on faith...


Quote:
Originally Posted by zipdogso View Post
PWM is actually the recommended way to power TECs by the manufacturers.
I guess heat has a momentum of sorts of it's own.


BTW, I originally though of Ohm's law because of a graph I found of temp vs. resistance. Of course for TECs, there is a vast difference and for every TEC that curve is different so an average can't really work without vast calculations (not to mention the inductance...)

I'm glad I asked
    
CPUMotherboardGraphicsRAM
e5645 @ 4.2Ghz P6X58D-E 5450 x2 6Gb Corsair Dominator 1600 
CoolingOSPowerCase
ALPENFOHN K2 Windows 7 Pro Antec Truepower Trio 650W Dell Optiplex 
  hide details  
Reply
    
CPUMotherboardGraphicsRAM
e5645 @ 4.2Ghz P6X58D-E 5450 x2 6Gb Corsair Dominator 1600 
CoolingOSPowerCase
ALPENFOHN K2 Windows 7 Pro Antec Truepower Trio 650W Dell Optiplex 
  hide details  
Reply
post #17 of 19
Thread Starter 
Quote:
Originally Posted by xtremetechuk View Post
Are both of you forgetting that ultrasonic has written a calculator, which can make fairly accurate calculations on TEC's. And fortunately does not make calculations based on the Ohm law. In-fact all of un-nefer's calculations are from ultrasonics calculator. I agree with zipdogso, I wouldn't even use the Ohm law. Your calculations will be far from accurate, and personally wouldn't even consider using them for ball park figures.

Basically temperature of TEC's greatly affects their characteristics, and because the ohm law can only make calculations based on resistance, amperage and voltage. Temperature can not be added to the equation, thus the results of the calculations will be highly inaccurate. I would love to know the calculations from ultrasonics calculator, however ultrasonic prefers to keep them a secret. Which is fine, considering he has spent the time researching.

GOOD LUCK
I will look that up, thanks.

So far as resistance goes, IF Ohm's law were true in the case of TECs, you could substitute R in the equation for an equation based on the temp of the TEC and produce an acturate formula, however, (I have since read) TECs (obviously now) have significant inductance from the heat they transfer. The combination of this and the massive temp difference across them makes the maths very complicated very fast hence, I imagine, the charts.
Edited by Crabid - 3/25/11 at 7:10pm
    
CPUMotherboardGraphicsRAM
e5645 @ 4.2Ghz P6X58D-E 5450 x2 6Gb Corsair Dominator 1600 
CoolingOSPowerCase
ALPENFOHN K2 Windows 7 Pro Antec Truepower Trio 650W Dell Optiplex 
  hide details  
Reply
    
CPUMotherboardGraphicsRAM
e5645 @ 4.2Ghz P6X58D-E 5450 x2 6Gb Corsair Dominator 1600 
CoolingOSPowerCase
ALPENFOHN K2 Windows 7 Pro Antec Truepower Trio 650W Dell Optiplex 
  hide details  
Reply
post #18 of 19
Quote:
Originally Posted by un-nefer View Post
Hmm, -20C is a pretty big ask for TECs TBH.

Sure it's easy to achieve -20C or higher at idle, but once your heat load increases, you'll need a pretty beefy TEC setup to stay anywhere near -20C.

I'll assume you'll be trying to achieve a higher overclock then 4.21Ghz on your i7, and for that, your i7 will be putting out over 200W of heat - we'll guesstimate 250W. I'll also assume your watercooling setup will keep your TEC hotside to 35C.

With the above assumptions and guestimates, you could use 4x 50mm CTE TECs and run them all at 24V and you'd achieve -18C under load (assumed above). But in doing so, the 4x TECs will draw a total of 51A at 24V - for a total electricity draw of 1224W.

Not exactly efficient, and a 1300W 24V PSU is not exactly cheap either.


At 12V, you'd "only" achieve around 5C under load, but the whole setup will be much more efficient, as the 4x TECs will only draw a total of 26A at 12V - for a total electricity draw of only 312W.

You could get to -6C with 4x 331W CTE 50mm TECs running at 24V, then the TECs would only draw around 27A and you might then get away with a 650W 24V PSU. But that's -6C, not -20C.
Hate to sh*** on your parade, but this is all bull.. you can't achieve these results.. after about 30º off your warm water tec 's have a really high diminishing returns, and this data is based of a bull**** calculator that some idiot invented without actually testing the data!
LKTEC
(13 items)
 
  
CPUMotherboardGraphicsRAM
i7 2600K @ 4.2ghz GA-P67A-UD4-B3 mobo 2x ATI 5770 in crossfire ripjaw 8gig 1.5v 8cas 
Hard DriveOSMonitorPower
69gig ssd Windows 7 Ultimate Old Hp f2105 Thermaltake 850 
Case
Haf 932 
  hide details  
Reply
LKTEC
(13 items)
 
  
CPUMotherboardGraphicsRAM
i7 2600K @ 4.2ghz GA-P67A-UD4-B3 mobo 2x ATI 5770 in crossfire ripjaw 8gig 1.5v 8cas 
Hard DriveOSMonitorPower
69gig ssd Windows 7 Ultimate Old Hp f2105 Thermaltake 850 
Case
Haf 932 
  hide details  
Reply
post #19 of 19
As longs as you can keep the TEC hotside at the temp I stated, then the TEC coldside will be very close to the temps I stated.

Please stop trolling the TEC forums. If you want to prove that Ultrasonic2's TEC calculator (or all of them for that matter) are incorrect, then do some tests, because no one wants to hear what you personally think, they want to see evidence or they want to see you stop posting.

Leslie
Crimson Basilisk
(25 items)
 
  
CPUMotherboardGraphicsGraphics
Intel i7 2600K @ 5.1Ghz AsRock Extreme4 Gen3 Palit GTX 680 Jetstream 2GB Palit GTX 680 Jetstream 2GB 
RAMRAMHard DriveHard Drive
G.Skill F3-12800CL9D-8GBXL G.Skill F3-12800CL9D-8GBXL Sandisk Extreme 120GB SSD Sandisk Extreme 120GB SSD 
Hard DriveOptical DriveCoolingCooling
WD Caviar "Black" 1TB SATAII LG WH14NS40 Blu-ray Burner XSPC Rasa CPU (Acetal) Black RX360 Triple Fan Radiator 
CoolingCoolingOSMonitor
XSPC DDC Two BayRes BMaverick Laing DDC (MCP350) Windows 7 Ultimate x64 Philips 201B 21" CRT 
MonitorKeyboardPowerCase
ViewSonic VX922 19" LCD Diamond Digital Black Antec True Power 750w Antec LanBoy Air Red 
MouseMouse PadAudioOther
Epic Gear Meduza (By GEIL) Steel Series QCK+ XL Asus Xonar DX ROCCAT Kave 5.1 Surround Sound Gaming Headset 
Other
Sennheiser hd555 
  hide details  
Reply
Crimson Basilisk
(25 items)
 
  
CPUMotherboardGraphicsGraphics
Intel i7 2600K @ 5.1Ghz AsRock Extreme4 Gen3 Palit GTX 680 Jetstream 2GB Palit GTX 680 Jetstream 2GB 
RAMRAMHard DriveHard Drive
G.Skill F3-12800CL9D-8GBXL G.Skill F3-12800CL9D-8GBXL Sandisk Extreme 120GB SSD Sandisk Extreme 120GB SSD 
Hard DriveOptical DriveCoolingCooling
WD Caviar "Black" 1TB SATAII LG WH14NS40 Blu-ray Burner XSPC Rasa CPU (Acetal) Black RX360 Triple Fan Radiator 
CoolingCoolingOSMonitor
XSPC DDC Two BayRes BMaverick Laing DDC (MCP350) Windows 7 Ultimate x64 Philips 201B 21" CRT 
MonitorKeyboardPowerCase
ViewSonic VX922 19" LCD Diamond Digital Black Antec True Power 750w Antec LanBoy Air Red 
MouseMouse PadAudioOther
Epic Gear Meduza (By GEIL) Steel Series QCK+ XL Asus Xonar DX ROCCAT Kave 5.1 Surround Sound Gaming Headset 
Other
Sennheiser hd555 
  hide details  
Reply
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Peltiers / TEC
Overclock.net › Forums › Cooling › Specialized Cooling › Peltiers / TEC › New cooling, some questions I need to ask first