Oscilloscope software captures only the last part of the waveforms

Thread Starter

RichW

Joined Nov 8, 2019
40
It seems I'm doing things the hard way. I now have an interesting problem where the oscilloscope software saves only the last 0.5ms of a 10 ms (on screen) waveform. So now I have 100's of files of only the last segment that is displayed on the screen. Has anyone ever seen this before?

The only thing I can think is that it is capturing/saving the waveform from right to left (I have no idea why as I didn't tell it to do this) and then moving on to the next waveform after saving. I'm going to try to increase the sample frequency (to 2 MHz since I am working at 1 MHz) and find the option to save left to right from trigger (if there is one), should this solve the problem?

I'm interested to know if anyone here has seen this before and how they solved the issue.

Thanks again for any advice!
 

Danko

Joined Nov 22, 2017
1,829
oscilloscope software saves only the last 0.5ms of a 10 ms (on screen) waveform....
If your oscilloscope has only 1 kSa internal memory stack
and sample frequency is 1 MSa/sec
and 2 channels are used,
then you have saved only last 0.5 ms of displayed waveform.
I'm going to try to increase the sample frequency (to 2 MHz since I am working at 1 MHz) and find the option to save left to right from trigger (if there is one), should this solve the problem?
At 2 MSa/sec you will have saved only 0.25 ms of waveform.

Which oscilloscope is using?
 
Last edited:

Thread Starter

RichW

Joined Nov 8, 2019
40
If your oscilloscope has only 1 kSa internal memory stack
and sample frequency is 1 MSa/sec
and 2 channels are used,
then you have saved only last 0.5 ms of displayed waveform.

At 2 MSa/sec you will have saved only 0.25 ms of waveform.
Awesome, I understand now, thanks!

Still not sure why it recorded right to left though, if it had recorded left to right then I would have at least saved some of my day. But, we live and learn!

Which oscilloscope is using?
It's hooked up to a PC through a Handyscope HS6, I switched from PicoScope for faster record times. Seemed to be faster but didn't realise I hadn't set the correct sample frequency. I'm guessing when I increase the sample rate the recording speed per capture is going to significantly decrease. I have read that the frequency much be set at least two times higher than the maximum input frequency (the Nyquist frequency). So, would 2.2 MHz be a good sampling rate for 1.1 MHz to avoid aliasing? Or should it be set even higher for more accurate measurement.

Is it best to set the internal memory to it's maximum and reduce the sample frequency, or go for an intermediate value and increase the sample frequency? Or just increase both to get the 10 ms waveform that is displayed on screen?

Thanks again.
 
Last edited:

rsjsouza

Joined Apr 21, 2014
383
So, would 2.2 MHz be a good sampling rate for 1.1 MHz to avoid aliasing? Or should it be set even higher for more accurate measurement.
That depends on the waveform. A sinusoidal can be reproduced with about 2x the fundamental frequency. However, if you have either distortions or a different waveform (triangle, square), then you will need a much higher sampling frequency, as its components will reach much higher frequencies.

Is it best to set the internal memory to it's maximum and reduce the sample frequency, or go for an intermediate value and increase the sample frequency? Or just increase both to get the 10 ms waveform that is displayed on screen?
I would start by adjusting the sampling frequency to meet the Nyquist condition of the higher harmonic of your signal and then set the maximum available memory that the oscilloscope can display at that given sampling frequency. Some oscilloscopes put a limit to that, but I don't know this brand to be absolutely certain.
 

Thread Starter

RichW

Joined Nov 8, 2019
40
That depends on the waveform. A sinusoidal can be reproduced with about 2x the fundamental frequency. However, if you have either distortions or a different waveform (triangle, square), then you will need a much higher sampling frequency, as its components will reach much higher frequencies.
That's a very good point, thank you! Yes, I am also looking for higher (and lower) harmonics of the driving frequency (f, 2f, 3f, 0.5f, 1.5f. 3.5f etc etc.). So, I guess I will need to adjust the sample rate even higher. Perhaps I'll reduce the number of samples and record signal from the left this time, if this is how this thing works.

I would start by adjusting the sampling frequency to meet the Nyquist condition of the higher harmonic of your signal and then set the maximum available memory that the oscilloscope can display at that given sampling frequency. Some oscilloscopes put a limit to that, but I don't know this brand to be absolutely certain.
It seems with this software, when you adjust the time base it also adjusts the sample rate and number of samples automatically. So setting say 500 us/div automatically changes them to 20 MHz and 100 kSa. If I change either of those last two things it changes the time base (and vice versa).

So anyway, I looked at the "data collector" options and there was an option to change the 'data size', I changed this to 100 kSa to match the main displayed number of captures and it works. I must have been capturing only a small amount of data.

Anyway, all sorted now, thanks for the help. Now I need to repeat what I did yesterday... I had perfect results and it wont repeat damn it! lol

Many thanks!
 

Phil-S

Joined Dec 4, 2015
238
Picoscope software has provision to set sampling rate etc.
A bit off topic, but I get the same problem with serial monitoring and terminal emulators like CoolTerm.
If I don't set the buffer size every time, it rapidly fails.
The Arduino IDE serial monitor does the same and it usually ends up restarting the PC as it consumes CPU and RAM capacity.
 
Top