RE(2): matched expo source / sink idea

[Haible Juergen]

	>Your revised circuit won't work right, because you are driving the
base of
	>Q1 with the base current of Q4.  Q1 should be driven with the
emitter
	>current of Q4, either as 2 PNP's in a darlington configuration, or
as a
	>NPN/PNP with one of them connected as a diode.

Hi Terry,

I agree it's probably not a good idea.

And I surely trust your expertise more than the Spice simulations I've done
(You can never trust Spice for slightly out of the ordinary configurations).
But as for not working right, I truely wonder why.
What's wrong with driving a transistor base with the base current of another
transistor ? What I *expect* is the following behaviour:

(1) I force a voltage at the emitter of Q4.
(2) There is no doubt a conductive path over EB of Q4, BE of Q1, and R3, to
have
a current flowing.
(3) The base current for Q1 would be provided by the base of Q4. I don't see
that
there would be a restriction for Q4 providing this current, other than a
much larger
current would flow thru the collector of Q4, than thru its base. (With Q1
being
similar to Q4, I expect collector currents of similar magnitude, but that's
not
critical.)

Where is the error ?

As I mentioned at the beginning, I don't think it's such a great
configuration,
i.e. your proposing a resistor load for my 1st version, and Osamu's placing
the transistor diode after the pnp follower, are both preferable. 

	JH.

(I should probably just go to the breadboard rather than typing this ...)