I found a particular Silicon-based sensor with an integrated opamp pre-amp. It has a gain of 10G (1E10, 200dB) using a 10G feedback resistor and a 190 Hz bandwidth. I call that a 1900GHz GBWP (Gain Bandwidth Product).
How do they do that? I'd like to be able to do it myself, outside of this sensor with the integrated opamp pre-amp. But I can't find an opamp with anything close to that GBWP. The highest I find at TI is 18GHz. The highest I find at Linear (LT) is 10GHz. And those also have far lower than a 10G open loop gain.
Are they cascading opamps but using a single feedback resistor across both? This is about low noise, so it has to have a very larger resistor. Two stages at 100K each won't do the trick. And I don't find anyexamples of cascaded opamps sharing a feedback resistor.
So, again, does anyone have an idea of how they do this?
I believe the opamp on that sensor might be on the same subtrate as the sensor. The same company has a small module with include opamp that I'm sure is not on the same substrate. The detail is hidden, but I believe this thing has a 1E8 (160dB) gain. Again, the highest open loop gain I've been able to find was under 130dB. But because this opamp isn't integrated onto the same substrate, I would think I could build my own from available parts. (And since there's isn't 200dB, I'm guessing they couldn't find off-substrate parts as good as their on-substrate opamp, perhaps.)
This is all about minimizing sensor noise by having a very large low noise first stage pre-amp.
Thanks very much.