2

I want to measure the relative intensity noise between 0 and 1 MHz of a laser using a photodiode (20 MHz BW), a voltage amplifier (10 MHz BW), and a spectrum analyzer. Since I don't have so much knowledge about electronics, I would like to clarify some things and like to know what the best "setup" is.

  1. In order to not crash the spectrum analyzer (0 V DC allowed at input) I need a DC block. Do I put the DC block before or after the voltage amplifier?

  2. The voltage amplifier can be DC- and AC-coupled. Which is the best setting?

  3. In order to suppress the repetition rate of the laser (50 MHz) I need to use a 11 MHz lowpass filter. Do I place this before or after the voltage amplifier?

  4. Where do I need to put 50 Ohm terminators? I suppose after the photodiode to convert the photocurrent into a voltage. But do I also need one after the amplifier and before the spectrum analyzer?

Basically I want to have as little electronic noise as possible from the setup to precisely measure the laser noise and not be determined by the electronic noise itself.

Photodiode: DET05D2 from Thorlabs

Amplifier: HVA-10M-60-B from Femto

Spectrum analyzer: E4443A from Agilent

user302220
  • 21
  • 3
  • Do you have a spec for sensitivity? Depending on how sensitive you need to be and how bright the laser you may not even need an amplifier. Otherwise, answers will heavily depend on the choice of photodiode and amplifier, so you need to provide that information. – user1850479 Dec 15 '21 at 15:12
  • Thank you for your answer. I need to attenuate the laser power in order not to saturate the photodiode. Therefore I need an amplifier to see anything on the spectrum analyzer. I put the specs of the diode and the amplifier above :) – user302220 Dec 15 '21 at 15:16
  • 1
    AC couple the amplifier and put the low pass after the amp. You don't need to add termination, it's already built in. The gain on that amp is very large, which may be too much for your application unless the RIN is low on the laser. Do you have some idea how big of a signal you want to measure? – user1850479 Dec 15 '21 at 15:29
  • Thank you very much for your reply. If I AC couple the amp I don't need the DC block anymore? So I also don't need termination after the diode because the amp already has it? Yes, the RIN of the laser is really really low, that's why the large amplification is needed :) The RIN should be on the order of 0.01%. – user302220 Dec 15 '21 at 15:31
  • Also in the spec sheet of the amp it says "output impedence 50 Ω (terminate with 50 Ω load for best performance)". Does this mean I need an additional 50 Ohm or has the SA one? Also RIN should be around -130dBc/Hz if it helps. – user302220 Dec 15 '21 at 15:38
  • You mention the "repetition rate" of the laser, but don't say what that rate is. How much separation do you have between your measurement band (0-10 MHz) and the laser's pulse frequency? – The Photon Dec 15 '21 at 16:54
  • The rep-rate of the laser is 50MHz and the noise I want to measure is from 0 to 1MHz. – user302220 Dec 15 '21 at 17:20
  • You can skip the low pass filter then. I would also try it with just a ~10uF ceramic capacitor in place of the opamp (no gain). – user1850479 Dec 15 '21 at 20:59

1 Answers1

2

In order to not crash the spectrum analyzer (0 V DC allowed at input) I need a DC block. Do I put the DC block before or after the voltage amplifier?

You place it where it will guarantee 0 V DC at the input to the spectrum analyzer. The only location that does that is immediately before the input to the spectrum analyzer.

Depending on the design of the amplifier, you might also need one at the input to the amplifier.

The voltage amplifier can be DC- and AC-coupled. Which is the best setting?

DC if you use an external DC-block in front of the amplifier. AC otherwise.

In order to suppress the repetition rate of the laser I need to use a 11 MHz lowpass filter. Do I place this before or after the voltage amplifier?

This might be misguided. Better might be to let the pulsed signal through to the spectrum analyzer, and then measure the RIN in the side-bands of the pulse peaks.

Where do I need to put 50 Ohm terminators? I suppose after the photodiode to convert the photocurrent into a voltage. But do I also need one after the amplifier and before the spectrum analyzer?

Your amplifier already provides a 50 ohm input impedance, so you don't need an external termination after the photodiode. It's not clear from the description on the Thor Labs website, but you may need to provide an external DC path to ground for the photodiode to operate properly. Using a resistance higher than 50 ohms for this will improve the gain of the system (but too high a resistor will introduce Johnson noise that might hide the laser RIN).

Your spectrum analyzer already provides a 50 ohm input impedance, so you don't need an external termination in front of the spectrum analyzer.

All that said, a RIN measurement is not an easy one to make correctly. It's easy to add enough additional noise in the amplifier stage to hide your laser noise. Also be very careful of back-reflections from the measurement system returning into the laser's output aperture --- these can dramatically affect the laser RIN (RIN is usually specified with some controlled back-reflection level to account for this).

The Photon
  • 126,425
  • 3
  • 159
  • 304
  • Thank you very much for the extensive answer. It seems that the AC coupling of the Femto amplifier introduces an additional bump in the spectrum. That's why DC coupling is the best to use. Since the theoretical maximum voltage after amplification is close the rating of the DC-block and the lowpassfilter I need to place them before the amplification. Does this make sense? – user302220 Dec 15 '21 at 17:16
  • @user302220, are you sure the amplifier doesn't produce a DC offset in its output? – The Photon Dec 15 '21 at 17:33
  • The website says "TRUE DC-COUPLING WITH ZERO OUTPUT OFFSET"... So I assume if I have a DC-block before, I don't need one afterwards. How would you couple now all the components after one another? – user302220 Dec 15 '21 at 17:51
  • Also the DC block is intended to be terminated with 100kOhm or more... Right now it is just directly plugged into the DC-coupled amplifier which has a 50Ohm load. Can this cause problems? It is necessary to put another resistance between the DC-block and the amp? – user302220 Dec 15 '21 at 19:25