Tuesday, 12 April 2016

Arrrgh! Getting fed up of this bloody PSU!

After studying the circuit diagram of module 5 for a long time, I began to think that the only remaining tantalum capacitor likely to have any influence on the timing of the HV soft start was the 330uF smoothing cap on the pulse transformer.

So, I changed it. In removing it I also removed the PCB trace between the through hole and the via to the groundplane, as the damn -ve leg was bent over and soldered. I ended up having to cut it off. Not having any 330uF caps, I opted after much pondering to use a modern 470uF radial electrolytic. This required some clever leg bending to solder the -ve to the top groundplane and route the +ve to the through hole.

I then reinstalled it and tested it. Still faulty! Arrrgh!

It was then I had a brainwave. Testing the 110v rail showed the same steady creep up to full volts, and I had always assumed that the 121v test point did the same. But now, on a hunch, I put the meter probe to the 121v TP and switched on - almost instant 121v! So - whatever is causing the issue must exist after the 121v section.

Looking again at the circuit diagram, and this time ignoring everything to the left of TR6, I turned my attention to the 110v adjustable regulator, the HV electrolytic in which I have already changed out. It was then I noticed that TR7's emitter connects not only to ML2's precision zener, but also to C7 to ground, and to R9 to the 110v rail - a bloody RC timing circuit!

I have the module dismantled and opened out on the bench. Tomorrow, I will pull R9 and C7 and test them. If my hunch is right (and theres absolutely NO reason that it should be!) one or the other will be out of spec.

On a slightly different note, I decided that im happy with the red paint job on the LiPo battery for the Clansman equipment, so have taken the masking tape off. A coat of clear lacquer might be needed to smooth down the paint edges.

No comments: