2
\$\begingroup\$

I'm trying to detect a small AC signal (~10kHz) generated by a phototransistor on a magnitude of 10nA~1uA using the following circuit by converting the signal to voltage signal first and then filter it:

schematic

simulate this circuit – Schematic created using CircuitLab

However, environmental light generates a constant offset of approximately 50uA, which is much greater than my signal. This means that if need to convert the current signal to voltage signal first before filtering, assuming I power the amplifier with a 5V voltage source, R needs to be smaller than 100K Ohm in order to prevent saturation of the amplifier. As a result, the voltage signal would be merely 1~100mV, which is too small for accurate detection due to noise issues. If the 50uA background could be filtered out, I can use a larger R to convert current signal to a larger voltage signal.

In theory, some form of high-pass filter should solve the problem, but I have no idea how to apply a high-pass filter that operates directly on the current signal.

Any hint would be much appreciated! thanks!

Note: phototransistor's collector should be fixed at 0V.

\$\endgroup\$
5
  • 1
    \$\begingroup\$ Welcome to EE.SE! You write phototransistor but your schematic shows a photodiode. If you click on edit and the schematic symbol, a schematic editor will open. \$\endgroup\$
    – winny
    Sep 7, 2020 at 19:46
  • \$\begingroup\$ @winny Thanks for your kind reminder! I've edited my post, however, I cannot find phototransistors in the circuitlab. \$\endgroup\$
    – Wjx
    Sep 7, 2020 at 20:09
  • \$\begingroup\$ I may remember wrong, but can you set the base of an NPN to photo input? \$\endgroup\$
    – winny
    Sep 7, 2020 at 20:13
  • \$\begingroup\$ @winny my phototransistor only provides two pins, a collector, and an emitter. the base is tied internally to the cathode of a photodiode. \$\endgroup\$
    – Wjx
    Sep 8, 2020 at 6:50
  • \$\begingroup\$ Yes, that’s what I’m trying to help you show in CircuitLab :-) \$\endgroup\$
    – winny
    Sep 8, 2020 at 7:03

2 Answers 2

1
\$\begingroup\$

Inspired by @analogsystemsrf, I designed and verified a circuit which can serve this purpose well.

schematic

simulate this circuit – Schematic created using CircuitLab

The key here is to design a stable, second order LPF on the feedback loop. Filters of 3+ orders would possibly cause oscillation so this is the best I could think of. Here current goes in from I_in pin, and the AC voltage goes out at V_out whild DC voltage can be obtained at V_FB.

Hopefully this could be useful for further viewers.

\$\endgroup\$
0
\$\begingroup\$

Sunlight can be removed, with a low_pass feedback.

  • Connect a series R + R network from output back to Vin-.

  • these new resistors are in parallel with your current_to_voltage Resistor

  • we need a very slow time constant, so add 1uF capacitor at midpoint of the 2 new series resistors

  • make both the new resistors of value 1MegOhm.

\$\endgroup\$
2
  • \$\begingroup\$ Hi! Thanks for your reply, I get your point but do you mean that the feedback part should be something like this? ---R---C---R--- in paralell with --------R--------- \$\endgroup\$
    – Wjx
    Sep 8, 2020 at 6:41
  • \$\begingroup\$ I suppose this is a high-pass filter instead? it has lower Z on higher frequency. But your point is helpful, I would check if there's any other way to use another amplifier to generate a LPF on the feedback loop. \$\endgroup\$
    – Wjx
    Sep 8, 2020 at 6:47

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.