1
\$\begingroup\$

I'd like to send two currents along a long twisted pair (well, ~5m, but I want to be very careful about noise in this experiment). The two currents are produced by two photodiodes with equal and opposite bias voltages, with the same incident light, giving equal and opposite currents. At the receiver end, these currents are sensed by a transimpedance amplifier into an ADC.

Sending voltages differentially gives a benefit to common mode noise rejection. Does the same benefit apply to differential sending of currents? I am unsure as to the mechanism by which "pick-up" from EM interference affects the signal. I'm interested in sensing signals up to around 100kHz.

\$\endgroup\$
3
\$\begingroup\$

Yes, differential currents help reduce externally coupled noise just like differential voltages do.

External noise sources aren't just voltage or just current sources since they have neither zero nor infinite impedance. See what Thevenin and Norton had to say.

You say you are interested in signals up to 100 kHz, but failed to mention the lowest frequency of interest. If you aren't trying to measure down to DC, then a transformer front end on the receiving amplifier might help a lot. The magnetic field in the transformer is only a function of the current thru the primary. The common mode voltage of the primary is largely ignored, except for whatever capacitive coupling there is between primary and secondary. Some transformers have a shield between primary and secondary for this purpose. You connect the shield to the ground of the circuit that receives the signals from the secondary. Such a setup can have very high input signal common mode rejection ratios.

|improve this answer|||||
\$\endgroup\$
  • \$\begingroup\$ I'm interested in signals above ~10 Hz, and I've got both DC and AC coupled paths (AC coupling achieved with a series capacitor). Good point about the noise - sounds like this approach should work nicely. The tricky part is probably the transimpedance amplifier on the receiving end, where I will have to avoid amplifying voltage and not current. Cheers. \$\endgroup\$ – Sean Apr 10 '17 at 14:10

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.