AC-Coupled Photodiode TIA Noise

Discussion in 'The Projects Forum' started by RitchRock, Apr 10, 2013.

  1. RitchRock

    Thread Starter New Member

    Apr 10, 2013
    8
    0
    I'm working on an AC-Coupled photodiode amplifier, but am encountering a problem with noise. The photodiode receives IR light, modulated by sound (20-20kHz BW).

    Without coupling capacitor C1, the amplifier becomes very sensitive to saturation. If I couple the light juuust right, it works OK (and sounds good) in this configuration, but since I need high intensity IR for high sensitivity, this configuration is not optimal.

    When I insert the coupling capacitor C1, as shown, the amplifier doesn't saturate, but what sounds like white noise appears. I tried adding the 1G capacitor to ground, as I thought C1 might need a discharge path to ground - no change. I also made a DC servo circuit, which worked well, but the noise stayed the same.

    I've made a little sealed coupling piece for the IR light and photodiode (so no outside light gets in). As soon as the IR light is taken out of the coupler, the noise goes away.

    Any thoughts on where the noise coming from and how I might get rid of it?
     
  2. crutschow

    Expert

    Mar 14, 2008
    13,052
    3,244
    Is it correct that the noise goes away when you remove the light source, with no other change? If so then either you are likely getting noise from the source or shot noise from the detector. Does the noise vary with the IR source intensity?

    What is connected to the IR source?
     
  3. RitchRock

    Thread Starter New Member

    Apr 10, 2013
    8
    0
    Yes, when the light source is removed the noise goes away with no other change. Also, it does vary with intensity.

    The IR source is a high intensity LED. I've tried biasing it both with a simple resistor and also a 2N2222 current source.

    Do you think the light source could be an issue? I had sort of ruled it out, as the circuit with C1 shorted (and no 1G) doesn't make the noise, but saturates easily.
     
  4. #12

    Expert

    Nov 30, 2010
    16,343
    6,828
    This might be too simplistic of an approach, but, R1 sets the gain. If the amplifier saturates, just lower the gain by reducing the resistance of R1.
     
  5. RitchRock

    Thread Starter New Member

    Apr 10, 2013
    8
    0
    I'll try decreasing gain and increasing the light intensity to get a higher SNR.

    I would really like to know why adding that cap increases noise so much though...thanks.
     
  6. crutschow

    Expert

    Mar 14, 2008
    13,052
    3,244
    The reason I think it may be shot noise is that shot noise is proportional to the current through the detector diode. So if the DC value of the current is much higher than the AC value, you will see the shot noise from the DC. You can test that by reducing the light intensity and see if the noise level drops.

    The reason the cap increased the noise is because the cap allows an arbitrarily high level of DC signal without saturating the op amp. This high current increases the shot noise.

    To minimize this effect you need to modulate the light source at a greater modulation index to increase the AC signal level over the DC noise level (S/N ratio).
     
  7. t06afre

    AAC Fanatic!

    May 11, 2009
    5,939
    1,222
    I have some Google search that may help you out. I think your approach may be somewhat wrong. As I think you will get biasing problems
    http://www.google.com/search?q=ac+cuopled+photo+diode
    http://www.google.com/search?q=ac+cuopled+transimpedance+amplifier
    Also if you remove the AC coupling from the first stage sensor stage. And simply put it after. Will that work. Or will you have problems with say saturation. Another thing is that such a design will require proper layout. This is something you do not breadboard. http://www.google.com/search?q=photodiode+amplifier+pcb+layout
     
  8. RitchRock

    Thread Starter New Member

    Apr 10, 2013
    8
    0
    Thanks Crutschow, this makes sense. I am planning on moving to modulation, but I wanted to do one thing at a time. Do you have a recommendation for a phase-accurate, high fidelity modulation scheme? I was thinking SSB, but since the light gets modulated after it leaves the emitter, I don't think that will be possible.

    I also was reading about lock-in to increase snr, but I'm not sure that is possible in my case, as the audio signal is unpredictable.

    to6afre: Saturation is the reason I wanted to AC-Couple. I built the DC servo circuit similar to the google search, but the shot noise was still there.
     
  9. RitchRock

    Thread Starter New Member

    Apr 10, 2013
    8
    0
    RE: Modulation - Would I need some sort of optical splitter / second Photodiode circuit for my reference signal?
     
  10. crutschow

    Expert

    Mar 14, 2008
    13,052
    3,244
    Not much familiar with optical modulation. Perhaps someone else has some info on that.
     
  11. ifixit

    Distinguished Member

    Nov 20, 2008
    639
    108
    Hi RitchRock,

    You're having a problem because the IR carrier is changing levels at a rate that is within bandwith of interest. I.E. 20Hz to 20KHz. The lower end of the band is in the range of amplitude change that might be expected from noise, or normal movement of the IR transmitter. Likely in the 0 to 100Hz range.

    To avoid this problem and get good quality audio you should tranmit on a frequency modulated (FM) carrier. The carrier frequency can then have a varing amplitude due to IR transmitter movement and other low frequency noise that does not effect the audio signal, which is FM.

    A carrier frequency in the 2 to 4mhz range with a 50KHz deviation should do the trick. You can likely buy one off the shelf if you don't want to make one. Of course, your IR devices need to be able to handle those higher frequencies.

    Good Luck,
    Ifixit
     
Loading...