Thank you everyone for some good questions, I believe I've addressed most of them below.
I'd prefer to stay away from a shunt to keep wiring easier (no need to add a shunt inline), but from a price stand point it may be worth me reconsidering this. I hadn't considered it initially so I will have to now.
Ignore the people who only read the title and thought "He wants a $3K oscilloscope!".
With those requirements you need an entry level oscilloscope and some very fancy probes.
Very good point, I should have been more explicit in my title/OP. Thank you for pointing this out.
OP wanted to measure current transients (20khz current clamp was also mentioned)
I do want to measure transients, but I don't expect they will be in the 20khz range, rather that the clamp I'd initially looked at should be sufficient to view transients as its response is (relatively) quick.
- do you really want to spoil the accuracy of a normal current clamp (1%) by the gain error of a 8 bit scope (3%).
Would prefer to get above 8 bits if possible. The picoscope 4444 accomplishes that.
- how many sample memory is necessary e.g. for measuring a complete run up phase of the motor including (parallel) CAN message sampling.
Hard to know how long of a sample is needed, longer is better
A few seconds should be sufficient.
- are pass/fail masks necessary for the production
No plans for this, although the fact the picoscope has it doesn't hurt
- is the scope only needed for viewing or is the measured data (complete or screen shots) to be stored on a PC.
Would like the ability to export to CSV or similar for post-processing in excel
- is remote control for the scope needed (e.g. by remote desktop)
No immediate needs or plans for this
- how often is the zoom function needed (and how easy can it be handled).
Zoom is important, I'm not sure if a knob or a mouse will be easier for me, but I'm comfortable with both.
Do you really need to know the current of a 200A motor to 3 decimal places?
No, to 1's place is fine in all honesty...