First of all why is the "black" area red ? Have you cranked the G2 for
some reason ?

Regardless, it seems you have eliminated yokes and STKs as suspects.
That leaves the DCM. You said the symptom did not change when you
changed the STK, but did it stay EXACTLY the same ? There are alot of
out of spec STKs out there. The ones that aren't completely dead
somettimes have poor high frequency response, and that is critical in
alot of sets.

At this point in time I do not know how to get into the alignment mode,
and although you never try to adjust out a circuit fault, just how it
responds to adjustments may lend a clue. If you happen to know how to
get into the convergence alignment mode, let me know. There is no doubt
one of these things will darken my bench someday.

I don't know if it has a cursor based system or the old siz/lin/trap
etc controls. If the set allows you to see the whole picture while
adjusting the crosshairs you can see if it changes when you do so, if
not, perhaps somewhere in the service mode there is a DC offset value.
If the pattern moves uniformly the problem is in the DCM. Even if it
doesn't the problem could be in the smoothing/buffer stage at the
output end of the DCM.

You don't even want to know what the DCM costs, so all other suspects
must be eliminated first. If the problem is in the DCM, and it turns
out that the output is clipping the waveform before it gets to the STK,
that might be no more than an OP AMP, but if it is in the waveform
generator itself, we are talking some hairy ****, err scary ****. Also,
because most of it is SMD components it is not practical to swap
comonents of red and blue channels to see if the problem follows.

In the last picture, I am not really seeing the red lines, but I do see
the abberation. When you had it "cross channeled" I will say that this
is not normal though, that is if you took a normally working set and
swapped the yoke plugs I would expect the H linearity errors, but the
errors in the vertical lines at the top are defiitely not normal. The
general trapezoidal shape is to be expected (that's why you can't just
switch them), but the 'wave' in it is not right. My remaining brain
cells still scream STK, but that has been eliminated. The exceedingly
rare shorted convergence yoke is also exonerated.

I know the set is an incredible bitch to work on, but at this point I
would compare the waveforms at the + and - inputs to the red V channel.
You should have a very similar waveform on each, or practically nothing
if it is setup as an inverting amp. Those STKs are nothing but
overgrown OP AMPs. You should be able to get the spec sheet at
datasheetarchive.com. You may find that a little (pf) capacitor has
suddenly decided to be a resistor and is making the STK behave badly,
wouldn't be the first time.

In some sets you can switch the yokes, AND the DCM outputs to the STKs
as a definitive test, but I wouldn't count on it here. It would
probably entail cutting foils and installing jumpers, and then you
can't be sure of yoke polarity.

You said one STK is getting hotter, which one ? The one feeding the
offending color or the other one ? If the offending channel is cooler I
would look for perhiperal components in that stage (which are few). On
the other hand if the offending channel is getting hotter I'd look for
something across the output. Zener diodes, small caps etc.

You got a fun one there. Good luck.

JURB