bias resistor position

[Evandrofoz]

I have a second order RC filter on a differential input of my NI6224. The problem is when I need to use a bias resistor for floating soucer the combination of the filter and the bias resistor act like a voltage divider and atenuate my dc signal.

Actually I am using the bias resistor after the filer. Can I use it before the filter. The circuit on left shows my actual configuration, the circuit on the right shows what I will try. 

Can I do this? Whats the best solution in my case? Thanks

[johnsold]

I do not see any reason why putting the bias resistors on the input side should not work. The input bias current is specified as less than 100 pA so the voltage drops due to the bias current are comparable to or less than the accuracy limits.

Also note that the input impedance includes 100 pF.  This will be parallel to C5. If you are trying to control the cutoff frequency closely, you may want to consider that capacitance. Its tolerance is not specified but is probably much larger than the tolerance on your filter parts.

If your DC signal changes slowly, you might be able to use a guard circuit to provide the bias with minimal effect on the DC measurement.

Lynn

[Evandrofoz]

Thanks! Can you describe this guard circuit?

[johnsold]

Probably the best source on guarding is the Keithley Low Level Measurements Handbook, Section 2. It is in the right column at this link.  You may need to register to get the book, but it is well worth having if you want to make low-level measurements of voltage, current, or charge. 

Lynn