Other Parts Discussed in Thread: OPA1622, LM4562, TINA-TI, OPA827, OPA828
I am a little mystified as to how the high bias current of these devices interacts with resistances in the bias position and on the inverting input. Part of that is my lack of an EE background. The series resistance limitations are clear -- keep it under 10K.
It is common to put 47KΩ-100KΩ bias resistors at the non-inverting inputs of voltage followers, some with input impedances down to 30KΩ...the LM4562 is a good example with 72nA Ib...but I'm unsure of what amp characteristics would turn this common practice into a noise liability. IIRC, the OPA1622 and INA1620 are spec'd at 60KΩ Zin and ~1200nA Ib.
1) How would either device handle a three-amp unity gain instrumentation front end application (U1), with high bias resistance and low series resistance? Most sources are <100Ω but some could be 1KΩ to 2KΩ, with the associated small capacitances from reasonable lengths of good quality cabling. This isn't designed for guitar or bass (obviously not a FET front end), but that would be 5KΩ to 13KΩ. Curious how noisy these scenarios would become.
2) Does anything change significantly if Rf and Rg are added to U1, as in a typical high-gain IA, assuming Rf is something like 1KΩ to 2KΩ and U1 gains are anywhere from 4 to 100?
3) How would either device handle a less-than-10K variable resistor at an inverting input (U2B)? Significantly worse than, say, a somewhat precision BJT device with 10nA of bias current - or not?
U2B is a ±10dB trim, by the way.
Thanks!
