Originally Posted by zipdogso;13115075
This is your problem..... dont take it from me anyone here with half an iota on how TECs work will tell you, you are wrong. As the Dt goes down the max Qc at the operating point goes up but the rate at which it is moved goes down untill such point as the Dt reaches 0ºC and equilibrum is achieved (hot and coldsides are both same temperature.) and no heat is moved. It is quite possible for there to be a negative Dt where the coldside is actually warmer than the hotside (coldside and hotside are determined purely by the direction of electron flow not neccessarily upon them being hot or cold.) the graph for Qc/amp is then inverted.
I'd like to take a poll on this.. Here is an analogy.. dig a hole next to the ocean. you'll have water coming back into your hole from the ocean, now put a little pump to pump the water back into the ocean. the water from the ocean will want to come back in and fill the hole up (0dt net) but if you don't have a big enough pump, you can't drain the hole, it will stay full, but that doesn't mean your pump stopped pumping water. Now put a bigger pump on it and all of a sudden your hole is drained, and will say drained as the pump overcomes the water comming back into the hole.. heat pumps work the same way... they stop pumping when you shut them off or at 0ºk when there is no more heat to pump
A.) you dont fully understand how TECs work, the 2 TECs you have are not really sufficient for your experiment because you dont understand the operating parameters...I will come to that .....bear with me.]
I never gave my operating parameters.. my tecs are well suited for what I want to do. the test was done on a stock 920 because I was building a 2600k which has lower power than the 920. so the test was done based on a worse case scenario
B.) the idea of Dt in watercooling is different to that in TEC. Dt in TEC is the actual temp of the actual plates on the sides of the TEC not the air/coolant etc waffling around them. You could never guage your results to that of a TEC chart like you attempted to because you are just recording the temp of the coolant. So there will obviously be much more "loss" due to thermal transfer making the TECs look much worse than they really are.
THIS IS EXACTLY WHAT I WAS TRYING TO ESTABLISH.. No doubt that the tec was performing to the charts.. my point was that thermal resistance plays a huge part in real world results!!!!!!!!!!! THank you very much..... This is what I was getting at.. again.. THANK YOU!!!
C.) the setup of the TECs is vitally important and you have said nothing of how you have done this. We have ascertained you wired them in series how did you you install them into the coolant line - in series (not so great), parallel (definately better.) and did you use a seperate loop/pump/rad for the hotside ?
Yes I did wire 2 in series to a 24v 25amp meanwell. I did put the tecs in series only for the fact this is going into a 5 1/4 bay and some sacrifices have to be made.. we can account for maybe 1 or 2 degrees at most based on other parallel series wc projects.
I have lapped all plates, checked the TIM footprint and torqued to manufacturers specs.
Next it appears from your picture you used a plug in meter...so I suppose you plugged your power supply into that, you didn't say.....you wanted to calculate the heat off the hotside so you used Ohm's law to arrive at the input power...BUT your meter is recording the input power to the power supply...not the input power to the TECs which is the figure you should be using so what you should do is set the voltage out on the power supply then look at the TEC chart to see the current draw........
No sir, I don't know the resistance so ohms law doesn't apply here.. I used the wattage at the wall and took a reading of volts to determine amps
So why did I say your TECs are not sufficient......think about it...you reckon you a load of 160w plus 30w residual stuff 190w in total. Right so we ascertained you using roughly 10v per TEC to start from experience I would say your Dt would be between 30º & 40º Dt at 10V so the draw is 14.5amps roughly at 14.5amps with a Dt between 30º and 40º the max Qc at that operating point is around about 65watts times that by 2 (2 TECs) is 130w....what do you reckon your load is ???? 190w. Hmmmm. So you aren't clearing your heat load. hence the crap looking results....
To use 2 TECs is part of your downfall you should be using one large one and I mean a large one.
Undervolting really should only be done with at least 4 TECs otherwise the Qc is no where near large enough. Of course if you understood how TECs work properly you would of understood the parameters in which you need to work and made a better informed choice of TEC to use.......
When someone builds a bigger single tec, I'll def try to build a chiller with it.. I just don't know of anything bigger than a 400w 62mm tec, which would be smaller then my 2 5mm @ 245w
lichEdited by lichking - 4/18/11 at 10:59am