More troubleshooting. Ordered a new filter capacitor and bolted it in place (literally).
Re-soldered the old crowbar back in and powered it up gently on the Variac. At about 60 Volts there was a flash from underneath and a puff of smoke. Killed the power and upended things to try see what had shorted.
There was nothing immediately visible, the new capacitor hadn’t flashed over, there were no scorch marks on the circuit board anywhere that I could see.
I then tilted the chassis and plink, this fell out:
The leg of a component that I had not caught, which had shorted out across the mains input. Oops. This is what happens when you work tired, kids.
Nevertheless, checked for any more and powered it back up again. Same result, 300mV under load, voltage rising then hitting the crowbar.
Lifted the main chopper transistor out of the +5 circuit.
Exchanged it with the one from the +15/-15/-12V circuit.
Stable five volts exactly. The +15v group rise up then cut out on the crowbar.
Ordered a couple new pieces from LittleDiode in the UK. They should be here in the next couple weeks.
Documented all the chips on the three boards this machine is populated with today.
Took a picture of the backplane. (High res version here).
Power from 1 thru 7 is +5 +5 GND N/C GND +12 -12 which is slightly bizarre.
Output to the front panel is via the ribbon cable, the board looks like this:
Brightness and volume controls, as well as a number of LED’s and the connection to the loudspeaker.
More PSU troubleshooting tonight.
Spent a bit more time Saturday trying to figure out what I was doing with this power supply.
Broke out the oscilloscope and Variac. Upon suggestion from someone else, I removed the crowbar thyristor and ran the circuit up on no load.
It decided to go flying up in voltage and doesn’t regulate. Something’s gone screwy because the voltage used to go up and stay stable; off load I could vary the value, and on-load I could, to an extent.
Started checking all the solder joints and resistor values. Decided at the same time to renew the resistors in case one had a high-voltage problem that didn’t show up testing with my multimeter.
There’s a 555 timer chip on the board that outputs this pulse:
That gets converted to a sawtooth and fed around the board on a bus to each section of the supply. Looking at the rest of the circuits, the chopper stays open until it reaches voltage, then starts to wobble at the same speed as the sawtooth, regulating to the correct voltage.
This is not happening on the 5V section, and the voltage is just skyrocketing; the crowbar cuts in and shuts it all down.
Just need to find why it’s not hitting the 5V and regulating properly.. might end up being the chopper transistor, that one’s been replaced in the past. We’ll see. I have a new crowbar to try (don’t think it’s the problem) and a new filter capacitor (that’s going in as a matter of course) but there’s a 5.6V Zener diode that operates the crowbar that I think is a bit squeaky and there’s definitely something wrong with the regulator circuit that it’s not clipping the voltage correctly.
Couple more diodes to change out on the board, might do that tonight. We’ll see.
I received a very nice email from a gentleman in the UK who used to own an AES SuperPlus IV (seems to be one of these with more RAM, a modem card and a detachable keyboard) and he was apologetic that he did not have any schematics from his research and time spent speaking with former Lanier employees etc.
His suggestion was to work out what’s on the board and go from there.
So I spent a couple hours after dinner tonight with my flashlight, multimeter, paper and pencil and ended up with this:
(Click here for a high-res version of that image).
There’s a couple places I should start poking with my meter but the crowbar does look to be possibly at fault.
New crowbar and filter cap arriving in the mail soon.