fast rectifier

[Martin Czech]

>    Someone on the list asked for rectifier cicuits with fast performance
>    a while ago. Unity gain compensated op-amp suffer from low slewrate
>    in such precision rectifier applications where there is no feedback
>    via the diodes for some time. So the idea is to use op-amps with external
>    compensation, and "switch of" compensation while the input voltage
>    crosses the 0 line, the most critical moment. 
> 
> I like the single transistor Moog circuit a lot better; it doesn't
> suffer from glitchy behavior due to slewing.
> 
> This is the circuit that is used on the VCOs to generate a triangle
> wave from a sawtooth -- it's a single common-emmiter transistor biased
> so that it runs inverting for half a cycle and the base-collector
> junction conducts and it runs non-inverting for the other half a
> cycle.

Mmm yes, is this both a emitter and collector circuit, so both inverted
and non inverted is available, right ?  This is certainly yhe fastest
possible circuit, I aggree.  No flames from me , but one question :
will temperature effects of the single transistor have some influence
on symmetry (not so much on emitter side, but collector ?). Will this
lead to the situation that both halves don't fit together properly?
Glitches again?  Or is the circuit ruled by resistor dimensions, that
is very stable? I think small signal equations won't lead very far,
better way is to check real circuit with hair dryer, or heat pistol.
Maybe I check this out after holiday.

m.c.