PSU problem

[orange]

I posted here before about my problems with computer power supply
sometimes not starting. Finally, I have replaced almost all capacitors.
Since I could not find those BIG two, 680uF 200V, I've used two 1000uF
200 V instead. It seems to work much better now, but still it sometimes
doesn't start immediately, then a click can be heard and it starts. I
don't mind waiting a second or two each time when power is turned on,
but could it be dangerous in the long run?
I think that once or twice it didn't start even after waiting (

 

[[email protected]]

Generally, the big electrolytic capacitors do not fail. It is usually
the smaller ones that go. I had a few where the capacitor in the
feedback circuit would develop high ESR and cause the PS to fail to
start. These are usually about 47 Uf at about 16 to 35 volts. It is
best to use an ESR meter to locate the bad electrolytic capacitors.
Also look for any domed or puffed-up capacitors. These are bad.
John

 

[Bob Parker]

Is this the same thread in which I suggested checking any startup
resistors in the circuit to see if their value had gone high, and
everyone ignored me?

 

[orange]

err, yes, sorry I ignored that.. could you please tell me more about
startup resistors, how do I find them?

 

[Bob Parker]

Most switch mode power supplies which are powered from the
rectified & filtered mains supply have one or more resistors from the
positive terminal of the main filter capacitor and into the control
circuitry, which bias the switching transistor etc and allow it to
actually start running.
These resistors have a value usually in the hundreds of K ohms, and
they are notorious for going high in value from having a high DC
voltage constantly across them.
I'd suggest looking for any resistors in that part of the circuit,
then after ensuring that the main filter cap is fully discharged,
measuring their value out of circuit i.e. by unsoldering one end.
I hope this has been of some kind of help to you. Good luck.

Bob

 

[Ross Herbert]

Most switch mode power supplies which are powered from the
rectified & filtered mains supply have one or more resistors from the
positive terminal of the main filter capacitor and into the control
circuitry, which bias the switching transistor etc and allow it to
actually start running.
These resistors have a value usually in the hundreds of K ohms, and
they are notorious for going high in value from having a high DC
voltage constantly across them.
I'd suggest looking for any resistors in that part of the circuit,
then after ensuring that the main filter cap is fully discharged,
measuring their value out of circuit i.e. by unsoldering one end.
I hope this has been of some kind of help to you. Good luck.

Bob

Bob, this type of failure is one which many techs miss. I have found
one particular brand/model of SMPS I encountered uses a 500K startup
resistor and it often goes high or o/c. The problem I think is due to
the type of resistor selected. Metal film seems more prone to failure
than metal oxide or carbon film. I also experienced this same problem
when using a metal film resistor as a current limiter on a half wave
direct 240Vac mains rectifier circuit. substituting a carbon film
stopped the problem thereafter.

Ross

 

[Bob Parker]

Hi Ross,
Thanks for your observations. Back when I was doing repairs full
time now about a decade ago, I was replacing those resistors with
over-rated metal film ones and *hoping* that they'd handle constant
high voltage DC better than the carbon film etc ones they replaced.
I didn't have any of them subsequently fail, but I probably didn't
stay in the repair business for long enough to find out that they
aren't as reliable as I hoped.
Next time I encounter an open circuit start resistor, I'll remember
your advice and go for carbon film.
This newsgroup is all about learning from each other's experiences!


Cheers
Bob




On Tue, 30 Aug 2005 07:21:54 GMT, Ross Herbert