Re: Tek 576 possible power supply failure
I put the schematic of those 2 supplies together with the voltages (hopefully readable) in failure mode in the files section. See if anyone finds the faulty transistor from that information alone...the winner is: Q837.
My first thought was of Q834, but alas no, it was Q837 with a sufficient short across CE (behaves exactly like a resistor). Its the current limiting transistor and the supply sprang to life as soon as I took it out. I think I could have guessed it by looking at the CB-voltages, which are very close.
This fault developed over time - at first the Tek went dark after an hour or so, a year later I barely could make a measurement before it happened. Thats what made me think of a capacitor rather than a transistor.
Curiously, all capacitors in this rather old unit seem to be OK...
My repairing of the step generator succeeded... it took me an afternoon understanding how it works (good manual), making tons of measurements to follow up the signal path until the counter, just to end up with a working step generator without having changed anything. The wiggling of the ICs and transistors must have done the trick.
There's an excellent review of the step generator here (in german):
The guy found out the there is an inherent design issue with the generator, as they are applying relatively slow ramping signals to the TTL logic. The old ones in the 60ies do switch at some point and thats OK, but the newer ones are faster and tend to switch forth and back during the time the signal is in the undefined area. At the very end this produces two or more clock pulses in a row, making the generator jump the corresponding steps. The best way to correct this is to replace the 7400 by a 74LS132 that has schmitt triggers at the entrance.