Commodore Superpet video problem -- ideas? - test-drb@ccmp.vtda.org

1 Feb 2013

...

 I'm having intermittent problems with my Superpet. It will run fine for some
 time, but then will "glitch" for a couple of seconds (including video
 distortion), and then hang. 

 Having done a bit of digging, I suspect that the flyback transformer is
 arcing internally. The video board is the one in this schematic
 (http://zimmers.net/anonftp/pub/cbm/s...032/321448.gif).

The link is mangled.  I think it should be:

http://zimmers.net/anonftp/pub/cbm/schematics/computers/pet/8032/321448.gif

...

 I can hear the arcing sound from area of the transformer when the glitch
 occurs, and I see no obvious external indications of anything sparking or
 shorting. I have looked at the voltages at the various test points noted
 on the video board schematic during normal operation, and they all appear OK.

 I'm sort of stuck at this point. Should I replace the flyback? The part
 number is 2432641AL 25M, and it's T721 on the board; does anyone know of
 an available equivalent?

I'm not familiar with the the Superpet but I have had similar faults in other
equipment which were caused by failed joints on the PCB.  If the problem
happens reasonably frequently, one approach is to darken the room and watch
the underside of the PCB for little sparks.  If it doesn't happen that
frequently, another approach is to try tapping the PCB at various points with
a well insulated implement such as a plastic ruler or other tool to see if the
fault can be mechanically provoked.

Do not overlook the possibility of arcing at the EHT connector on the CRT or
on the CRT base.

I would hate to end up replacing a major component and finding that the
problem was still there.

...

 Alternatively, could this behaviour be caused by a bad electrolytic
 capacitor? I'm specifically suspicious of C754 (47uF/250V), but my knowledge
 of CRT theory is not deep, and I lack a good way to test the capacitor in
 circuit.

It's not the sort of thing I'd suspect an electrolytic capacitor of but I
suppose it is possible.  Usually when they are stressed enough to make noise,
there are visible signs of stress too.

An oscilloscope connected between its negative terminal and ground it would
probably show a significant increase in ripple or reduction in voltage when
the problem happens if the capacitor is at fault.

Regards,
Peter Coghlan.