Current Mirror: The V-mirror

[rtarbell]

I saw this specific mirror referred to in the related "Current Mirror Discussion" thread. I had a question about it, but I didn't want to thread jack.

This mirror has been referred to as the "V-mirror", or the "Gilbert Current Mirror".

--> If we are using an opamp to equalize the drain voltages between the two transistors, isn't there a potential for positive feedback (depending upon the load connected to the output side of the current mirror)? I believe that the path from the output of the opamp to its inverting terminal, and the path from its output to its negative terminal, *could* be the same path.

If I remember my nuts and bolts properly, I think that an opamp must have more negative feedback than positive feedback if it is to act stable (instead of a switch like a schmidt trigger). In this schematic, there is the potential for more positive feedback than negative feedback. Am I missing something?

[darkfenriz]

It assumes infinite impedance of the current source and finite and positive impedance at the output, so there is more nagative feedback then positive.
In theory it could work, but will be very sensitive to capacitive loading (stability!) and to any overdrive situations.