[Chris55000]

Hi!

By comparison of the values of the reservoir capacitors!

The +12.6V series stabiliser is fed from BR3 and winding q–r of the mains transformer, with a 2000uF reservoir, so that implies winding q–r is designed to provide a much higher current than the other three, which only have 500uF reservoir capacitors!

I estimate the current taken from the +50V series stabiliser TR6 to TR10 is going to be in the region of 50mA or so, (you can use a bench power supply to determine it exactly), so if you use the module I suggest, the current taken from the 12.6V rectifier BR3 should be an additional 200–250mA, which winding q–r on T1 should be able to supply!

The output from BR3 would need to be about 15–18V for the +12.6V series stabiliser TR11–15 to work properly, so assuming it's 18V and 50mA is drawn from the +50V line, equates to about 4W of supply power, assuming the step–up converter is 75% efficient!

Therefore the approximate supply current needed to provide 50mA at 65V is about 4/18 or about 220mA!

The reason I didn't suggest using winding o–p and BR4 to supply this module is because it's almost certainly going to been wound of the thin wire used in winding m–n and attempting to draw another 220mA from it would almost certainly have damaged it!

That bring me to your final question – why was the original winding m–n 77.5V a.c?

The reason is because of the voltage drops in the two 2 ohm series resistors, the first between the rectifier BR2 and the fuse, plus the second after the fuse, means that in order to provide "headroom" for the +50V series stabiliser TR6–TR10, Cossor determined at design stage that the stabiliser requires +65 to +67.5V, therefore allowing for the ripple voltage across C4 plus the voltage–drop in the relatively high–resistance secondary m–n, and two diodes, and the two series 2 ohm resistors, about 75V r.m.s. off–load is needed for the circuit to work efficiently, so this is what Cossor designed their transformer to provide!

As to why winding m–n failed, the answer is too much iron and too little copper! Cossor designed these transformers to run at very high magnetic flux densities:–

e = 4.44 × f × B × N × A where;

e = induced secondary voltage; f = supply frequency, B = magnetic flux density in core, A = cross–sectional area of core!

By using a very high magnetic flux density the turns/volt figure for the windings can be maximised, but this unfortunately leads to a higher voltage stress between individual turns!

What happens is that inevitably at some point, the enamel insulation will start to flake off the winding wire in localised spots deep in the windings, which results in local large circulating currents between turns, causing further overheating and even more heavy current flows, and the whole thing becomes self–perpetuating until the winding burns out, as happened in your case!

Chris Williams