finally got somewhere on the CX7A power supply. pulled an arc. was not supposed to have been pulling an arc, however ;) there are a lot of step-downs from a central 34 volt master supply for the semiconductor stuff in that hybrid ham transceiver, and at least one of them is shorted. in chasing down the cause, I determined the 34 volt master works.
or at least did until I whacked the supply to a short. back to it in a few days. it's nasty getting to the pass transistors in the machine, and that's where I pulled a temporary short. bad design put the transistors below a bunch of connection jacks, daring you to try and work on the thing.
the supply is supposed to go 0 volts with no current via a foldback loop in case of overcurrent, so the master should be OK. if it's not, I go to a 6-part regulator using a packaged 3-terminal regulator chip, and move forward. because the 34 volt master is fed from something approaching 50 volts DC, I'm using the high-voltage input version of the regulator.
signal/one had a history of issues with that power supply board, and I'm hoping the usual modifications using 3-terminal regulators for the 3 critical semiconductor feeds settles that bookful of problems. if so, it's pull the output tube, and start sneaking up the other supplies. for instance, it's quite helpful to get the screen supply working first, because that lights the Nixie tubes that display frequency. but light up the screen grid in a power transmitting tube without the plate supply of 1600 volts on, and you end up with a flash-bang exercise in making what is now a $540 power tetrode into a $0 gassy triode that just gets hot and does nothing remotely useful.
have written about this in other venues... but the machine was bought off eBay perhaps 8 years ago because it was top of the line and scarce when it came out in 1969... and at $2395 list, way, way out of reach for a new Novice ham in his senior year of high school. monkeys had apparently tried to fix it, losing all the undercover shields. the PS had a bunch of roasted parts, the plate compartment had some issues, appeared to have a grid-cathode short in the output tube, blew the top off one of the 20-watt power transistors driving the tube, burned up circuit board for the display and counter section... and then monkeys tried to fix it.
it's much better now. or should be.
just have to get power on it and see if the sparks and magic smoke stay inside the parts.