tridy wrote:...it's the length of each "high" and "low" that makes it distinguishable from the rest of the noise. So, I don't know how to set a trigger on it. Otherwise, it's less than a second that I am interested it. but there is some delay between when the signal is sent and me pressing the stop button to stop recording.
When you don't really know the characteristics well of the signal you're trying to analyze, then Martyn's suggestion is good, and it will get you the entire transaction on one screen. No cut and paste required.
From there, you can slide over to the region of interest, and zoom in to examine things in more detail. Because of the deep memory on the PicoScope, you'll still have 500 samples per msec, even when capturing a large 20 sec span.
If you then find that the signals are always consistent with your last screen shot example, you can set your trigger to Pulse Width, low, >800 usec. Amplitude anywhere from 1-4V. All the 'noise' on the left is <=400 usec in the low state, followed by a 850 usec low Start pulse. (Perhaps with 5% pre-trigger, if you want to confirm what's happening just before.)
And then set the time-base to ~100 msec/div. That will let you focus exclusively on the ~1 sec you're interested in, and even visualize the data traffic in real-time. You also won't have to try to trigger manually, to catch your burst transmission.
From that point out, your low 'gaps' seem to be consistently 1.2 msec, with a varying number (1,2,3,n) of positive pulses between them.