Best Picoscope for Averaging Waveforms

Which product is right for your exact requirements
Post Reply
rmayAZ9558
Newbie
Posts: 1
Joined: Sat Jun 20, 2020 5:07 pm

Best Picoscope for Averaging Waveforms

Post by rmayAZ9558 »

I've read through this forum and found several threads concerning waveform averaging, and it looks a little cumbersome to set up for such a common function. But it wasn't clear from those threads whether it is possible to configure a continuous average/display condition so that the screen plot updates always show an averaged result.

We are looking for a storage scope or waveform digitizer with averaging for a pulsed laser experiment. We fire a 1 ns wide pulse and watch a resulting decay from the process it initiates. The 5000 series appears to have the speed we need (500 MS/s at 12-bits), but it appears the averaging function is implemented by accumulating buffers and then doing a running average on it where the most recent waveform is added to the average and the oldest discarded. Or, it is apparently possible to stop the process when the declared number of buffers is full. But it isn't clear whether this sequence can be repeated continuously, automatically.

What we need to do is average a large number pulses to get enough signal to noise to do the measurement that is needed. The laser rep rate (pulse rate) is 1 MHz (1 ns wide pulse), and at each pulse we collect 200 data points triggered with the pulse start, 2 ns per point. So we collect 400 ns of data per pulse, then have 600 ns before the next pulse event. But we need to average typically 10,000 pulse waveforms (basically a simple exponential decay curve) and then display the result, average another 10,000 pulses and display that result, etc. If necessary, we can slow down the rep rate as needed for the capability of the scope, but we need to execute this sequence continuously.

Can this sort of thing be done with the PicoScope? That is, can we define an averaging scheme where 10,000 (or some similar number) of waveforms are averaged and only the average is displayed, and this process repeats indefinitely? If so, is it possible to save the averaged result if the process is stopped in order to save to a file the most recent averaged result?

Gerry
PICO STAFF
PICO STAFF
Posts: 1145
Joined: Mon Aug 11, 2014 11:14 am

Re: Best Picoscope for Averaging Waveforms

Post by Gerry »

Hi rmayAZ9558,

To answer your Questions:

Yes, by default, you have a continuous updating average waveform trace using a Math Channel, and this averaging is progressively, cumulatively averaged. So, you're correct in that, once the number of waveform buffers that you have decided to capture have been filled, the 1st buffer will be discarded, and over-written with the latest captured data, and then the 2nd, so on (until you stop the capture process) while the averaging will continue as if you have an infinite number of buffers. Alternatively, you can set an alarm to become active once all the buffers have been filled, and then trigger an action of restarting the captures from the beginning. You can also add different types of actions, e.g. sounding/displaying an alarm or saving a file after the buffers are full before restarting the captures.

Capturing the pulse tail is no issue using our PicoScope 6 software, and if you set the trigger correctly and use the appropriate timebase, you should capture at least 200 sample points. Then 10,000 of those (generated as described previously) can generate an average waveform which would be the only thing you would see, if you turn off the input channel display of the captured waveform (which you can do by going to Views->Channels and un-checking the letter of the input channel).

The PicoScope 5000 series typically requires an amount of time to re-arm the trigger, in preparation to capture the next pulse tail, which is >1us, so you may need to slightly slow down the Laser rep rate, but if you have setup an alarm to automatically re-start the capture, the sequence will be executing continuously.

There are just 2 issues that I can see here:

1) The noise level of the PicoScope 5000 series is quite high, and is worse for smaller input ranges. As you're averaging waveforms in order to overcome noise I'm guessing that you're working with low level signals. This may or may not be a problem, depending upon the following:
(a) what is the peak signal voltage that you expect to capture.
(b) how much detail (what resolution, in number-of-bits) do you need in your captured waveform.
If the noise level is a problem then to reduce the additional noise level you could just continuously average over more than 10,000 waveforms, and just manually stop at a point where the signal integrity is good enough.

2) If you manually stop the capture process then the most recent averaged result is likely to be a partial average, of less than 10,000 waveforms because the averaging process will not have reached it's conclusion. To guarantee a fully averaged result as the most recent one you would have to either set the alarm actions to be 'Save file then Restart capture', with the file save taking a considerable amount of capture time to complete, or as mentioned for the first issue you could just continuously average until you get what you need and then just stop the process (especially as you're not interested in the individually captured channel waveforms).

So, a 3rd question I need to ask you is (c) are you attempting to capture a signal that is changing at a rate that is too fast to be using an averaging technique, spread over significantly more than 10,000 generated pulses (if not, why not just average for much longer, to mitigate the 2 issues I raised)?

Regards,

Gerry
Gerry
Technical Specialist

Post Reply