You say that the caps "look" OK - but have you actually tested them?
The main cause of faiilure of these things that I have experienced is
cap failure.
Either take them out, measure them - including (especially ?) the esr
- or just replace with brand new.
Start with the startup filter cap.

RP

On 02/12/05, Roland <caonmail@mweb.co.za> wrote:
> I did repair a few of UC3844 a few years ago.
> Usually the 3844 is powered, on startup via cap/resistor network. This is
> only required for a few pulses until the transformer switches and one of the
> secondries supplies power to the chip. It sounds like what you are seeing is
> simply the startup switching, and not the proper running.
> A few things to try;
>
> 1 - it sounds like one of the outputs may be overloaded, so it cannot
> startup.
> To test, remove mains and apply the correct output voltage(+5? +12..)  with
> a PSU after the output bridge and see if there is excessive current draw
>
> 2. - or the startup circuit may be faulty.
> To test, power the unit up and use a PSU with a diode- in-line to supply 12V
> directly to the 3844, and see if it runs smoothly
>
> If something goes bang, at least you'll know where to look  (:-o
>
> Regards
> Roland
>
>
>
> ----- Original Message -----
> From: "Alessandro Queri" <piclist@tiscali.it>
> To: <piclist@mit.edu>
> Sent: Thursday, December 01, 2005 3:15 PM
> Subject: [ee]: UC3844 switching PS repair
>
>
> > Hello everybody,
> >
> >  http://www.fairchildsemi.com/ds/UC/UC3844.pdf
> >
> >
> >  I 'm trying to repair myself a switching Power Supply made by ZyXEL. The
> PS
> > suffers from 5V line oscillating (actually the PS starts, shuts down,
> > starts...). I checked against the output filters and CAPS and they seem
> ok,
> > so I took the Datasheet for the UC3844 and went decoupling the AC line
> with
> > a 2:1 transformer. As many of you know, here in Italy the line is 220, so
> > I'd like to decouple and lower line to get 110 and save my oscilloscope
> ;-)
> > . First the IC doesn't start even if the PS is rated 100-240VAC, so I had
> to
> > ramp up the AC line to 184 (strange, isn't it?). The supply on the IC
> > (capacitive reactance supply) seems to be a bit noisy floating from 14 to
> > 16.5. On the other hand, when externally supplied, the IC starts over 16V
> > and stops running below 9, so that's ok. Now I read from the datasheet:
> > typical osc freq 52 KHz but I cannot understand where to test it: on pin 4
> I
> > have 92 Khz while on pin 6 I have 46 Khz and a really UGLY waveform (seems
> > chopped, as being amplitude modulated, not as in fig. 5). Is the former
> FAR
> > away from the working point or the latter a little below? Has any of you
> > clues about usually suspected components in faulty PSs? Many thanks in
> > advance for your help.
> >
> >  Alessandro Queri
> >
> > --
> > http://www.piclist.com PIC/SX FAQ & list archive
> > View/change your membership options at
> > http://mailman.mit.edu/mailman/listinfo/piclist
>
> --
> http://www.piclist.com PIC/SX FAQ & list archive
> View/change your membership options at
> http://mailman.mit.edu/mailman/listinfo/piclist
>

-- 
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist