[ronbryan]

The original reverse terminator network on my probe was a 50 ohm pellet paralled by a 1k BB resistor trimmer and a 2.2pF 0805 capacitor, which pre-dated the version shown in the circuit diagram in post #1. I have replaced that network with just a small wire ended 51 ohm Electrosil C3 metal oxide resistor, the nearest tiny component I had.

Without the correct termination, as calculated by Jeremy above, I had a look at the termination glitch generated from the reflection, which appears 20ns from the front edge according to the manual. The photo is taken using a Tek 465 100MHz scope connnected to the P6201, 50ns sweep, so one small division on the graticule is 20ns. With this setup, it is nowhere near fast enough to catch any spikes, but it does show a slugged response a little after the transition. Maybe this would have been improved if I could have refitted the 2.2pF speedup cap and trimmed the 51 ohm back to 47 ohms, but it is not easy to add tiny parts in such a small space. (I probably should have looked at the 50MHz pulse generator output direct via a terminated co-ax cable, to check the waveform, but ...)

I corrected the 85mV output offset when DC coupled, by adjustment of a couple of pots in the DC-LF section in the scope end box, but the probe needs at least a half hour warm-up before drift is more or less stabilised.

Ron