Hi,
Will doing some testings today, I realized that when measuring regulated voltages, from time to time the picoscope measures big negative spikes (sometimes even exceeding the -20V).
I'm using a Picoscope 2405A. I put a trigger in single or repeat mode at 3V rising. About twice a minute, I'm getting a descending spike that triggers (often it goes bellow the 20V).
I tried the following things:
-disconnect the circuit I was testing (and just connect the probes directly to the output of the lab power supply). Note that I kept the circuit disconnected for all the other testings
- change power supply (3 different ones tested)
- switch the probe used (among the 2 supplied with the picoscope)
-switch the input channel (A or B) used
- change the supply voltages (5V, 8.5V, 11V)
-change calibers (+- 20V and +-10V)
None of those tests changed anything. So the only reason I can still think of is a problem of the picoscope itself.
What do you think?
Any Idea how to solve this problem or what else might be the reason of it?
Thank you very much in advance
Best
Felix
PS : I attached the screenshot of one of those spikes
Could you try capturing the same voltage on a +/-20V input range, at 5ms/div. If the sample rate is fast enough to capture the impulses then we will be able to see the full range of them and whether or not they are periodic (at mains frequency). Could you then post the actual data file (rather than an image of it).
Hi,
Thanks for your answer.
I did the test again (at 11.1V according to the power supply).
I first used 5ms/div as you suggested. As one get only a single spike, I tried 5s/div (2 spikes, but it isn't enough to see if they are regular) and then finally I tried 20s/div, where I got 8 spikes, without a pattern.
How can I increase the number of samples above 49k? For my model (Picoscope 2405A), the buffer size is 48kS. If I put a number of samples bellow 48k, I get approximately the number of samples I ask for, but when I ask for something bigger (then I always get around 49k samples). When I go for slow time/div, then I start getting the number of samples I ask for even when above 49k.
Note that I'm using the linux version of picoscope 6 (downloaded a few month ago), and that the picoscope is connected on a USB3 port.
So for the 5ms/div, I didn't managed to get anything new.
I attach the data for the 1ms/div (but with only 49k samples).
You can't increase the number of samples (I did check on your buffer memory, that's why I mentioned "if the sample rate is fast enough to capture the impulses", and why I didn't suggest a longer Time period, which would be better if you had enough samples).
There's not much point in having too long a capture period because too many impulses will be missed. If the impulses are periodically related then it looks like you could already be missing one in the capture you performed. So, it would be worth trying again with 5ms/div, but setting the maximum number of waveforms to 100 (in Tools->Waveform buffer) so that we have much larger statistical sample of data (increased probability of capturing periodically related impulses) that we can look through using the waveform buffer.
It would also be worth generating a 50Hz squarewave from the signal generator, feeding that into channel A, to generate a repeat trigger, and then feeding your voltage into Channel B, then switching to persistence mode and letting it build up a time averaged capture. You will have to get a screenshot of the persistence mode capture to post as you can't save the persistence data.
Hi,
For the first measurements you asked, if I understand well, I just run it with no trigger. I added the run as I understood it (5ms/div, no trigger), but just decided to increasse the number of waveforms to 1000 instead of 100 because otherwise I got about 2 spikes. I put the file on the drive : https://drive.google.com/drive/folders/ ... sp=sharing (5msDiv1000waveformsNoTrigger.psdata).
For the repeat mode, I have the impression that the scope isn't superimposing enough waveforms to get a good result, and I haven't found out how to increase it. I still attach a couple of screenshots. One interesting thing is that both channels seem to spike simultaneously.
No, you have to run with a Repeat Trigger (triggering off of the impulse) so that at least one impulse is always being captured. With Repeat trigger many full buffers of data will be captured and sent to the PC (i.e. it's single triggers just repeated many times). You don't need to capture 1000's just a 100 should be enough.
For the Persistence mode plots, could you run one again, for a much longer time, (and turn the decay intensity up to maximum - in the Persistence Options - which is the icon next to Persistence Mode) and then give us a screenshot.
Did you manage to work out what was causing this? I'm also measuring very similar negative voltage spikes with a picoscope 2405a, from several different sources and using different laptops, which makes me think I have a problem with the picoscope itself.