Hi;

(I've tried to be thourough here)

As you know I've taken a new job. As such I am being "introduced" to all the
dogs. Such fun.

Two Toshibas, one was a midaged one, H line or so, the other, today was a D
line.

Both had the same initial symptom, blue horizontal covnergence was off, and it
got worse when it warmed up, all of the sudden. Replacing the STKs stabilized
it, but they both needed adjustment, and I don't mean just a little. Controls
just barely had enough range. Now I really really hate to adjust a problem, but
I did because we were under the gun, remember these are the dogs. I checked all
the output circuitry and the problem must be in the digital board in the newer
one, but what about the D line ?

Thing has pots, and the same waveforms are applied to the common pots for each
function, thereafter it is buffed by an OP AMP.

Am I missing something here. Something different about the blue ?

You've seen the NAPs with the blue tube mounted closer to the mirror. I assume
it's because of the difference in the wavelength of the light. I guess this
allows them to use the same lens for all colors, but others do and they're just
fine. I think I have seen some sets with different blue lens.

Can I conclude that blue convergence is pushed a bit harder in these sets ? Red
and blue convergence yokes are the same, and everything's the same in some sets
and they seem to work just fine. Is it possible that the blue convergence yoke
runs hotter making it more prone to a failure like partial pulverization of the
core or something ? why else would it be able to be adjusted ? I checked best I
can for a hard circuit fault and none seems to exist.

The only test I can figure out next is to tie the red H output to the waveform
for the blue and compare the uvoltage/u waveforms. Actually even if I was
aware of a tester that could ring a convergence yoke, it isn't worth it for two
jobs. If any problems appear I'd rather test by substitution. Switch the red
and blue and see if the problem folows.

See the thing is, every control has it's intended effect. There are no clipping
artifacts on the screen. Supply is good, at least I think so, the +&- 12s are
11.57 and 11.something. The 5V is right on the money. Yoke return resistors are
good.

The red was fine and geometry was good. The blue had a vertical line bow that
opposed the side pincushion errors. There was also a horizontal linearity error
in the direction of uncorrected convergence. Low gain would explain that, but
not the bow. What it comes down to is if the impedance/inductance of the
winding in the blue convergence yoke is it, the STKs will put out less voltage
at the same current.

Has anyone ever even seen a defective convergence yoke ? I haven't (well
maybe). Or, could this be one of those elusive STK eating problems ? This set
has had the STKs changed before, but that doesn't prove much, except that the
others who did it didn't do it well enough so you can't tell. (I don't know
about anyone else but I consider that to be a standard for a totally correct
job) Luckily the pads are fine.

I don't think it's a main yoke because the focus is good. Any ideas ? I don't
just mean something helpful on this job, I mean what if this gives someone an
idea to fix their nemesis ? That would be good.

I think a way to test these things is in order. Actually since it is a current
source output stage, I think I/we can figure out a way to test it definitively.
Perhaps by interrupting the signal connector and inputting our own test
waveform. It can probably be a square wave, and once we see a good one this
would be a definitive test. If the output voltage is lower with the standard
input, then the reactance of the yoke is low, with everything else being equal.
Or a spectrum analyser, but that is out of reach for most shops for it's (now)
limited purpose.

Even with a scope, there are other possibilities for such a device. The only
problem is we need a standardized test signal that is effective, a picture of
the waveform produced by a new (or good) part, and a square wave generator.
Perhaps one resistor beyond that, so you know the source impedance.

A spectrum analyzer would give us the data to build a proper filter for the
DUT, but really knowing the scope has greatly enrichened my experience in this
field. I could simply compare that voltage waveform to that of the red. Hmmmm,
check it out, I don't need ANYTHING from the manufacturer to do this. No data
no nuthin.

I'm on the M___________ and unlike Pulp Fiction, I'm not sending the Wolf, I am
the Wolf.

Well, thanks for being bored enough to read all this and thanks in advance for
any ideas, not only about the Toshiba, but figuring out how to test things. I'm
for anything that enhances the repair industry.

JURB

If you want to get the convergence yoke out of the equation, just ring
it Bob Parker sells one for not to much.

"JURB6006" wrote in message
...
Hi;

(I've tried to be thourough here)

As you know I've taken a new job. As such I am being "introduced" to all
the
dogs. Such fun.

Two Toshibas, one was a midaged one, H line or so, the other, today was a
D
line.

Both had the same initial symptom, blue horizontal covnergence was off,
and it
got worse when it warmed up, all of the sudden. Replacing the STKs
stabilized
it, but they both needed adjustment, and I don't mean just a little.
Controls
just barely had enough range. Now I really really hate to adjust a
problem, but
I did because we were under the gun, remember these are the dogs. I
checked all
the output circuitry and the problem must be in the digital board in the
newer
one, but what about the D line ?

Thing has pots, and the same waveforms are applied to the common pots for
each
function, thereafter it is buffed by an OP AMP.

Am I missing something here. Something different about the blue ?

You've seen the NAPs with the blue tube mounted closer to the mirror. I
assume
it's because of the difference in the wavelength of the light. I guess
this
allows them to use the same lens for all colors, but others do and they're
just
fine. I think I have seen some sets with different blue lens.

Can I conclude that blue convergence is pushed a bit harder in these sets
? Red
and blue convergence yokes are the same, and everything's the same in some
sets
and they seem to work just fine. Is it possible that the blue convergence
yoke
runs hotter making it more prone to a failure like partial pulverization
of the
core or something ? why else would it be able to be adjusted ? I checked
best I
can for a hard circuit fault and none seems to exist.

The only test I can figure out next is to tie the red H output to the
waveform
for the blue and compare the uvoltage/u waveforms. Actually even if I
was
aware of a tester that could ring a convergence yoke, it isn't worth it
for two
jobs. If any problems appear I'd rather test by substitution. Switch the
red
and blue and see if the problem folows.

See the thing is, every control has it's intended effect. There are no
clipping
artifacts on the screen. Supply is good, at least I think so, the +&- 12s
are
11.57 and 11.something. The 5V is right on the money. Yoke return
resistors are
good.

The red was fine and geometry was good. The blue had a vertical line bow
that
opposed the side pincushion errors. There was also a horizontal linearity
error
in the direction of uncorrected convergence. Low gain would explain that,
but
not the bow. What it comes down to is if the impedance/inductance of the
winding in the blue convergence yoke is it, the STKs will put out less
voltage
at the same current.

Has anyone ever even seen a defective convergence yoke ? I haven't (well
maybe). Or, could this be one of those elusive STK eating problems ? This
set
has had the STKs changed before, but that doesn't prove much, except that
the
others who did it didn't do it well enough so you can't tell. (I don't
know
about anyone else but I consider that to be a standard for a totally
correct
job) Luckily the pads are fine.

I don't think it's a main yoke because the focus is good. Any ideas ? I
don't
just mean something helpful on this job, I mean what if this gives someone
an
idea to fix their nemesis ? That would be good.

I think a way to test these things is in order. Actually since it is a
current
source output stage, I think I/we can figure out a way to test it
definitively.
Perhaps by interrupting the signal connector and inputting our own test
waveform. It can probably be a square wave, and once we see a good one
this
would be a definitive test. If the output voltage is lower with the
standard
input, then the reactance of the yoke is low, with everything else being
equal.
Or a spectrum analyser, but that is out of reach for most shops for it's
(now)
limited purpose.

Even with a scope, there are other possibilities for such a device. The
only
problem is we need a standardized test signal that is effective, a picture
of
the waveform produced by a new (or good) part, and a square wave
generator.
Perhaps one resistor beyond that, so you know the source impedance.

A spectrum analyzer would give us the data to build a proper filter for
the
DUT, but really knowing the scope has greatly enrichened my experience in
this
field. I could simply compare that voltage waveform to that of the red.
Hmmmm,
check it out, I don't need ANYTHING from the manufacturer to do this. No
data
no nuthin.

I'm on the M___________ and unlike Pulp Fiction, I'm not sending the Wolf,
I am
the Wolf.

Well, thanks for being bored enough to read all this and thanks in advance
for
any ideas, not only about the Toshiba, but figuring out how to test
things. I'm
for anything that enhances the repair industry.

JURB

Interesting. On this job one of the ICs did fail after replacement. The
original problem was the blue horiz channel, after running about ½ an hour, the
grreen vert channel fried.

I did notice the the loop they bend into the pins was off. It wouldn't let the
IC seat completely so I rebent them. I mounted them and then pushed each one
with a screwdriver handle to relieve the stress.

Perhaps that's a sign that you got the crappy ones ? I'll see if they're still
around and get the lot number. Of course with AOL dropping newsgreoups I might
not be able to post it . . . .At least right away. We'll see.

JURB

Just use the Toshiba replacements rather than the generics. Several
vendors, including MCM and Tritronics have reportedly had problems with bad
STK392-110 generics. The Toshiba parts are not that much more than the subs
and are the least expensive of all of the TV manufacturers supplying these.
I use the Toshiba replacements in all brands and have had no problems with
them.

Pioneer has subbed the STK392-180 in their sets that use the -110. I has a
higher power rating but is otherwise identical. Has anyone been using these
in other sets? Anyone got a good supplier at a reasonable price for them?

Leonard

"JURB6006" wrote in message
...
Interesting. On this job one of the ICs did fail after replacement. The
original problem was the blue horiz channel, after running about ½ an
hour, the
grreen vert channel fried.

I did notice the the loop they bend into the pins was off. It wouldn't let
the
IC seat completely so I rebent them. I mounted them and then pushed each
one
with a screwdriver handle to relieve the stress.

Perhaps that's a sign that you got the crappy ones ? I'll see if they're
still
around and get the lot number. Of course with AOL dropping newsgreoups I
might
not be able to post it . . . .At least right away. We'll see.

JURB