Bob Blick wrote:
>> As for the opamp and FET bit -- the 10-ohm resistor is (I assume) there 
>> to stop the opamp overdriving the gate on the FET. 
> 
> It's the other way around, it isolates the opamp from the large amount
> of gate capacitance of the MOSFET. Otherwise the opamp would probably
> oscillate.

Ah, well that makes more sense. The IRFP450 has nearly 4nF of gate 
capacitance (worst-case) so it's easy to see how that could potentially 
make the circuit oscillate, especially if the input to the opamp (the 
current-set input) was changed rapidly.... as might happen if you 
toggled the switch rapidly, or used a D/A converter to drive it.

> Probably again, a stability issue. The LM358 has a nasty class-B output
> stage and without the resistor the circuit probably suffered from
> oscillation, or at the very least, ringing when hit with stepped changes
> to its input. The resistor reduces the gain of the opamp.

Hmm, OK, but I would have thought it would have had very little effect. 
Most of the current in that part of the circuit is coming from the 
source side of the FET (a potentially quite high current section).

The only thing I can see that resistor doing is pulling the output 
towards ground (or rather, towards whatever voltage the top of the sense 
resistor is sitting at) when the two transistors in the LM358's output 
stage are operating near their Vce-drop limit.

Though the LM358 -- while not rail-to-rail -- does have an output range 
that includes ground, which brings us back to "why?"

Argh.

Thanks,
-- 
Phil.
piclist@philpem.me.uk
http://www.philpem.me.uk/
-- 
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist