I believe that third one still a bug. Including pictures, input signal is a 1 kHz square wave 50% duty cycle, the hold off is set to 100ns and only by changing the time base (nothing else) some times it will triggers on both falling/rising, notice didn't change anything else other then time base.
Scope Details:
MSO2202A, SW ver 00.03.01, HW ver 2.2
This is very, very interesting. For context, having the effective holdoff "change" as the timebase is manipulated is standard fare for analog oscilloscopes (or at least, the one analog I've played with). The reason is that the holdoff period doesn't start until the horizontal sweep is completed; which means that the distance between triggers is timebase + holdoff. So a change in behaviour due only to a timebase change is completely expected...
Except, in the video in
my earlier post, I demonstrated very explicitly that the Rigol DS2202
doesn't do this. The holdoff is measured from the trigger point, and the selected timebase has absolutely no bearing on the trigger behaviour -- down to the exact nanosecond value of holdoff that flips it between states. So it's extremely strange that your Rigol MSO2202A is behaving more like a Hitachi 20MHz Analog CRO than a Rigol DS2202.
To confirm exactly how it's behaving, could you please show the behaviour with the following 6 holdoff settings:
200us/div, holdoff = about 0.3ms (that's 300us, or 300k ns)
200us/div, holdoff = about 0.7ms
200us/div, holdoff = about 1ms
And the same three holdoffs, but with 500 us/div.
That way, we can see if it's just the timebase futzing with the holdoff settings, or something more broken.