Switching PSU Theory of Operation / Repair Tips & Techniques - test-drb@ccmp.vtda.org

26 Jul 2000

Geoff Roberts stated thusly:

...
  This is starting to sound like it has either a
separate + and - supplies with a common ground, or
...
  it's some kind of voltage doubler.   Since I
can't imagine why they'd need to do a doubling act
...
  for the primary side of a SMPS I'm inclined to
suspect the former. 
Here's a quote from a former Prime engineer (who's been as helpful as his
memory permits):

: These things are much more complicated than that. Functionally the Line is

: directly connected to a special rectifier configuration called a voltage 
: doubler. It produces rougly 280 volts DC. There is no isolation
transformer, 
: which makes these things potentially lethal. The 280V DC is input to 
: switcher, which runs the 280V DC through the torroidal transformers to 
: produce the other voltages. The Big Bridge is to convert the output of the

: switch from high frequency (probably about 25Khz) AC to DC.

Howwver, your former remark would also seem to be true, as the two rectifer
outputs are +280vdc and -280vdc, respectively.

[..snippage..]

...
   Yes, but
aren't the 2 capactiors in series, something like :

 ------------+-------------------- +ve
             |
           =====
           -----
             |
 ------------+
             |
           =====
           -----
             |
 ------------+--------------------- -ve  
 Well they could be if they were across a potential higher than the invidual voltage
rating of the caps.
...
  That gives you an effective voltage rating equal to
the sum of the ratings of the caps.  The penalty is
...
  that it divides the effective capacitance by the
number of caps in the series circuit (assuming all are the
...
  same value - it's a bit more complex if they have
different values) Have you got a + where there should be -?
...
  The above diagram looks wrong to me.  If this was a
+/-  supply, then the lower cap should have it's anode 
...
  connected to the common (centre) ground point, or
it's effectively reverse polarised. It would be more 
...
  common to find caps in parallel to increase the
capacitance (opposite of series - the total c of the 
...
  circuit is equal to the sum of the values of all the
caps in it.)

 > And you're seeing 600V between the 2 points I've marked '+ve' and
'-ve',
 > right? That's still twice what I'd have expected -- I'd expect peak
mains
...
   voltage (i.e.
115*sqrt(2)) across each capacitor, or about 320V between
 +ve and -ve in that diagram.  
 The Prime techie mentioned a bridge, so I assume it's not a doubler circuit,
...
  that leaves some kind of +/- supply but I'm blowed
if I know how they achieve that
...
  out of a bridge.  I'd definitely like to see a
circuit for this one.

  I am still puzzled as to how it gets 600V without
either other large
 capacitors, or a mains-frequency transformer.  
 Ditto. 

                                                    --------------------- +
to BBU
                                                    |
                                                    |
             +---------{ MCI 307138 }---------------+--------------------
+ve
            / \                                     | +        |
           /   \                                  =====        /  12k ohm
          /     \                                 -----        \  5%
         /       \                                  | -        /
    ----*         *---------------------------------+----------*
    |    \       /|                                 | +        \
    |     \     / |                               =====        /  12k ohm
    |      \   /  |                               -----        \  5%
    |       \ /   |                                 | -        |
    |        ----------{ MCI 307138 }---------------+---------------------
-ve
    |             |                                 |
    |             |                                 |
    |             |                                 ---------------------- -
to BBU
    |             |
    |             | 
    |             |
115v Load    115v Neutral

...
   I will just
ask again -- this thing does run off 115V, right?  
 I keep wondering the same thing, if it was multiphase it might make sense, hmmm,
...
  I *wonder* if this was originally 3  phase and been
converted  to single?
 2 mains rectifiers in series?  With a common rail?  Sounds ugly but possible I
suppose.

There is an adhesive label on the front of the PSU that says "115 V". Since
it's
a removable label, that there is some way to change it over for 230v
operation.
With that in mind, I can report that there is a connector on the main PCB
that
has wires that loop back to itself. I'm wondering if it's replaced with a
plug
either without wires or with wires in a different configuration that it
changes
the PSU from accepting 115v input to 230v input. But about phases, I am very
ignorant.

[..snip-o-rooney..]

...
  > > > PALs? Are you sure? It's not at all
common to have significant amount of
...
  > > > digital electronics in an SMPSU. Most
of the time it's either simple
 > > > analogue circuitry either as discrete components, or one of the
well-know
...
   > >
SMPUS control chips.  
 How old is this box of tricks anyway?  I must agree, if there is a lot of logic
there,
...
  from the description of all the led monitoring etc, it
might be more to do with that than
...
  part of the supply proper.  Possibly control logic to
ensure the system won't power
...
  up if the supplies are out of spec, or shutdown the
whole thing if one fails etc...

I think this is correct.

...
  > > > In which case look for dried-up (high
ESR) electrolytic capacitors. This
...
   > > is a
very common heat-related problem.  
 Indeed.   Probably the commonest fault in PSU's, which tend to get pretty hot
...
  anyway.   The startup resistor is still a potential
culprit here too. 
Geoff, see my other post which is a reply to the same message to which you
were replying in this message to which I was replying. [hope that's clear!]

regards,
-doug quebbeman