Hey Bastiaan, How to you remove the heat?
Why don't you use a liquid coolant in a jacket surrounding the cell -- like Rossi seems to do? That would accomplish reasonably accurate calorimetry for you "automatically" with nothing more than a flow meter, two thermometers, a known electrical power source.for calibration, and a computer/data logger. All of those are cheap and easy these days. If you don't want to complicate the system, you don't need the coolant jacket -- you can run the device in a well insulated water bath. Calibration will compensate for any losses from the bath.
