I've learned to run them from a constant current source, not a voltage source. This helps to avoid getting into the region of the device where the Delta T gets inefficient. Its not how the book explanation says to drive them. The manufacturers all seem to quote the same old tired explanations of how to use the response curve data. I find if I starve them for voltage, rather then overdrive them, I can find the operating point easier. I just had to figure out where to drive a unknown large TEC in a DPSS head, cooling a pump diode, and I switched the bench supply to constant current and more easily found the "sweet" spot. The Delta T and stress across the TEC is actually less then it would be if I ran it at some arbitrary constant voltage, for a given diode temperature.
But its easier to tell a non-engineer just to buy a different sized part then it is to get them to design a proper controller. This is why TEC has such a bad reputation.
Your mileage may very, but I've found it to be easier. Granted most people buy OEM power supplies of constant voltage.
Steve
Qui habet Christos, habet Vitam!
I should have rented the space under my name for advertising.
When I still could have...