The power supply for the early 434 is a primordial switched mode power supply. I've just downloaded the manual and had a shufty. We are talking a switched mode power supply dating from a 1971 instrument, and was probably designed in the late 60's.
I haven't had that detailed a look. The two flashing neons are associated with the starting the switcher going. They are a neon relaxation oscillator and should flash alternately every five seconds or so, with no precision at all. There are then two sets of discrete circuits which turn on and turn off Q1080. The turn off circuit looks at the collector current of Q1080 and when it reaches a specific value measured by T1080, turns the transistor off. How frequently this turn-on/turn off happens is controlled by U1130, a ua723 that is configured as a switching controller whose period of oscillation is controlled by the value of the +15V line. If that is the correct value it oscillates at about 25kHz. That frequency changes slightly to ensure that it stays at 15V. So the frequency of the whole switched mode supply is set by that 25kHz frequency.
The uA723 is still a current part by the way, and costs less than a quid. That is one hell of an innings for a piece of silicon.
The problem with switchers in general, and early ones specifically is that operation is somewhat arcane. First thing I'd do is unplug each transistor in turn and check it. Check one at a time, and if OK replace back in the board. Are all the diodes OK? Particularly VR1084/5 and CR1084 whose purpose is to stop the c-e voltage of Q1080 exceeding 250V. If one of those has gone open circuit that could explain a dead Q1080. Go through that lot and come back with what you find.
I'm not surprised that noone stepped up to the plate candidly. The list is littered with experienced people (me included) who have wasted whole weeks trying to resurrect a Tek switched mode supply, The top of the dresser in the spare bedroom has three 7000-series supplies that defeated me big time.
Craig
I haven't had that detailed a look. The two flashing neons are associated with the starting the switcher going. They are a neon relaxation oscillator and should flash alternately every five seconds or so, with no precision at all. There are then two sets of discrete circuits which turn on and turn off Q1080. The turn off circuit looks at the collector current of Q1080 and when it reaches a specific value measured by T1080, turns the transistor off. How frequently this turn-on/turn off happens is controlled by U1130, a ua723 that is configured as a switching controller whose period of oscillation is controlled by the value of the +15V line. If that is the correct value it oscillates at about 25kHz. That frequency changes slightly to ensure that it stays at 15V. So the frequency of the whole switched mode supply is set by that 25kHz frequency.
The uA723 is still a current part by the way, and costs less than a quid. That is one hell of an innings for a piece of silicon.
The problem with switchers in general, and early ones specifically is that operation is somewhat arcane. First thing I'd do is unplug each transistor in turn and check it. Check one at a time, and if OK replace back in the board. Are all the diodes OK? Particularly VR1084/5 and CR1084 whose purpose is to stop the c-e voltage of Q1080 exceeding 250V. If one of those has gone open circuit that could explain a dead Q1080. Go through that lot and come back with what you find.
I'm not surprised that noone stepped up to the plate candidly. The list is littered with experienced people (me included) who have wasted whole weeks trying to resurrect a Tek switched mode supply, The top of the dresser in the spare bedroom has three 7000-series supplies that defeated me big time.
Craig
