Medium voltage switching-mode power supply.

I'm thinking about the possibility of replacing the regular type of PSU that
powers a tube receiver with a switching-mode type. The current PSU has a
choke with gas filled regulator and provides outputs at 250v, 230v, 200v,
and 150v. I understand that I could get by with a much smaller transformer
if I used a swithching-mode PSU.

Does anyone know of any published circuits of a medium voltage S-MPSU? TIA.

 

I haven't any circuits to show you but I would do this with either a
forward converter or flyback convertor topology with stacked outputs.
I would wind the inductor or transformer with several different
isolated windings that had turns counts that would simultaneously
produce 150 volts, (200-150)=50 volts, (230-200)=30 volts, and
(250-230)=20 volts. Then after rectification of each, I would stack
the resulting DC voltages in series and regulate the total to 250
volts (unless the regulation of one of the lower totals was more
critical). This approach minimizes the high frequency voltage swing
of any section, reducing the stray capacitive currents. I might even
make the 150 volt section of 3 similar 50 volt sections stacked.

 

John

Thanks for the ideas.

The original receiver has one regular PSU outputing the voltages mentioned.
I'm going to reconstruct the rx by having three modules, rf, if and af. I'm
trying to get the whole rx into a 6U 19" rack cabinet, but it's tight for
space, and that's why I think of getting rid of the original PSU that has
quite a large transformer. I'm faced with a choice of either using one PSU
to power everything, or to put a PSU on each module. I kinda like that idea
for various reasons but it's not essential by any means, and might still go
for the one PSU solution.

I think what you are saying is wind 4 seperate secondaries and treat each
one as a seperate PSU in it's own right. But don't regulate each secondary
seperately.

Anyway, I think a key point here is the notion of winding ones own
transformer. This is a must. So, I've got to find a transformer supplier and
a circuit. Maybe if I look around I might come across some low-voltage
circuits and it's simply a matter of just upgrading the component voltages.

I think then task is to find a low-voltage circuit, then more-or-less copy
that design, but replicate the secondary circuit for as many secondary
windings that I have.

For instance:

http://www.coilws.com/Switch Mode Power/5W_SMPS.htm

This particular circuit is not regulated, but I think were saying that if I
were as an example to use this circuit I could stick to the circuit
arrangement except for component voltages in the secondary circuit, and I'd
have 4 rectified secondary circuits instead of one. Which I would join in
series.

Of course I must calculate the needed power output of the desired PSU. And
I'd have to figure out how to regulate.

 

Yes. There would be quite good regulation of all the outputs if you
regulate their sum. The big problem in a receiver would be to contain
all the high frequency noise of the supply. You would probable have
to enclose it is a separate metal box and filter all input and output
lines, besides the normal DC filter capacitors.

To use my method, certainly. for any other method, almost certainly.

Finding a core and bobbin is your challenge (besides designing the
rest of the circuit.

THis design is sort of regulated, by using the second secondary to
approximate what the main secondary is doing.

That is what I was describing.

Some other examples:
http://henry.fbe.fh-darmstadt.de/smps_e/smps_e.asp
http://www.maxim-ic.com/appnotes.cfm/appnote_number/2039
http://www.powerint.com/PDFFiles/an31.pdf
http://www.smps.us/Unitrode.html
http://www.fairchildsemi.com/an/AN/AN-4134.pdf
http://www.fairchildsemi.com/an/AN/AN-9015.pdf
http://www.onsemi.com/pub/Collateral/AND8039-D.PDF