21 Sep
2008
21 Sep
'08
3:20 p.m.
From: ard at p850ug1.demon.co.uk
At least the refernce voltage is more reasonable now :-). That old TL494
was almost certainly defective.
Hi Tony
I'd put these under #5. When finding a failed part, I generally don't
suspect other parts unless I can find some connection with the failure
I know. Not saying I've not seen multiple failures in old equipment, just
that it isn't as likely. Checking the diodes for the output is a good idea
in most any case.
Sure.
U4:
1 - 0V
2 - 2.40V
3 - 0V
4 - 0V
5 - oscillating
8 - 7.91V
9 - 4.43V
Hi Josh Getting closer. Pin 9 should be almost zero. The
two transistors ( Q17, =
Q18 )
form a latch. Normally this latch is not tripped and Q17 is on while Q18 =
is off.
Two sources of signals can trip the latch. One is an over voltage on the =
5 volt line
and the other is an over current in the switcher transistor. The over cur=
rent is detected
by the optical coupler ( D1 in the switcher board ).
All the levels on U5 look as expected under the current conditions.
There are a number of possible reasons that the Q17,Q18 latch has fired.=
Not
all are normal operation of this circuit.
1. D18 may be shorted and any voltage on the output trips it
2. Replacing the bad U4 has caused the voltage adjustment to be too high,
causing the circuit to trip.
3. Q18 may be shorted.
4. Q17 may be open.
5. Something else
How about a genuine overcurrent-in-the-chopper problem? You seem to have
neglected that inout to the latch. I suspect a shorted rectifier diode on
the output side of the PSU could do that, or a shroted smoothing
capacitor.
Check for dead shorts between each of the outputs and ground. If you find
one that tests shorted, desolder the assoicalted diode and capacitor and
check those for shorts.
Do we know that this model of supply works correctly with no load? Some
SMPSUs will go into overvoltage shutdown if they have no load. Equally a
high ESR capacitor on the 5V output would put high-ish voltage spikes
there (around 10V typically) which would trip that latch and shut things
down.
I guess we should recommend that he connects a dummy load. Connecting
a 6V car lamp is usually enough. As you say, the turnon spike with
no load might be enough to trip the overvoltage latch.
We know that U4 was bad so it may have been the
only bad part. This
makes #2 a likely possibility. Try adjusting the pot R39 such that the wi=
per
is closest to R38 ( maximum resistance ). Cycle the power switch.
Normally I wouldn't suggest adjusting any pots in a non-working circuit
but this one should be harmless. Don't worry about remembering the
My rule is that I'll only tweak a preset if :
a) Misadjustment can't cause further damage.
b) I know the procedure for settign it correctly and have the equipment
to do so.
Here. even if the PSU overvoltages it won't do any damage because you
don't have the rest of the computer connected to it. And the prcedurte
for settign it is to adjust if for 5V on the 5V output connection,
something you only need a voltmeter for. location on this pot as it would have needed some
adjustment anyway,
with the replacement of U4.
Do not fiddle with any of the other adjustments!
Whateverr you do, don't disable the overcurrent trip circuit. Doing that
is a sure way to gets bits of (expensive) chopper transistor distributed
round the room and the magic smoke escaping from all sorts of components.
Please don't ask how I found that out (OK, I was young and foolish!).
-tony
_________________________________________________________________
Want to do more with Windows Live? Learn ?10 hidden secrets? from Jamie.
http://windowslive.com/connect/post/jamiethomson.spaces.live.com-Blog-cns!5…