Help with AGC circuit

[Mark Atom]

I'm designing an AGC circuit using Teledyne Cougar parts. The AGC amp is
the AGC230, the detector is the DAC4101. I'm trying to hold the output of
the AGC amp at -5 dBm. -5 dBm into the detector produces about .7 volts
output. I used and error amp with a reference input connected to a gain
stage to drive the AGC amp. This works OK, but not great. There is still
some variation in the output of the AGC amp when the RF input changes. The
detector output is linear over a wide range (-40 to +10 dBm), but the
control voltage vs gain curve for the AGC amp is not. I suspect that I'm
operating on a sensitive point of the AGC amps curve, where small changes
in control voltage make large changes in gain. However, I thought that the
Feedback path would then to linearize those changes.

Any help is appreciated.

Mark

 

[John O'Flaherty]

I'm designing an AGC circuit using Teledyne Cougar parts. The AGC amp is
the AGC230, the detector is the DAC4101. I'm trying to hold the output of
the AGC amp at -5 dBm. -5 dBm into the detector produces about .7 volts
output. I used and error amp with a reference input connected to a gain
stage to drive the AGC amp. This works OK, but not great. There is still
some variation in the output of the AGC amp when the RF input changes. The
detector output is linear over a wide range (-40 to +10 dBm), but the
control voltage vs gain curve for the AGC amp is not. I suspect that I'm
operating on a sensitive point of the AGC amps curve, where small changes
in control voltage make large changes in gain. However, I thought that the
Feedback path would then to linearize those changes.

You might try increasing the gain of the gain stage to reduce the
error. You might end up oscillating, but there may be some room before
that happens.