Overclock.net › Forums › Cooling › Specialized Cooling › Peltiers / TEC › No experiance using TEC's, air cooling?
New Posts  All Forums:Forum Nav:

No experiance using TEC's, air cooling? - Page 3

post #21 of 25

I've had this discussion with you before man, I think your whole "the TEC needs to move 2x the heat load" thing is flawed, or at least misleading or an over simplification for newcomers to TEC cooling. If by 2x the Qmax you mean requiring 440w qmax for a 220watt heat load but that 440w qmax is at a 0 degree dt between hot and cold sides then yes your theory is correct, but when we have the charts to do the calcs why bother just going for the 2x tec qmax to heat load idea when you can more accurately calculate your actual heat load, and then see what dT your specific tec will hold to cover your actual heat load, not qmax = 2x your heatload and then leave the resulting dT to chance, your actual heatload and the approximate dT resulting from the heat that your tec can move at a given voltage.

Saying qmax must be 2x the heatload is just a guesstimate to cover your heatload and give you some sort of delta between hot and cold sides of the tec. I just find it misleading to say your tec must move 2x the heat load, it is physically impossible to move 2x the heat load, if you have a 220w heat load your tec can only move 220watts of heat, but if you picked that tec as having a 440w qmax at 0 degree dT then yes it is a good rule of thumb, because it should move the actual heat load of 220watts to a decent dT of ~20 or so degrees (this is purely made up numbers for the sake of demonstration).

I just find this one bit of terminology or rule of thumb that you keep preaching a little misleading, other than that your knowledge of tecs etc seems perfectly fine, I've just never been able to read your statements about qmax must be 2x heat load and agree with it in the terms you are stating as a rule of thumb. just out of curiosity, who told you this rule of thumb? because i've never heard it preached in the form you do and I've been reading these tec threads for years. I think you may be slightly confused with the rule of thumb which is actually to pick a tec that has qmax of about 2x the heatload at your TARGET DT(say 20*dt) not 0*dt at full power of the tec. this doubling of qmax to heatload at your desired dt at full power of the tec is actually to give you the headroom to undervolt your tec to the point where it perfectly moves your heat load at the desired dt and be very efficient in doing it. for instance, say we keep the 440w qmax, but instead of 0 degree dt its actually your desired dT of ~ 20 degrees, now we can undervolt the tec so that it moves only 220w of heat but still at your desired 20 degree dT. This is the only rule of thumb that i can think of that closely corresponds with yours but can be misinterpreted as what you are stating as a rule of thumb. It is very close, what you and i are saying, but your rule of thumb is leaving lots of crucial information out. ie if you state qmax must be double the heatload at 0degree dt then i can accept it because i know it will move your actual heat load and provide a delta between hot and cold sides albeit it leaving it up to chance rather than just running some simple calcs from the charts, but if you are saying that your tec must have double the qmax to heatload and do that while providing a delta of say 20ish degrees between hot and cold sides of the tec then i have to disagree. at that point in time you would not be moving 440 watts, you will still be moving 220watts but your dt will be much greater which will become very inefficient.

you can overshoot your required qmax a little to cover any inefficiencies in heat transfer etc or just accept a lower dT than the ideal, but this does not mean you need a qmax of 2x your heatload unless you are just guesstimating based on qmax of 440w = 2x heatload of 220w @ 0*dT and hoping this gives you a decent delta because your tec only has to move 220w.

I'm sorry i just thought we should settle this as you are a regular here and i find it misleading when you tell this rule of thumb to most newcomers to this section of the forum. i'm still prepared to discuss it with you but i will continue to try to get my point across to you as i can't make yours make sense with the information you are providing. Discuss biggrin.gif
post #22 of 25
My "rule of thumb" design starting point always is Qc = DT 20 @ 1/2 Imax. Qmax to me is useless, it only indicate its max Qc handling when it reach DT 0.
post #23 of 25
My starting point is to calculate actual heatload or maximum heatload of whatever I'm cooling, select desired dT range of say 20-30, decide on a voltage of my psu, ie I prefer to use variable analog dc voltage rather than PWMing 12v and with my heatload, dt and voltage range I run through the charts of a number of different tec's to find the most efficient one. I usually end up with a tec with high Vmax, low Imax meaning it has a high couple count and then undervolt it to its most efficient range or an efficiency that i can live with anyway, usually between 33-66% of Imax. I like high couple count tec's as it gives me a large voltage range to control it over, so i can use variable dc of 9.6-26.4vdc with a meanwell PSU with vtrim function. This is for a chiller set up though. I would approach it a little differently for direct die setups. I still calculate my qmax but i calculate it at the actual dT i want rather than just leaving it to chance, i find this way you are far more likely to achieve the dT close to what you want rather than just "rule of thumbing it" and not doing the calcs.
post #24 of 25
perhaps i should start using the term Qc (watts) @ 20*dT instead of saying Qmax @ 20*dT. Qmax, as you say, would indicate it's max Qc which is always achieved at 0*dT. Perhaps this is where Zythe and I are getting mixed up, ie he is correctly stating Qmax at 0*dT should be double heatload, but this still leaves the resulting dT to chance rather than simply running some calcs to get a better idea of how efficient it will be, i can find a large range of tecs that will have the same Qmax at 0*dt but it does not tell me which one will be the most efficient in doing it, which is something I always want to know.
post #25 of 25
Originally Posted by LiamG6 View Post

I prefer to use variable analog dc voltage rather than PWMing 12v
Couldn't agree more, TEC should only use DC, PWM's efficiency really bad at low duty cycle and wasteful.
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Peltiers / TEC
Overclock.net › Forums › Cooling › Specialized Cooling › Peltiers / TEC › No experiance using TEC's, air cooling?