(03-03-2013, 09:05 AM)AlanBeckett Wrote: Never mind that, tell me how to fix the PSU
Alan
Your wish, sir, is my command!
Well, it sounds as though the EHT generator is doing something sensible & that's a good start. As for the PSU voltages "being all over the place", we are now faced with the usual questions: (i) are the faults confined to the PSU section only? (ii) are the faults confined to the non-PSU sections only? (iii) are there faults in the PSU and the non-PSU sections?
Considering the age of this item, I would expect (iii) to be the most likely answer, and that encourages the 'scatter gun' approach now to be taken throughout the entire unit: check every R that is 47-kΩ and above; check various capacitors (except small pF values); clean all rotary switches and rotary controls. As per usual, that will probably not remove 100% of all the faults (so a detailed cct. analysis will then be necessary). The alternative approach to fault-finding is to use the strictly logical / analytical style, starting with the PSU. That method may involve more thinking and can be more mentally-rewarding but possibly requiring more hard work.
I would go down that latter route: divorce the PSU from the rest of the circuit and get that section going first - then move on to the rest of the cctry. I say that with particular regard to this item since I see that it generates + HT and - HT rails. Although it is almost certain that the Y-amps. will use differential ccts., (hence the ±150v. & ±85v. rails), we can't overlook the possibility of a derived bias line being incorrect and consequently a valve from being seriously over-run. So I would start by electrically separating the PSU connections from the rest of the electronics. Unfortunately, with Marconi kit of this vintage, it's my experience that that is a lot easier said than done, since Marconi did like to build their kit on the basis of mechanical integrity first and electrical maintenance capability last: they had an irritating habit of fitting components where they were physically convenient to fit, not necessarily in positions that approximately followed the electrical cct. diagram.
Having divorced the PSU O/P rails from the rest of the kit, I would then use the 'scatter gun' approach within the PSU and then see what the results were. If faults still remain, then I would then fit dummy loads on each PSU rail: just light loads, e.g. the cct. states 300 mA from V604 cathode, so I'd draw 20 mA, say. Then the analysis of the PSU cct. can begin.
As is usual with commercially-drawn cct. diags. (and Marconi is a prime example of this) the PSU cct. is not drawn in such a manner that makes its operation very clear. My first area of concern would be that little bit of cctry. that states "Heaters V1 & V2, lines A and B". I'll make an inspired guess that these are the first valves in the Y-amplifier(s). This heater line is rectified d.c. which changes according to whether the time-delay relay is energised or not. With that in mind and the fact that that d.c. is sourced from a 250 v.a.c. winding causes me some concern.
If that checks out O.K., I'd then move on to an analysis of the rest of the PSU cct.
The diag. shows that MR601 and MR602 form a rect. cct. that produces an 0v. / +H.T / +2 X H.T. arrangement: +H.T. to V606; +2 X H.T. to V604, resulting in + 300 v. and + 150 v. after regulation. (The 0v. line for those rails goes back to the -ve. O/P of MR601: that is not obvious to spot: a poorly-laid out cct. diag.) The error-amplifiers to each series-pass 6080 are much the same as each other, and no neon-stabilization is employed (although with a casual glance at the cct., it looks like that that is not the case: Marconi's cct. diag. poor layout again). Overall, these error amplifiers look quite conventional.
The -ve rails share a common series-pass stabilizer, V600: a 6CD6G. (Note: these valves are now rather scarce, although I do have one in my collection, condx. unknown). The overall idea here is to 'regulate' the 0v. rail for the -ve. rails, but since that is impossible, the -ve. rails, in turn become stabilized relative to 0v., with the -85v. rail being additionally stabilized by V.603. and V608 stabilizing the supply voltage for the error amplifier, (which also furnishes a stabilized +85v. rail). The overall operation of this -ve. stabilizer is not very clear from the cct. diag. My brief analysis is of it as follows.
V602 forms a long-tailed pair phase-inverter with the control grid of the pentode section stabilized via V603. @ - 85v. Since the cathode voltage of V602 will vary according to the load on the -150v. line and the grid of the triode section is also taken to that rail with its anode at 0v., a variable error-related current will pass through the right-hand half of V601 producing an error voltage at V601 pin 2 and thus to the control grid of the series pass valve, V600.
There are two unstabilized rails: "to 'X' amp." and "to EHT unit". The purpose of the latter is fairly obvious; for the former, an inspection of the cct. of the X-amp. should reveal all.
In the bottom LH corner of the cct. there is a line labelled "+150 v." I cannot see how this is a source of voltage @ +150 v., but, OTOH, I cannot see why it should be tied to the + 150v. rail, either. It may have something to do with reducing the h-k voltage difference for those valves whose heaters are fed from LT3 and LT2, although in that case, why the heater of V609 is similarly involved, I cannot understand. Perhaps it is something to do with the inductive kick-back from RL600.
As for the time-delay cct., I see that RL600 is a latching relay whereby the heater for V609 is only energized during the warm-up period. The functions of the relay contacts are obvious from the cct. diag.
I hope that helps,
Al. / March 3, 2013 //
