Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

What is photodetector bandwidth?

  1. Aug 1, 2013 #1
    Hi,
    I would like to know how a detectors bandwidth can be measured? for example, if there is any relation with the produced current or what? I can't understand bandwidth concept.
    Thank you.
     
  2. jcsd
  3. Aug 1, 2013 #2
    Before you can measure the bandwidth, you must understand what you are measuring. First you need to decide which specific detector you are going to use.

    What is it you don't understand? How were you thinking of measuring the bandwidth?
     
  4. Aug 1, 2013 #3

    davenn

    User Avatar
    Science Advisor
    Gold Member

    you would need to look at the manufacturers data sheet for a specific photodetector

    Dave
     
  5. Aug 1, 2013 #4

    berkeman

    User Avatar

    Staff: Mentor

    As already asked, what specific detectors are you looking to use? Are you going to use a pre-packaged module, or will you build it yourself with a PIN photodiode and amplifier? If you are building your own, be sure to use a good reverse bias voltage on the PIN photodiode to help get the best bandwidth out of the circuit.
     
  6. Aug 2, 2013 #5
    It is a PIN photodiode and I am in simulation level. not any experimental work yet. Is it possible to measure bandwidth in simulation level or you need to build and measure it?
     
  7. Aug 2, 2013 #6
  8. Aug 2, 2013 #7

    berkeman

    User Avatar

    Staff: Mentor

  9. Aug 2, 2013 #8

    dlgoff

    User Avatar
    Science Advisor
    Gold Member

    I'm wondering if that's at an operating temperature of -40ºC. :bugeye:
     
  10. Aug 2, 2013 #9

    sophiecentaur

    User Avatar
    Science Advisor
    Gold Member

    In principle, it's no different from measuring the bandwidth of any receiver. You need to produce a modulated signal from a test source and then see what your detector (demodulator) gives you. The wider the bandwidth of the receiver, the more data or the faster the analogue signal it can detect. (Is that what you had in mind for your question - sorry if it's talking down but it's hard to assess your initial level of knowledge.
    An obvious (?) way would be to amplitude modulate light of a suitable wavelength and see how the shape of the modulating signal waveform is affected - e.g. look for the effect on a very fast pulse. Alternatively, use a swept frequency for modulation and see the amplitude of the signal that your detector demodulates at different modulating frequencies.

    Btw, thw signal bandwidth is distinct from the "spectral bandwidth" which is the range of wavelengths that the detector can actually detect - see the spec sheet in the link.
     
  11. Aug 2, 2013 #10

    Thanks for your explanation, Sophie. I know something about bandwidth but I am concerned about any theoretical solution for measuring bandwidth, if any. i.e. how is it possible to measure the bandwidth if I have the current (I output) results for a photodetector?
     
  12. Aug 2, 2013 #11

    sophiecentaur

    User Avatar
    Science Advisor
    Gold Member

    Like I said before, you need to provide the detector with a modulated input signal, swept or impulse. You then look at the Ioutput waveform and see how the system has modified it. This is what you'd do to measure the bandwidth of any channel. Of course, a single figure for 'bandwidth' would refer to the 3dB bandwidth (or whatever else you care to specify) but that, in itself, is not enough to char
    acterise the system fully. How do you want to specify the performance? What parameters are relevant to your purpose?
    Actually, I am not sure what you actually mean by your question. What is your level of knowledge of signalling channels in general? What do you understand the term 'bandwidth' to mean, in the context of this application? It's a word that has at least three connotations.
     
  13. Aug 4, 2013 #12

    meBigGuy

    User Avatar
    Gold Member

    Given a spice model for a photodetector, you can connect it to an amplifier model, include estimates of parasitic (pcb?) capacitance and measure the system response (or just measure the photodetector current to begin). The model will have a calibrated way to provide stimulus. You sweep the frequency of the input, and see how the amplifier or current output varies as the frequency increases. The 3dB bandwidth is simply the frequency at which the output drops 3dB.
     
  14. Aug 4, 2013 #13

    sophiecentaur

    User Avatar
    Science Advisor
    Gold Member

    Isn't spice a simulation? How does that help with actually measuring the characteristics of a real detector? Have I missed something about Spice? Is there a hardware version?
    3dB bandwidth is a partial solution to finding the 'usefulness' of the device (for instance, the noise bandwidth). It's fine if you can treat it as a single pole low pass filter but, more detail would be needed if you want to include it in a channel with shared Tx Rx filtering (a common requirement).
     
  15. Aug 4, 2013 #14
    Interesting question. Bandwidth of photodiodes depends on things like material, doping levels, and physical size (physically larger = lower bandwidth). You would need to extract equivalent circuit from these physical characteristics.

    You may want to do a literature search of this. Are you in a university environment where you have access to journals?

    I have attached an example, Let us know what you find.
     

    Attached Files:

  16. Aug 4, 2013 #15

    meBigGuy

    User Avatar
    Gold Member

    The OP said "It is a PIN photodiode and I am in simulation level. not any experimental work yet. "

    One of the first steps to understanding the actual bandwidth would be a simple frequency sweep in spice. No argument that more analysis is appropriate, as are more complete models. 1 step at a time.
     
  17. Aug 4, 2013 #16
    Yea, we are all working on very little information from the OP. I am assuming he knows how to drive his simulator (whatever it is) and needs to know how to develop a model for his photodiode.

    Maybe I am wrong and OP needs tutorial on spice.

    @OP: more info please. What exactly do you mean by "I am in simulation level".
     
    Last edited: Aug 4, 2013
  18. Aug 4, 2013 #17
    If you want a rough idea of the bandwidth, the easiest way (and it works great in simulation) is to put a sharp pulse into the detector and measure the rise time of the response. If it is close to first-order then the bandwidth is about 0.35 over the rise time of the detector. (so if the 10%-90% rise time of the detector is 1 ms, the BW is about 0.35/1ms or about 350 Hz.)
     
  19. Aug 7, 2013 #18

    Njorl

    User Avatar
    Science Advisor

    I'll assume your simulator is a pencil and paper. Calculate the capacitance of your detector. Multiply by the load resistance. Take the inverse. That's a decent approximation of your bandwidth. You could use the intrinsic resistance built into your detector instead of load resistance, but nobody ever does that.

    The capacitance calculation is the only difficult part. You can treat a mesa detector as a parallel plate capacitor - area is the size of the mesa, separation is the thickness of the undoped region, don't forget the dielectric isn't a vacuum. Interdigital finger detectors are a more complicated calculation.
     
  20. Aug 7, 2013 #19

    sophiecentaur

    User Avatar
    Science Advisor
    Gold Member

    What I don't understand is how, if you are doing a simulation, someone hasn't already defined the frequency response of the photodetector.!?!?!?! Else, how can it have been simulated?
     
  21. Aug 7, 2013 #20

    berkeman

    User Avatar

    Staff: Mentor

    For the OP -- I did a google search on Modeling A PIN Photodiode In Spice, and got a lot of good hits. Try the same search if you'd like to read through the list.
     
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook




Similar Discussions: What is photodetector bandwidth?
  1. Bandwidth problem (Replies: 2)

  2. PIN Photodetector (Replies: 6)

  3. Bandwidth and bitrate (Replies: 1)

Loading...