Peltier modules are voltage specific from what I recall and need to have decent regulation. So in that regard I would use a PWM circuit, or try anyway, in order to control the cooling level of the module. At least on an experimental basis. On a design basis, I would use a thermostat with feedback similar to how a heated crystal oscillator works so as to keep the imager temperature consistent regardless of the environment. You would want something along the lines of a PID controller as opposed to a slam-bang thermostat.
Thanks for the Infos on the Regulator, and yes, they are Voltage Specific, the one i linked above wants 3,6V and 1.4A to run on full power. Looked a bit into it, and Driving Peltiers seems like a whole rabbithole on its own. However found some old Thread on another Forum about PWM Driving Peltiers, seems to be a bad idea to drive them with PWM, seems to want a Linear Regulated Driver.
.
https://www.mikrocontroller.net/topic/313553 .
[Its in German and about Astro Stuff]
.
I really dont want to build anything which ends up in Programming something, and every else available to Purchase as a ready-to-use-Unit seems to either come in a Huge Form Factor, is very unprecise or surpremely expensive (while those expensive nice ones then lack on-the-Fly adjustment options, and seem to allways need to be connected to a PC to change anything, which also sucks for this Project).
.
https://www.meerstetter.ch/products/tec-controllers .
I am kind of surprised, that there arent any Controllers for Low power which are easy to use and not huge, but will do some More Research on this and see if i can figure out Some Simple Solution. Otherwise i might end up building some Really basic Linear Driver with some OP-AMPs and a FET or Transistor and a bunch of Potentiometers to be able to adjust everything.
I assume they should be at least be driven Current Limited or something to not kill them instantly. Guess the "or try anyway" attempt would be worth a Try and just Use some cheap available Thermostat Unit. Not too sure how well it would end up working, because for the #1. Design i wrote about above, there isnt all that much Thermal Mass involved on the Cold side, so it could be end up beeing way to slow, and causing a lot of Temperature Changes all the Time. Pretty sure, it would work fine with most other designs, where there is a Cooler Block of certain Mass involved, this could end up beeing very precise and reliable if tuned in a bit.