OK... I set the switch while in real-time to Smooth and the display looked just like the very slow rise/fall. I suspect the Normal/Smooth switch contacts were dirty. I have exercised them a few times. So that would explain the "fix" :)
Meanwhile, I shorted the coupling cap (a .01 uf disc) and ran it to only one sampling head without the tee. It looks a lot better now. Photos added at 100 ns/div (one full cycle at 1 MHz) and at 2 ns/div.
However, I cannot get the pulses to display on 10 ns/div or faster. The timebase will trigger (trace onscreen, flat at zero) but absolutely no edges regardless of the setting of Time Position (or B Delay)! I would have thought there is something defective in the 3T77A on the faster ranges - but it works with the 600 MHz sinewave input all the way to the fastest range of 0.2 ns/div.
I also thought there might be a significant delay from trigger to display since the 3S2 has no delay line. So I connected the TD pulser directly to the Ext Trig input with the tee, then ran ~20 feet of GR cables to the sampler input. Not surprisingly, the displayed edge (with reflections) is now delayed by 30 ns or so... but it's still invisible at 10 ns/div or faster.
What's going on??
