PVM Reliability Problem

[mikejmoffitt]

I rescued a 14M4U in the worst condition I've ever seen - half the screws holding the tube in are gone (and the screw posts were snapped off), all the internal interconnects for the front panel were just missing, the power button plunger was snapped off, the bottom is bent and cracked...

Well, by some miracle after reconnecting everything (including the ground wires just lying around unconnected inside!), and hardwiring the PSU to be on (the switch AND switch cable are gone), it actually works and looks really good!



Unfortunately, it will seemingly at random go into conniptions and get this vertical deflection jitter. It's not the typical "screen shakes up and down" type, as it only affects a few lines at a time. This shows up as a few lines overlapping, and then a subsequent gap before the vertical deflection gets back on track.

Here's a picture of it misbehaving (very hard to capture on camera!) The "jitter lines" move around, it's not just in the same spot.


It has nothing to do with the source; various consoles, different screens, resolutions - it'll come and go in all situations. I suspect it might be a cold joint.

Just wondering if anyone has seen something like this and has a tip before I go ahead and reflow the deflection section of the board (maybe it's HV regulation too?)

[mikejmoffitt]

Anyone have any ideas? This one's really bumming me out - it's so close to being a great monitor!

[viletim]

The vertical deflection is upset. I'd guess if it's had trauma there's probably a crack in the PCB or a solder joint. Phenolic paper boards can be quite brittle.

It would be faster to troubleshoot the traditional way than to meticulously check for cracks though.

[mikejmoffitt]

I suspect that the problem isn't just the vertical deflection itself, but rather one of the analog values derived from the settings DACs. Specifically, I think it is the "V Center" setting that is having jitter.

The vertical jitter is only ever within a short range (manifesting in overlapping or gapped scanlines). Further, modifying the V-CENTER service parameter, I can make it more upset or less upset:
* At 255, the jitter happens frequently and all over the screen. Even at 255, the screen is still a little too low.
* At 128, it comes and goes, in varying intensities.
* At 0, I have not yet seen it happen. Here, the screen is so low that the current-feedback lines at the top of the screen are visible (they should be hidden in overscan)

I traced the schematics, from the DAC generating the V. CENT control signal, down to this circuit in the deflection area:



I think the problem may be that one resistor biasing one of the transistors is having intermittent contact. It looks like the signal from the DAC provides a return path for the voltage above D535. Unfortunately I can't do much with this schematic, as I've forgotten a lot of my analogue transistor network skills and will have to do some re-learning. I can at least inspect the circuit in question as a start, though.

There is one other problem with the monitor. When I got it, the SCREEN Control was far too high, with a washed out screen and visible retrace lines. I turned it down, and things were fine. However, it seems like the overall brightness (DC offset) of the image varies slightly. I'll adjust the black levels correctly, then look away. When I look back a minute later, it may have shifted to be a little brighter or darker. This is independent of content, and I've seen it happen to just one color channel just one time. This is a bit odd, but I'll investigate this only once I'm done fixing the deflection. I do not believe it is a clamping issue.