19-07-2016, 08:43 PM
(19-07-2016, 05:00 PM)ppppenguin Wrote: I have a feeling that Mark Hennessey has a fair bit of accumulated knowledge on PSU design so it would be worth waiting for his input.
Not at those voltages
And what are those funny glass fings 
Still, the topologies are familiar enough...
The unregulated supply appears to be a pair of followers fed from a pot that is fed from unregulated supplies. To improve on that, I guess there are two options:
1. Stabilise the voltage across the pot to improve things a little...
2. Go the whole hog and add a bench of negative feedback, etc. Which is what the posher version obviously does.
Guessing Option 1 is the preferred route?
There's 2 references needed - one to fix the voltage across the pot, and another to offset the pot below ground to account for the "Vbe" of the followers. I'm guessing the latter is a relatively small (variable) voltage, and that might well be within the remit of the good old TL431 (that can stretch to 36V). But the high voltage one would worry me most, being a softy when it comes to this sort of thing. What are we talking about - 350V? So attention to detail is obviously needed. A string of zener diodes is the obvious "KISS" solution - I've reasonable confidence in them at higher voltages. Seven 50V jobs would be a start, leaving nothing too worrying over a single component.
Or, how about neons?
Anything more complex or precise is probably wasted given the lack of NFB.
