And deeper the rabbit hole goes...
So while the thing was open, a little mishap caused two transformer taps to get shorted for a split second (I am not 100% sure which, but I think it was the ones highlighted in orange in the image attached, pins 7 and 8 on the main PCB). The deck still
worksworked, but I noticed the VFD being brighter than expected, with some crosstalk between segments I'd never noticed before. A bit of probing showed that the -24V rail (highlighted in green) was at about -31V, roughly the voltage of the input to the -24V regulator (the red box). Looks like the pass transistor for -24V (Q708) went short. I put in a BD140 (the only PNP I have on hand with sufficient wattage and the correct ECB pinout), and while it's now regulating again, it's about 4V low. But the input to the base of Q708 is low, too, so I don't think the different transistor specs are to blame. I replaced Q707 (which drives the base of Q708) with another BD140, no difference other than a very slight additional drop in the voltage of the -24V rail. (I kept it, in case it's advisable to put it back.)
Both sides of zener D716 and its parallel C715 measure at -19.5V, so probably one of them has failed short? (Powered off, 16 ohms across them.) This node, which also has a 47K resistor up to ground, connects to the center tap (highlighted in blue) of the VFD filaments' transformer winding. I don't understand this part of the circuit.
The +7.5, -7.5, +12, and +5.5V rails are all fine. (5.5V is actually at about 5.38V.)
So I went to measure voltages today to give you more info, and lo and behold, in between last night and now, the thing won't come alive. All I did was unplug it, put the disconnected parts inside, and put the cover back on.
- The VFD filaments (which are 4.4V AC powered from their own transformer winding) do glow, and the capstan motor does its power-on thing, but no display, no reaction to buttons.
- All voltages check out, save for the -24V rail.
- The control CPU/ASIC has 4MHz oscillation on the corresponding pin (manual says it should be 4.1Vpp, but it's about 5.4V peak to peak.)
- CPU supply voltages are OK.
- Testing more voltages shows that the POWER.OUT output (pin 34) is high, indicating the CPU is not detecting power on the AC detect signal input (POWER.IN, pin 25), and checking that pin measures about 0.17V, whereas it should be high.
- Furthermore, the reset pin (pin 27, active low) is high, and scoping that pin on powerup shows no "several 100msec" low as is supposed to happen according to the schematic.
- The schematic says the input and output of the reset generator (IC804) should be at 5.5V and 5.3V, resp., but they're both at 5.38V, so no 0.2V voltage drop on the output.
- If I short the output of the reset generator to ground (short pins 2 and 3), the capstan motor turns on, and letting go, the motor turns back off. This indicates to me that the capstan control output of the CPU (pin 13) is working, since that pin actively pulls down to turn the motor off.
So as best I can guess, the CPU is not getting a reset, and it's not getting the high on the AC detect signal. Tracing that pin out, it's got a 22K pull-up resistor and a 0.1uF cap to ground, and then goes off through diode D711 to the transformer tap that powers the 5V rail. (Powered off, D711 measures about 1.5V forward and 0.6V in reverse, as measured in-circuit.) Measuring the voltage on D711, it's the 0.17V on the anode side, on the cathode it's 8.6V DC, 1.5V AC.
(The rectified but unregulated voltage to the 5V supply is supposed to be 8.5V, but it measures 9.9V. But since the regulated 5V rail is within 2% of where it's supposed to be, I guess this isn't a problem?)
I'm obviously in way over my head now. I'd be appreciative of any help you can give. (No rush as such, I'm about to leave for a trip for 2 weeks, so won't be able to look at it.)