A key aspect of this design was that it needed to reject atleast a 12V common mode signal. To achieve this, I attenuate both the common mode and differential signal at the first stage of my CSA (any other ideas??? I’d be interested to hear about them). This brought signal+Vcm to within the operating limits of my op-amp, but also substantially affects my S/N ☹. Anyhow, lets see how it turned out in practice with a few test points.

Note that I’m only looking at DC common mode here since this is what I’ll typically encounter. I’ll likely do a sweep of the AC rejection later.

Test signal is 50kHz, 50mVpp, AC coupled across the inputs of the CSA. The common mode voltage is provided by an external bench supply.

Vcm=0V, Vo_mean = -0.3mV

Vcm=5V, Vo_mean = 71.6mV

Vcm=10V, Vo_mean = 132.55mV

Well that was a bit disappointing, but I think my test setup is flawed. I added some series impedance to isolate my Vcm source, but this would have created a voltage divider at the R1 and R5 in my schematic. I’ll likely be doing some tweaking of both my circuit and test setup. (pretty sure I can DNP both R1 and R5 in this REV of my CSA)