I am trying to obtain the transfer function for a photodiode. Since I couldn't find any standard result online, I figured I would try to obtain it from the circuit equivalent shown below, but I am having some trouble with this. From what I understand, Ri and Ci are dependant on the incoming signal so I wouldn't be able to model this using set values for Ri and Ci. Is this true? Is there a standard result that I just haven't been able to come across?

enter image description here

  • \$\begingroup\$ Ri and Ci do not depend on the incoming signal, but rather the geometry of the photodiode and (possibly) the magnitude of any bias voltage. \$\endgroup\$ Commented Mar 3, 2021 at 1:00

1 Answer 1


Your model seems reasonable, although for most photodiodes Rs and Cp will be very small while Ri will be very large. If the leads on the diode are kept short, usually Ls will be small as well. The reduces to current source driving an RC filter with time constant Rl*Ci, thus a typical photodiode looks like a first order low pass filter.

Note that Rl (the load resistance) is an external resistor in the read out circuit. Thus you can adjust the bandwidth of a photodiode by using a larger or smaller load resistor, at least up until the point where Rl gets large enough that Ri is no longer negligible or small enough that Ls cannot be neglected.

  • \$\begingroup\$ Thank you for this answer. I see that some models incorporate an ideal diode in parallel with the current source. Could you tell me why it is the case that sometimes this could be included in the model and others it's seemingly not necessary? \$\endgroup\$
    – Inf_E
    Commented Mar 3, 2021 at 2:56
  • \$\begingroup\$ @Inf_E A photodiode will act as a diode if you forward bias it, so for completeness you could include the diode. Since you will probably never forward bias a photodiode (which would prevent it from working as a photodetector), often the diode is omitted. \$\endgroup\$ Commented Mar 3, 2021 at 3:00
  • \$\begingroup\$ Ls seems small, until you're trying to receive a signal of several Gbps or more. Then every nanohenry (the effect of a 1 mm bondwire) makes a difference. \$\endgroup\$
    – The Photon
    Commented Mar 3, 2021 at 5:00
  • \$\begingroup\$ How would the current source be modeled if the photodiode is acting as a photodetector for a laser signal? Simply set to some constant amperage? \$\endgroup\$
    – Inf_E
    Commented Mar 3, 2021 at 20:15
  • \$\begingroup\$ @Inf_E It would generate one electron per photon detected. \$\endgroup\$ Commented Mar 3, 2021 at 22:03

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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