I don't have a photo of the broken set or the manual yet, but here are the working ones I just fixed. The picture is around 25 inches, and each unit is heavy as all get-out. I stacked the two PVM's and put them on a dolly. This was the only way I could move them around.
.jpg)
On the first power up, the set produced no raster, and R806 started smoking. After some investigation, I found that the HV protection was kicking in, cutting off the convergence circuit. The heater circuit is tapped off T806, which is driven by the convergence output. The set can't produce a raster without current getting to the heater.
811 is the convergence output transistor (SG264). Most waveforms in the set were correct, except the voltages on Q814, Q809, and Q810, which are all in the SW circuit. If the high voltage exceeds a threshold set by R862, R861, and D809, then it will trigger Q814. Q814 is a silicon-controlled switch that acts like a transistor in a switching mode. It stays latched after a positive voltage is applied at its base. Like an SCR, you can get it to go out of conduction by applying a negative pulse at its base. It can also drop out of conduction if the A-K current drops to a low value.
Aside from the smoking resistor, Q814 was saturating around 8 seconds after power-up. This indicated that the fault existed in the circuit highlighted in red. The components in yellow were removed from the circuit and tested. Surprisingly, each component tested good and within tolerance.
At this point, I got distracted with a different project and returned to this a week later. I decided to retest some of the capacitors in the circuit with my Knight R/C Tester. I ended up finding two shorted capacitors (C834 and C822) that the digital meters measured as good. Always test your capacitors at high voltage! I replaced the capacitors, and the result is below. There were still some pincushion problems. I tracked it down to a shorted transistor, Q803, which is the output transistor for the pincushion signal. R806 is the collector resistor for this transistor, so that is probably why it smoked. I don't have any 2SC926A transistors on me, so I ordered a KSP42 transistor to replace it.
