Lock-in amplifier with pulsed sinewave

AI Thread Summary
The discussion revolves around using the SR865A lock-in amplifier to measure the frequency of a pulsed sine wave signal, which is only active for 5% of its period. Users suggest that the lock-in amplifier struggles to maintain a lock due to the non-continuous nature of the signal, recommending alternatives like spectrum analyzers or phase-locked loops (PLLs) for better results. Filtering the signal to reduce noise before measurement is emphasized as a crucial step. The conversation also touches on measuring the frequency of ring-down signals, with participants agreeing that a spectrum analyzer is more suitable for such tasks. Overall, the consensus is that the lock-in amplifier may not be the best tool for these specific measurements.
fatima ece
Messages
2
Reaction score
0
Hi,
I am trying to use the SR865A lock-in amplifier to measure the frequency of a pulse of a sine wave( example: waveform below), - the sinewave signal is only ON during finite time not continuously-, I am using an external reference signal that is also a pulse of a sinewave with the same frequency,- the pulse has a frequency of 100 Hz and the sinewave have a frequency of 300kHz, the sine wave is only ON for 5% of the period,- , however, the lock-in amplifier shows the unlock sign -although is shows the right frequency on-screen but lower magnitude-and the unlock sign only deactivated when I have a continuous reference, is there any way to fix this problem? or any other possible solutions or add-ons I can use to measure the frequency of such a signal?
1642620248715.png

-Also, is there a way to measure the frequency of a ring-down signal of a resonator -also a pulsed signal-?

Thanks
 
Engineering news on Phys.org
What about passing it through a high-pass filter to remove some of the low frequency part before you measure it?

Or use a spectrum analyzer.
 
  • Like
Likes Twigg
Why a lock-in amplifier? Normally their value is in measuring the phase information in relatively stable waveforms.
 
  • Like
Likes Twigg
The signal I am trying to measure is small and buried in noise that's why I am using a lock-in amplifier
 
Have you tried a spectrum analyzer? That seems like the "normal" solution to me.

In any case, for extracting signals from noise, filtering is part of the solution. Get rid of as much "noise" as possible before the measurement.
 
fatima ece said:
the sinewave signal is only ON during finite time not continuously
What is the source of this pulsed sinewave signal? Is the source a continuously-running sine oscillator, and some transmit amp is only enabled for those 5% of the time pulses? Or is each time the oscillator starts up a different sinewave? In other words, is the phase constant from pulse to pulse, or is the starting phase random for one pulse to the next?
 
  • Like
Likes Twigg, DaveE and hutchphd
Is this about DTMF demodulation? Or a similar problem with expected discrete frequencies, or is it more of an analog measurement problem?
 
fatima ece said:
Also, is there a way to measure the frequency of a ring-down signal of a resonator -also a pulsed signal-?
Spectrum analyzer if you can trigger it repeatedly. Otherwise for occasional transients, you'll need to capture the waveform and do DSP of some sort.
 
fatima ece said:
Hi,
I am trying to use the SR865A lock-in amplifier to measure the frequency of a pulse of a sine wave( example: waveform below), - the sinewave signal is only ON during finite time not continuously-, I am using an external reference signal that is also a pulse of a sinewave with the same frequency,- the pulse has a frequency of 100 Hz and the sinewave have a frequency of 300kHz, the sine wave is only ON for 5% of the period,- , however, the lock-in amplifier shows the unlock sign -although is shows the right frequency on-screen but lower magnitude-and the unlock sign only deactivated when I have a continuous reference, is there any way to fix this problem? or any other possible solutions or add-ons I can use to measure the frequency of such a signal?
View attachment 295736
-Also, is there a way to measure the frequency of a ring-down signal of a resonator -also a pulsed signal-?

Thanks
A phase locked loop will lock on to a pulsed sine wave. (This is how the colour sub carrier is extracted from the reference burst in colour television). I can't see the point in trying to lock it to a local oscillator when we don't actually know the frequency. Using a PLL is the same as using a narrow BPF. For example, if you passed the waveform through a narrow 300 kHz filter, you would have a continuous signal with not so much amplitude variation, which you could then measure.
 
  • Like
Likes Twigg and Baluncore
  • #10
Condition your reference signal with a narrow band PLL that will make it continuous. Select a phase comparator for the PLL that only integrates when there is reference signal available.
 
  • Like
Likes Twigg, berkeman and DaveE
  • #11
I don't quite understand what is going on here.
This almost sounds like the type of measurements I do all the time, but I wouldn't use a lock-in; it sound like a job for a gated spectrum analyzer .

I had a quick lock at the specs for the SR865 and it does look like you can indeed get it to show the FFT of the input signal. However, in this mode you are not actually using it as a lock-in but as simple digitizer; you are not "locking" to anything so I am not surprised that it doesn't work. So as has already been mentioned this is not the right tool for the job.
If you are interested in measuring the magnitude/noise of the signal (NOT the frequency) then it might work depending on the performance of the lock-in;; you would then use a sine wave (NOT pulsed) as the reference. I don't know how well the SR865 copes with pulsed signals, it will e.g. depend on the time constant and filtering you can use.. Something like a Zurick Instrument HFLI lock-in does this quite well, but is of course more expensive.

Also, I don't quite understand your other question about the frequency of the ring-down. They answer is -once again- that this is not normally something you would use a lock-in for, something like a spectrum analyzer with a trigger or gate input would be more appropriate. That said, it is not the type of measurement one would do very often for the simple reason that the dominant frequency will be exactly the same as the frequency you are applying; usually this is known.
If you are talking about measuring the frequency noise during ring-down, then that is not trivial. A spectrum analyzer would still be your best bet for an initial analysis but there are lots of subtle issues here (frequency vs phase noise etc) so an accurate measurement is not necessarily easy.
 
Last edited:
  • Like
  • Informative
Likes Twigg, berkeman and DaveE
Back
Top