And I was having such a good night at the bench...
After doing a bunch of tracing signals and staring at schematics, I isolated the crosstalk problem (an attenuated version of the marker output appearing on the trigger output) to the two emitter followers used to drive the marker and trigger outputs. The transistors are mounted side by side near the front panel, and share a dedicated +12VDC line. It appeared that the "leak" was through the 12VDC. Tracing a little further back, I found that both transistors are driven from another emitter follower.
I decided to duplicate the circuit on a breadboard, using some 2N2222As (of which I have hundreds), and power the breadboard circuit with a separate 12VDC power supply. If the noise was being passed through the 12VDC, I should get a clean signal on my breadboard. Success! Tapping the signal that drives the marker and trigger output (before the emitter follower that drives both), I got a clean copy of the trigger. Taking the signal after that emitter follower and driving just the buffer amp on the breadboard also showed a clean signal. I wanted to see I could take power from the 184, albeit at a different spot, and also get a clean signal.
Next I moved the power from the power supply to the 12VDC line in the 184, right at the onboard emitter follower. The noise was back, albeit at lower levels. Then I took 12VDC back at the power supply. The noise was just about gone. Dividing the circuit in half, I decided to try the 12VDC rail at the countdown board.
I turned off the power, checked the schematic to find the right pin (pin A) connected the jumper from the post on the countdown board, and turned the power on.
It was then I discovered that I had connected my breadboard to the 125VDC line (pin H) rather than the 12VDC line (pin A). With the input of that circuit connected to the output of the trigger generation circuit
before emitter follower.
In addition to blowing out the 2N2222A on the breadboard, I also smoked the onboard emitter follower. It appears the damage goes further back because there's no output on the 0.1us marker now and the lower frequency markers are no longer constant.
Looking at the schematics, worst case I put 125VDC into one or more chains of transistors back to the 10MHz oscillator.
Feeling like a total idjit, I unhooked all the test gear, put everything away, and turned off the lights.
I will figure out what I broke tomorrow.