Il 17/10/2012 00:20, Jamie ha scritto:
Perhaps we miserunderstood something :) I built your first circuit (bjt + zener) and it changed too much the current over temperature span. Then, I modified it with the suggestions of John S and now it doesn't change the current but 1-2 mA. In both cases the measures are very close to the sim outputs.
I didn't build your second circuit (2 bjts). I've just simulated it. I wrote it's better than the first one but it exhibits a "not so small" temperature dependency. Of course, this time the simulation couldn't be so accurate.
Thank you anyway because (currently) I solved using a circuit based upon you first idea.
Marco