Just unboxed my new scope and getting to grips with it - apologies if this is just a setting which can be changed somewhere, but it seems unintuitive and a limitation...
I'm probing a CAN signal - actually one of the training outputs from my MSOX3104A - and before I can decode it, I need to know the bit rate.
I couldn't find an automated measurement that gives this quickly and easily. IIRC it was added in firmware after release on the Agilent, and it's handy. I think it just measures the minimum observed pulse width (either positive or negative) within the captured data.
Feature request 1: bit rate measurement, please!
In the absence of an automatic measurement, I figured I'd just do it the old-fashioned way. Capture a packet, find the narrowest pulse and assume it's 1 bit wide. Then, put the cursors on the edges as accurately as possible, and measure their separation.
The odd behaviour is that the delta-X measurement on screen doesn't show the distance between the cursors if one of them is off screen; instead, it 'clips' to the edge of the screen, and if both are off screen then it just reports the width of the screen, or zero if both cursors are off to the same side.
That's really misleading. It clearly remembers the correct cursor positions if they're pushed off screen, because they come back to the right place if I adjust the time base to bring them back.
If I were feeling unkind, I'd call this a bug, though it's clearly intentional behaviour. I just can't think of any circumstances when I'd want to measure between a cursor and the edge of the display, as opposed to between the two cursors.
Feature request 2: delta-X time measurement actually measures the time between X1 and X2 even if one or both of them is off screen - not between a visible cursor and the edge of the screen.
(Example photos: both cursors on screen, showing delta-X = 8.01us, and the effect of just speeding up the time base 1 click; this pushes X1 off the left side of the screen, and the delta-X readout becomes 6.77us which is effectively meaningless).
I asked about Bit rate measurement long time ago... Will repeat...
In meantime depending on signal being idle high or low, i enable +Pulse Width or +Pulse Width and stats and look for minimum value. 1/t gives frequency i.e baud rate.
But yeah I miss that too, coming from Keysight. Not really needed (you can do what I explained, or just use cursors on visually shortest pulse) but a giant time saver.
Especially if you are doing µC work and doing custom baud rates etc...
As for cursor behaviour it is not a bug.
Scope's capture ENDS with the end of screen.
So it will remember what you set for cursor, because you can magnify the portion of screen to set cursor more accurately, magnify other event set that one, and then st time base to have both cursors on the screen to measure.