How Can I Convince My Advisor About Measuring the DC Component of Our Signal?

[Backpacker]

Ok, so I am just an undergraduate intern this summer.

We have light source shining through an optical chopper, reflecting of a lit display, and going to a detector. So, we have a signal on top of a DC signal coming from the fact that the display is also emitting light. We want to measure both the DC signal and the chopped signal and compare them. The signal from the optical chopper is our reference for measuring the chopped signal.

My advisor is convinced that we should be able use a second lock-in amplifier to measure the DC component of our signal by somehow using the same chopping frequency as a reference signal. He won't really listen to me, but I am convinced otherwise (that we need more than just another lock-in). (Essentially, he believes that we should just be able to look 180 degs phase shifted from the reference and be able to see just the DC signal. I've tried telling him that that is not how a lock-in works).

He wants me to call tech support for Stanford Research (the makers of our lock-ins) to ask them for advice. It seems silly.

So far, I've thought of a few other ways to accomplish the problem:
-split the signal from the detector, send half to the lock-in and the other half through a low-pass filter
-split the signal, send half to the lock-in and mix the other half with a signal generated by a frequency generator, then send that to another lock-in (using the generated frequency as the reference)

He has pretty much dismissed both of these ideas, thinking that his solution is more elegant. What should I do?

Thanks!

 

[Manchot]

I think that your advisor is somewhat correct. If your lock-in is anything like the SRS lock-ins that I have dealt with, it has an output on the back called the "PSD Monitor." This output is the signal after it has been filtered and demodulated (i.e., multiplied by the reference), but before it has been amplified and filtered again. So, if you look at this signal, the DC signal will now be located at your chopping frequency, and the chopped signal will be located at DC. You could then lock on to the PSD signal to find the original DC signal.

However, there are a couple problems with this. The first one is the main filter on your first lock-in. Depending on how you configure it, you will attenuate your DC signal in a chopping frequency-dependent way. Nonetheless, I would think that you could fairly easily account for this through some sort of calibration. The bigger problem is the harmonics of the chopper. The demodulation process is just a multiplication, and so not only will the DC signal contribute to the chopping frequency at the PSD monitor, but the second harmonic will as well.

There may be other options, as well. What are the relative magnitudes of the DC signal and the chopped signal (e.g., is one much larger than the other)?

 

[Backpacker]

Hey,

So I've talked to my advosor, and he more clearly described the problem that he thinks we will have.

Look at the attached picture. He wants to measure the blue line, but fears that if we simply measure the DC signal that it will give us the red line.

I told him that I don't think it should be a problem because the DC signal is much bigger than the chopped signal (although we don't know exactly how much bigger because we haven't made any measurements yet). But he wants me to figure out if there is some way to measure the blue line using what tools we have.

Peace!

 

[Manchot]

Is there any reason you can't just block the chopped signal and measure the DC value directly?