Nanao (Eizo) KZ-20EN Monitor
What can one say... but "AARRRRGHHH!!". This is the unluckiest piece of hardware I've ever dealt with, even worse than Golden Square computers. I'm not sure if this monitor comes from Nanao or from Kaga Denshi; all the units I have are labeled by both companies and I suspect that one is a subsidiary of the other. The CRT (it's a Mitsubishi tube) carries a white label which reads "KAGA KZ-20EN-x" where x is a revision level I think; I have rev. B and rev. L monitors. The flyback coil is protected by a metal housing with a Nanao label and the same model number.
My first experience was with a revision L monitor. First thing I noticed was that the last set of permanent magnets on the CRT neck (used to adjust convergence) is not secured and the last two magnets are free to drift around. Random quality control problem, you say? Tampering by a previous owner, maybe? No, sir - I now have four of these units and they all have this fault. Anyway, I hooked up a game board and adjusted the convergence, and put a drop of hot-melt glue on the magnets to keep them in position. Problem sort of fixed.
Next fault with one of these monitors was an intermittent fade-to-black (loss of CRT heater voltage).Wiggling the neckboard fixed this, so I thought "no problem" - dry joint or dodgy contact. As it turned out, the poor contact was in the CRT socket, and the reason for it was that the board had been sealed on with one of those locking compounds like Lok-Tite®, and the holes in the socket were clogged up with the $#&@ stuff! After some cautious experimenting with solvents I happened to have lurking around the place, I found one that would loosen it a little, and I used a tiny rat-tail file to clean out the remains. Problem fixed. But who did this and WHY?
Problem child #3, a revision B in a Legend of Hero TONMA cabinet. The symptom was an intermittent vertical shrinkage, and oscillation between shrunk and normal size. My thought was a flaky vertical drive transistor. This is a 2SD1138, and not something I keep in my junkbox, particularly since they're about AUD$16ea. I pulled apart a couple of junkable TV sets looking for one of these transistors; no luck (though I did find a lot of interesting/useful bits and pieces which I'm sure I will use someday). So just on the offchance that it was a dodgy connection, I reseated all the connectors. Result: No picture, no high voltage, not a sausage, next time I powered it up. It was getting 100VAC from the isolation transformer, so the fault must lie on the main monitor PCB. Great progress! It was late, so I went to sleep.
The next morning, I awoke in the mood for violence. In arcade cabinets, the monitor chassis is generally screwed to one side of the (wooden) cabinet and it's impossible to service without removing it and disconnecting half a dozen cables. I pulled off the PCB and got it upside-down on the workbench, intending to go over it looking for dry joints. As soon as I got it under the light, I saw that this board had been repaired many times before, and not always in the most skilful manner. There were tracks pulled off, pads missing and component leads soldered directly together, and generally a mess. Plenty of dry joints, which kept me busy for a while. When I came to look at the vertical deflection section of the board, I found a large star-shaped crack under the vertical drive transistor's heatsink. The fragile phenol PCB (when did the Real World switch to fiberglass? Years ago!) had fractured when the heatsink had received a shock of some kind, and maybe half a dozen tracks were broken. I patched these with wire and crossed my fingers. On powerup... voila! the picture was rock-stable. The PSU caps are a little tired, though; I'll get around to replacing them sometime.
Next day I noticed a brightness flicker in the darn thing. I think it was the brightness trimpot. If you look at the bottom left photo in the group of four below, you can see in the lower left corner of the monitor PCB where I replaced the right-angle trimpot with a vertical trimpot. Look just below the group of four trimpots on the left edge of the board and you can see the replaced trimpot. Inconvenient; I now have to tweak the brightness with a screwdriver.
At the moment (3/9/97) I have two of these monitors in 100% working order after the above toils, one more which is fuzzy (the EHT is tired), and another with intermittent minor horizontal oscillation. I'm inclined to leave these two problems, partly because nobody is likely to notice them except me (I'm a monitor perfectionist), and partly because I'm really sick of poking around inside these machines.