Page 111 - The Art and Science of Analog Circuit Design

One Trip Down the 1C Development Road

MIRROR MIRROR
#1

Figure 8-8.
The control circuit.
the amplifier stable. The basic current feedback topology I had chosen was
the feedthrough problem. Now it was obvious why the discrete circuit was
operating inverting. The problem goes away when the non-inverting input
is grounded because then the inverting input has very little signal on it.

Redefinition

At this point I realized I must go back to the beginning and look at volt-
age feedback. I started with the basic folded cascode topology and
sketched out the circuit of Figure 8-9. It seemed to work and there were
no feedthrough problems. It also appeared to simplify the control re-
quirements, since there were no PNPs to steer. While working with this
circuit I realized that the folded cascode transistors, Q7 and Q8, could be
used as the steering devices, and sketched out Figure 8-10. This looked
great since it had fewer devices in the signal path and therefore better
bandwidth. The only downside I could see was the critical matching of
the current sources; all eight current sources are involved in setting the
gain. While I was pondering how to get eight current sources coming

94