[sdiy] Some observation with op amps and AC signals

[Czech Martin]

Some practical obervations:

for the improved temperature chamber servo I'm currently finishing,
I use a bridge sensor and an instrumentation amp, consisting
of two OP07 noniverting buffers and a OP07 differencial amp.
Overall gain about 1000.
The classic circuit with a trim pot to tune the CM supression.
Still cheaper then to feed the sensor from a precision band gap
or burried zener reference. This will cancel out any supply
or thermal variations of the regulator.

E.g.:

http://www.mines.edu/Academic/courses/physics/phgn217/lab8/lect13/sld012.htm

Now I wanted to trim the CM suppression. So I fed in a 6Vpp CM signal,
arround 10Hz, and watched the residual output. I assumed I would see some
noise and the sine CM signal. But when I trimmed the circuit the 
residual output got lower, but obviously the second harmonic creeped
up, until it was as strong as the fundamental.

This did no go away with reducing frequency.
Obsiously, if we look at cancelation circuits like this,
the residual can be very distorted.

The CM amplitude was 20mVpp, so I guess the circuit has a CM
rejection of 1/300 or -49.5dB @ 10Hz.

The ratio of differential mode and CM amplification is about
109.5 dB.  He, this is not so bad, or am I blind?

Perhaps this woul look better if the DM amplification
would be not so high.
Still, some strange feeling is there, perhaps using IA for
CM line suppression, like these SSM21XX parts, is not so good.


m.c.