That is exactly the way it works for the calibration.
You need to do this for every input range you are using, because different input ranges have different offsets.

Benno

You solder it together.
2 BNC cables with 1 end open.
1 cable the inner core to a 4mm banana to the CAN+ pin
and the outer shielding to the signal ground pin
2nd cable inner core to a 4mm banana to the CAN- pin
and the outer shielding to the signal ground pin (same as the 1st cable)

Benno

Be aware that the voltage used to power the picoscope can be different (motst that I know use 5V DC)
If you put a 12V barrel in int you will blow up the scope.

Benno

it is the power loss of the long cable, if you have an option to power the scope seperatly then it can work, although during startup the scope still uses the usb power.

Benno

Tested this with PS7 and it is the time at the end of the traced.
so when the whole trace is completed. (simple measure if you set timebase to 5 sec / div)

Benno

What threshold levels have you used for the digital inputs ?
Is your GND grounding solid ?

You can also use 1 analog input on 1 of the digital inputs with strange spikes so you can see what the analog signal does when the spikes appear.

Benno

Looks like your start capacitor is turned back on after about 13 min running.
The currents during the tripping are exactly the same as the startup currents.
So I suspect the relais that connects the startup capacitor to the 3th phase.

Benno

The protocol still is CAN the serial RS232 is embedded in the can messages as normal data.
Baudrate looks about 1,3 Mbaud.

Benno

You can look at YouTube

https://www.youtube.com/results?search_ ... alibration+

https://www.youtube.com/results?search_ ... mpensation

Benno

Got message file is corrupt. (more like encrypted or something like "%TSD-Header-" at the start if the file.

Can you retry uploading it to we transfer.

Benno

I have made a comparison between different methods of current measurement.
See attached .psdata file. Measured on a 1 ms pulse of 11A DC rising and falling edges are of interest for you.

Benno

can you post a .psdata via https://wetransfer.com/ Also the current probe works with hall sensors and thye have some opamps in it to amplify the output of the hall sensor, they usually have a bandwidth of around 10kHz so a delay of some sort is to expect from this kind of sensor. If possible you can...

Do you have low pass filtering on in the settings ?

Benno

Can you post a .psdata file.
And what channels are measuring what ?

Benno

The only way I know is to use reference waveforms.

Benno