I=92m looking at the thermal noise from a resistor down the bottom of a probe. Cable and probe are maybe 1/2 a meter long. I added a shield line to try and reduce the capacitance between the =91active=92 end of the resistor and ground. (The other end of the resistor is tied to ground at the bottom of the probe.)
The driven shield seemed to work great at the higher impedance levels
1Meg and 100k ohms. (Though a more careful examination showed there were some issues.) When I tried 10k ohms there was some serious gain peaking at the higher frequencies...above 100kHz. I mucked about a bit and made sure this wasn=92t the common problem of an opamp driving a capacitive load.Late yesterday it struck me that there is capacitive coupling from the shield back to the input. I had been mistakenly thinking of the shield as only a capacitance to ground. The capacitance of the inner conductor to the ground (Cig) is 60pF, from the inner conductor to the shield (Cis) is 85pF and from the shield to ground (Csg) is 160pF.
I was first using an opamp follower to drive the shield, but later added a bit of gain...and then threw it away.
+-----Cis--+ | | | |\ | +-----+--+ \ | | | >----+---+----+ | +-- / OPA | | | Rmeas. | |/ 134 | R1 Csg Rmeas. | | R1 | | +---------+ | | | GND GND GNDR1 was 50 ohms (to get rid of Csg ringing) And then this,
+-----Cis--+ | | | |\ | +-----+--+ \ | | | >----+-R4R4--+----+ | +-- / OPA | | | Rmeas. | |/ 134 | R5 Csg Rmeas. | | R5 | | +--R3R3---+ | | | | GND GND GND R2 R2 | GNDWhere R3 and R4 were 1kohm and R2 and R5 were 100 ohms.
This seems like it must be a known problem and I wondered if there are any simple solutions. I thought that a bit of inductance (L =3D Cis*Rmeas^2) in the right place might help, but I only managed to make a nice oscillator.
Thanks for any help or advice, George H.