Hi;

I don't recall the exact model right now, but it's got the digital
convergence. Set came in with serious errors mainly in the red
horizontal channel.

The trained eye could tell it was not the output IC, no foldover, no
massive offset. So I set the convergence.

It has gone bad again. Tomorrow is going to be one of the rare
occasions I go on the road. I suspect I will simply set the convergence
again (in front of the customer, oh joy), but the real cure needs to be
found.

Somebody correct me if I am wrong, but those PA000X ICs do not store
data onboard correct ? Data comes from the EPROM. Thing is, how in the
heck can something only be affecting the red horizontal ?

Right now I think tomorrow I am going to focus on whether the error is
exactly the same as before. From the description it sounds the same. At
this point I wonder if something is forcing the parameters back to some
virgin factory state or something. Like a default.

The most whacked out parameter was horizontal linearity, and now the
complaint is "the red won't move over far enough". I remember the Hlin
value was like -45 and it came into alignment as it approached 0. If I
find the same thing tomorrow, well we will have another piece of
information.

I'd appreciate anything you can tell me about these sets. Due to the
cost of prints and the fact we don't see all that many of them I don't
have service info. I drew the conclusion that the PAXXXX ICs don't
store the parameters because I have swapped red and blue on other sets
to see if the problem followed and the set did not require
reconvergence. Logical right ? Well, sometimes logic doesn't work.
Logic also says that "**** happens" when data are involved, look at all
the RCA EPROMs some of you haven't had to change due to the chipper
checker. The part simply isn't bad, all it took was some bad data
coming in, or even a static discharge to a data line at just the wrong
time. Watch me find a bad ground near the tuner on this Pioneer now L O
L. ( I doubt it )

Anything you got, really. This is a mid-late 90s set, and they are
pretty much the same. A feature here and there may differ, but I know
they kept the same design for a few years as far as the convergence
goes.

Thanks in advance.

JURB

Reflow several hundred joints on the convergence board. Seen this several
times, including a Pro97 that I used to own. Could still be the output IC,
too, but with that symptom it is almost always bad joints.

Leonard

wrote in message
oups.com...
Hi;

I don't recall the exact model right now, but it's got the digital
convergence. Set came in with serious errors mainly in the red
horizontal channel.

The trained eye could tell it was not the output IC, no foldover, no
massive offset. So I set the convergence.

It has gone bad again. Tomorrow is going to be one of the rare
occasions I go on the road. I suspect I will simply set the convergence
again (in front of the customer, oh joy), but the real cure needs to be
found.

Somebody correct me if I am wrong, but those PA000X ICs do not store
data onboard correct ? Data comes from the EPROM. Thing is, how in the
heck can something only be affecting the red horizontal ?

Right now I think tomorrow I am going to focus on whether the error is
exactly the same as before. From the description it sounds the same. At
this point I wonder if something is forcing the parameters back to some
virgin factory state or something. Like a default.

The most whacked out parameter was horizontal linearity, and now the
complaint is "the red won't move over far enough". I remember the Hlin
value was like -45 and it came into alignment as it approached 0. If I
find the same thing tomorrow, well we will have another piece of
information.

I'd appreciate anything you can tell me about these sets. Due to the
cost of prints and the fact we don't see all that many of them I don't
have service info. I drew the conclusion that the PAXXXX ICs don't
store the parameters because I have swapped red and blue on other sets
to see if the problem followed and the set did not require
reconvergence. Logical right ? Well, sometimes logic doesn't work.
Logic also says that "**** happens" when data are involved, look at all
the RCA EPROMs some of you haven't had to change due to the chipper
checker. The part simply isn't bad, all it took was some bad data
coming in, or even a static discharge to a data line at just the wrong
time. Watch me find a bad ground near the tuner on this Pioneer now L O
L. ( I doubt it )

Anything you got, really. This is a mid-late 90s set, and they are
pretty much the same. A feature here and there may differ, but I know
they kept the same design for a few years as far as the convergence
goes.

Thanks in advance.

JURB

I have found in most all of the sets I worked on with the STK4277-SL
convergence output that they seem to have a life of about 5 to 7 years then
either completely die or go intermittent. Don't be afraid to gently peck on
the IC or heatsink to determine if it has a loose connection internally.
Check ALL the joints on the convergence panel, every Pioneer I have seen has
cold solder somewhere. The IC is not massively expensive either, and in the
interest of having no recalls I generally recommend replacement "just in
case" if the set is over 5 years old. The PAxxxX IC's do not store data,
they are the DAC's. Also on Pioneers, and especially on Toshibas, check the
heatsink and make sure it was cut flat on the mating surface with a
straightedge. I have seen a few where the center of the sink was cut too
deep making the IC contact only on the outside. You can usually tell when
you remove the IC, just look at the pattern left by the heatsink compound.
Heat kills, so if you have no gap at the ends and .005 in the middle (where
most of the heat is dissipated) taking a sanding block and some ultrafine
sandpaper and providing a flat surface may be a good idea if you want the
repair to last as long as possible.

Jammy
wrote in message
oups.com...
Hi;

I don't recall the exact model right now, but it's got the digital
convergence. Set came in with serious errors mainly in the red
horizontal channel.

The trained eye could tell it was not the output IC, no foldover, no
massive offset. So I set the convergence.

It has gone bad again. Tomorrow is going to be one of the rare
occasions I go on the road. I suspect I will simply set the convergence
again (in front of the customer, oh joy), but the real cure needs to be
found.

Somebody correct me if I am wrong, but those PA000X ICs do not store
data onboard correct ? Data comes from the EPROM. Thing is, how in the
heck can something only be affecting the red horizontal ?

Right now I think tomorrow I am going to focus on whether the error is
exactly the same as before. From the description it sounds the same. At
this point I wonder if something is forcing the parameters back to some
virgin factory state or something. Like a default.

The most whacked out parameter was horizontal linearity, and now the
complaint is "the red won't move over far enough". I remember the Hlin
value was like -45 and it came into alignment as it approached 0. If I
find the same thing tomorrow, well we will have another piece of
information.

I'd appreciate anything you can tell me about these sets. Due to the
cost of prints and the fact we don't see all that many of them I don't
have service info. I drew the conclusion that the PAXXXX ICs don't
store the parameters because I have swapped red and blue on other sets
to see if the problem followed and the set did not require
reconvergence. Logical right ? Well, sometimes logic doesn't work.
Logic also says that "**** happens" when data are involved, look at all
the RCA EPROMs some of you haven't had to change due to the chipper
checker. The part simply isn't bad, all it took was some bad data
coming in, or even a static discharge to a data line at just the wrong
time. Watch me find a bad ground near the tuner on this Pioneer now L O
L. ( I doubt it )

Anything you got, really. This is a mid-late 90s set, and they are
pretty much the same. A feature here and there may differ, but I know
they kept the same design for a few years as far as the convergence
goes.

Thanks in advance.

JURB