Suppose I have photodiode with a Dark Current of 5 pA and that using the NEP of that photodiode at some frequency \$\lambda\$ I determined that the smallest signal I could detect was of 1 pA.

First of all, does it make sense that my signal is smaller than the Dark Current?, my understanding is that they are additive, so when I have no signal I would see Dark Current \$\pm\$ Noise, but that would always be smaller than 6 pA, thus when I get 6 pA I know I had a real signal.

Now, assuming this is the case, what should I do with my ADC in order to have all my bits carrying information from the signal?.

Again, my understanding is that if I have, for example, 16 bits, and the Dark Current uses 8 of those, then there are only 8 for my signal.

Should I set the ADC to ignore signals smaller than 6 pA?, 11 pA?.

Signal theory is really not my area and I am not sure where to start learning it.


closed as unclear what you're asking by Eugene Sh., Edgar Brown, Dwayne Reid, Sparky256, Bimpelrekkie Feb 12 at 14:32

Please clarify your specific problem or add additional details to highlight exactly what you need. As it's currently written, it’s hard to tell exactly what you're asking. See the How to Ask page for help clarifying this question. If this question can be reworded to fit the rules in the help center, please edit the question.

  • \$\begingroup\$ No hope of working, learn more about communication theory, SNR , BER, signal vs Noise BW, BPF design, Q, group delay, detector jitter vs SNR, hysteresis, etc \$\endgroup\$ – Sunnyskyguy EE75 Feb 5 at 22:49
  • \$\begingroup\$ Also the gap between your knowledge and reality is very large, many of the statements just don't make sense. Signal theory is really not my area that is clear. Start by reading a book and/or following a course. \$\endgroup\$ – Bimpelrekkie Feb 12 at 14:39

To get started, the NEP is the power of the signal that give you an SNR of 1 (not very useful for communication) at a bandwidth of 1 Hz (Damn slow). To do a quick back of the envelope calculation, pick a bandwidth, close to half your desired data rate, and a desired SNR, and get \$P = SNR \cdot NEP \cdot \sqrt{BW}\$. You will probavly get something higher than 1 pA.

First of all, does it make sense that my signal is smaller than the Dark Current? The current you will measure will be the photo-current (your signal) plus the dark current. Be careful, just like the photo-current, the dark current is susceptible to shot noise, and this is ignored by the NEP calculation, so if you are detecting signals of the same order of magnitude as your dark current you should look more closely at that. Also, you will need a trans-impedance amplifier to convert from current to voltage for your ADC. This will come with noise and bandwidth considerations too.

Now, assuming this is the case, what should I do with my ADC in order to have all my bits carrying information from the signal?. Most communication systems use balanced signaling, for example 8bits-10bits encoding, and then use AC-coupling to remove the average of the signal. The signal from your ADC is now symmetrical, around 0V or what reference you used to bias the ADC input. Background light is less of an issue. If you want to stay DC-coupled because of your modulation, for example with Pulse Position Modulation, it's very likely that you will need to estimate the low and high level of your signal with digital signal processing to get a reliable system.

  • \$\begingroup\$ Thanks, first of all I should note that there are conflicting definitions of NEP, the link at the end explains them. The point is, I was using what you define as NEP \$\times BW^{1/2}\$, second, I don't just get values that are bellow my Dark Current, they are a whole order of magnitude bellow. But that is using a SNR of 1, should it be larger?, how large should it be?. Second, I want to know how big the signal should be and how to calculate it. Link to the paper about NEP: thorlabs.com/images/TabImages/… \$\endgroup\$ – Fernando Franco Félix Feb 5 at 23:16
  • \$\begingroup\$ I'm not sure what you are referring to. NEP is in \$W.Hz^{-1/2}\$. Multiply by \$\sqrt{BW}\$ to get \$W\$. See equation (2) in your link. If your bandwidth is anything larger than \$1 Hz\$, \$P\$ is larger than the NEP. \$\endgroup\$ – pserra Feb 5 at 23:21
  • \$\begingroup\$ @FernandoFrancoFélix And yes, the SNR should definitively be larger than 1 is you want an acceptable error rate. You can pick 10 until you start learning about BER. \$\endgroup\$ – pserra Feb 5 at 23:24
  • \$\begingroup\$ what is BER?, I NEED to learn everything about this \$\endgroup\$ – Fernando Franco Félix Feb 5 at 23:28
  • \$\begingroup\$ Well, This is not exactly the place for that. Find a book, or some class (This one for example: ocw.mit.edu/courses/electrical-engineering-and-computer-science/…). I you want to build a quick hobby system, Analog Devices has a nice photodetector wizard you can play with. \$\endgroup\$ – pserra Feb 5 at 23:34

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