PDP-11/34A / BA11-K PSU

From: Tony Duell <ard_at_p850ug1.demon.co.uk>
Date: Mon May 26 17:30:45 2003

> > > Yes, but I have no clue about SMPSUs and there theory of operation.
> > Well, you may not need to really understand what you are doing at
> > first. About 95% of problems in these bricks are cured by testing=20
> > and replacing the power transsitors and the electrolytic capacitors.
> Well. I am an engineer. I want to understand and analyze, not blind
> poking.=20

OK... That's good. I hate lucky-dip servicing too... There are several
good references on SMPSU design (the section in 'The Art of
Electronics' 2nd edition is a good staring point), and these DEC PSU
bricks are a pretty standard step-down regualtor circuit. The only
unconvention feature is that the controller is a 723 chip, not a special
SMPSU chip. The 723 is basically a votlag reference and a (separate)
op-amp in one package, so it's not hard to understand.


  o-----\ /----------------+-----UUUUUUUUUUUU-------+-------o
         \ </ | L |
        -------- | |
Im | ----- =====
       Control ^ D ----- C Out
                              /|\ |
                               | |
                               | |
  o----------------------------+------------------------+-------o


Here's the basic circuit of the power-handling part of the step-down
regulator. When the control signal is asseted, and the transistor is
turned on, D is reverse-biased (so does nothing), and the left terminal
of L is +ve. Current starts to flwo through L, but the back emf of L
effectively drops the voltage that appears at the load (and, indeed,
energy is stored in the magnetic field of L). When the control signal is
deasserted, the transistor turns off. The magnetic field of L collapses,
so that now the left terminal is -ve. A current (effectively supplied by
L) no flows via D (the 'flywheel diode'), and supplies the load. C, of
course, acts as an output smoothing capacitor and preveents the load
voltage changing too much.

Yes, that's an oversimplification (but not as bad as some explanations
I've seen in print) but it will at least explain what the main components
on the schematic do.

-tony
Received on Mon May 26 2003 - 17:30:45 BST

This archive was generated by hypermail 2.3.0 : Fri Oct 10 2014 - 23:36:16 BST