INA333: Switched Gain Resistors

Hi Laura,

keep the stray capacitances and parasitic capacitances low. And keep an eye on leakage currents.

Kai

Hi Laura,

There are some things to be concerned about with using a mux for the gain setting:

1) As you've mentioned, the Ron of the mux is going to contribute gain error and non-linearity with fluctuations in on resistance and channel-to-channel mismatch.

2) The mux capacitance is going to be the biggest concern here, because this capacitance shows up right on the inverting nodes of the internal op-amps, and this is a good way to make an op-amp unstable. Unity gain is always your worst case from a stability standpoint, so if unity is one of your configurations you will need to keep the parasitic capacitance extremely low (I'd say 10pF or less, but I think you will struggle to find this in a mux). The more gain in your circuit, the more capacitance you can get away with.

3) There will also be some settling time associated with the mux capacitance and charge injection when switching channels. You will need to make sure you allocate ample time for the INA to settle before you do anything with the signal. The same thing applies if you are muxing the input to the INA.

4) As Kai mentioned leakage currents will also be a concern because of the relatively large impedances (50kOhms) around the internal op-amps. leakage currents will effectively see 100KOhms + RG and this may produce significant error.

I need clarification on item 4. Are you referring to input bias currents for the internal op-amps? Are you concerned that 100kOhms + RG + Ron(switch) will be too large?

Thanks.

Hi Laura,

Ron shouldn't be large enough to cause concern in this regard, I was actually referring to the leakage current through each of the off mux channels.