The big TEC driver thread!

[Solarfire]

Works for me. I'll PM you my home address.

I loaded the code you sent me, it compiles. Looking at reworking for PID now.

BTW, the temperature limits are hard coded and based on a 100Kohm thermistor. I will start with 25c as the setpoint, unless you feel it should be lower. I may also change the thermistor to 10K once I see a board, as these are more common.

25°C is a good setpoint! No problem with the 10k either.
If you need anything just let me know.

[RedlumX]

Crude on-off type of "regulators" are discouraged from any TEC manufacturer, so better make it very slow as the thermal cycling will reduce lifetime quite a lot. It is not the way a TEC is intended to be used. The worst thing are unfiltered PWM regulators such as the one from oatleyelectronics.com. Usually one aims for 10% PWM ripple max for a TEC driver.

[LaNeK779]

aaarghhh... all i need is a simple - effective circuit to control a tec... without having to build the pcb myself.... and without soldering smd compopnents myself....


what do you think about our beer keg device mr redlum???

[lasersbee]

What I can see is a variable voltage control to the TEC with
feedback sensing to hold the TEC at a preset heat transfer
level.

Jerry

[dnar]

PID or PI control will be fine. I have yet to see a TEC die from thermal shock under my control.

[RedlumX]

aaarghhh... all i need is a simple - effective circuit to control a tec... without having to build the pcb myself.... and without soldering smd compopnents myself....


what do you think about our beer keg device mr redlum???

Well if the TEC just goes on and off it should be OK if it is neither too fast nor too slow; but I have seen unfiltered PWM controllers for TECs that go on or off like 100 times a second, which is very bad. Actually one should switch fast enough such as to avoid thermal cycling, ie, the temperature should not vary much in-between. And one one should not too fast in order to avoid AC losses in the TEC .

Professionally designed drivers go with, say 100khz.. 1Mhz and use inductors in a buck configuration to cut down the ripple. Again, a maximum PWM ripple of like 10-15% seems to be the general recommendation.

[Solarfire]

Here's an interesting source for high-end TEC-Controllers and LD-Drivers. A little pricey but very compact DIP packages for implementing on PCB's.

http://www.analogtechnologies.com/tec-controller.html

[michelrietveld]

this is a must have tec driver for my el cheapo rgb ;-)
Michel

[dnar]

how about i send you 10 boards, you work your magic and then send back as many as you wish, keeping a number of boards for your effort and time?

Thanks to Solarfire for finding these boards, and LaNeK779 for shipping 10 board to me (received today) I'll shoe-horn the PID code in this weekend and test using 33W to 60W TEC's and a 445 @ 1A.

Stay tuned.

[dnar]

Ok, I have the Micromatch programming port connectors due to arrive tomorrow, Saturday is set aside to bash the PID code in to these boards.

What is the consensus on the temperature set point? From comments above, I am leaning towards 22c. Also, what are the typical TEC power ratings? I have TEC's between approx 30 to 66 watts to test with.