I did find one thing that was causing some erratic behavior on the screen, and that was a switch on the back panel of the scope that had failed so that it was open all the time except for when you were putting pressure on the lever.
I think you're on something.
The switch is labeled Brightness; maximum/normal, and as far as I can tell it just joins or opens a connection between pins 6 and 8 on U4
All of this seems quite strange. It seems this is controlling the flood gun grid (TP3) and it seems that it would inhibit signals which turn off and on the flood guns which I believe may result in the weird things you've been seeing because they are a very important part of the behaviour of this CRT. Can you tell if such a switch is "original" or was added later? its function seems more like a hack or a debug tool rather than something you would really need to use...
Sadly this did not fix all the random behavior, the trace still flashes randomly sometimes.
So you're saying that the ghost image is now flashing rather than being all the time there? I think this may be an improvement.... then you have to trace the timing of the signals which may cause the flickering. Or perhaps it's still the switch which is not connecting properly... honestly I would bypass the switch for the time being.
what is Q10 doing?
https://en.wikipedia.org/wiki/Common_baseEDITED :
U4 is an Emitter Coupled Logic (ECL) gate. In ECL you don't couple gates by voltage but rather by current. TTL is a current-sinking logic since a current must be drawn from inputs to bring them to a logic 0 voltage level. I think by connecting the two gates output together, they are realizing a AND logic gate here.
That transistor I think is (EDIT: besides providing the logic function "AND") a current buffer which interfaces the result to a voltage-coupled device (the CRT flood gun grid) by means of a load resistor (R46 and R47) which would be too hard for the gate to drive directly. Note how the
emittercollector is referenced to -100V supply.
I find it hard to believe that there is a mistake in the HP manual
Mistakes may be much more common than what you may think. I would just really trust the PCB because at some point it must have worked. Looking at the datasheet I see only pin 3 and 4 might have been swapped in the schematic... I don't have the PCB to confirm and trace where the pins go.
U8 and U7 are the only ICs on this board that do not have a HP part number on them, and although they have the same P/N, one is a plastic package and one is a ceramic, leading me to believe that they have been replaced?
Now if you think somebody has had a go, keep you eyes more open than usual around those signals and check them all bu t I wouldn't be surprise if they used different packaging or slightly different ICs at the factory.