I measured a 10 kHz modulated white light using a photodiode (Thorlabs PDA10A):

enter image description here

Unexpectedly, the photodiode output sensitively depends on the orientation and position of the photodiode. For two positions with subtle difference, I got two measurements, as:

enter image description here enter image description here Left: Random phase superposition (unexpected); Right: Stable measurement (expected).

Other observations:

  • This issue is not due to the driver or the signal generator. Everytime I moved the photodiode away from the stable position to a new position, likely there will be random phase superpositions as above left.

  • Also, I have tried changing light source to a 632.8 nm laser diodes, yet same thing happens.

  • If leaving the stable status running for long time (>10 min), the random phase superposition happens again. However I am not sure if this is because of the driver or the photodiode.

  • The photodiode I used, there is a piece of cover glass in front of the sensing area. I am not sure if multi-reflection could cause this effect.


  • Why this random phase issue happens?

  • How can I suppress it?


  • I have used a lens to focus the light onto the photodiode.

  • I cannot reproduce the stable status anymore, even with my previous setup. So something must be wrong with the driver.

  • With an FFT on 4kHz data, we see: (note the waveform itself is not as good as it looks)

enter image description here enter image description here Left: without driver, LED directly connects to signal generator, still phase-random; Right: with driver, and we see 4 kHz, 8 kHz, 12 kHz ... frequencies.

  • For the driver part, I was using analog modulation. Is analog modulation possible to produce such a random phase phenomenon?
  • $\begingroup$ The first thing I would check is to see if there is any stray EM signal that could be affecting your readings. You could, for example, focus the light to a point farther away, use shielded wires, etc. $\endgroup$
    – S. McGrew
    Commented Sep 21, 2019 at 13:55
  • $\begingroup$ Does your scope have an FFT option? Would be interesting to see the frequency spectrum in the two positions. $\endgroup$
    – boyfarrell
    Commented Sep 21, 2019 at 15:46
  • $\begingroup$ @S.McGrew I have changed the setup to your suggestion, but it seems the issue still exists. $\endgroup$
    – WDC
    Commented Sep 21, 2019 at 21:58
  • $\begingroup$ @boyfarrell, please see my edits for the FFTs. it seems there are many sub-frequencies in the waveform. $\endgroup$
    – WDC
    Commented Sep 21, 2019 at 21:58
  • $\begingroup$ Ok. What happens if you reduce the frequency to say 100Hz? What happens if you connect the scope directly to driver? Do you see a clean signal? $\endgroup$
    – boyfarrell
    Commented Sep 21, 2019 at 23:00

1 Answer 1


I have changed the cable (connecting between signal generator and the laser driver) from

enter image description here to enter image description here

Now the output is expected, and does not depend on orientation orposition of the photodiode.

  • $\begingroup$ Glad to see that a coax cable (a shielded wire) solved your problem. $\endgroup$
    – S. McGrew
    Commented Sep 22, 2019 at 3:30

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.