vertical lines in a Toshiba TV
 

I've got this Toshiba set that died on me and I recapped most of it because it needed it. It seems like it ran quite hot in it's life and there were multiple burned areas on the PCB

I went through and replaced any bad diodes and resistors in the areas that were burned and stuff, and tested and replaced anything near a heatsink.

After all that, it works and generally looks fantastic and bright, but there is a visible vertical line pattern in the raster that appears most strongly in solid colors, and most strongly at a particular brightness. I would say just roughly guessing it's about 50 IRE bright.

I think possibly what happened is that I used some higher heat resistant caps because I wanted to be safer about it, but that maybe one or more of those caps are higher ESR or something and they are causing the lines somehow. Maybe not doing their job to suppress them perhaps.

Problem is I don't know what to look at here as I did them all at the same time, and I didn't notice the issue until a bit later.

Any ideas?

https://i.imgur.com/DcOXxUF.jpg?1

https://i.imgur.com/biQLo7I.png

https://i.imgur.com/67bzE3k.png

https://i.imgur.com/rDw4mWZ.png

https://i.imgur.com/qOzbO2Q.png

https://i.imgur.com/qz0wDus.png

Is that picture taken with no signal connected? If so, try connecting a signal and see if they dissappear. This will tell you if it's an RF interference or a conducted interference from horizontal into video on the chassis.

If it disapperas with a normal strength signal, I'd just forget about it. If it's still there, then you might try chasing ground connections between the circuits and also wire dress.

Thanks for the reply.

It's 100% most definitely in the picture when a signal is connected.

I tried dressing some stuff, but not everything. I covered the entire flyback and also the hot lead into the tube with copper electrical insulating tape and it made no difference I could see.

I don't know how to "chase ground connections" exactly. Any guidance on how to do that?

Check ground connections from varioius circuits to the chassis to see that they are secure; resolder if needed.

I've reflowed the entire board already for sure. I don't think this is a loose connection thing. I suppose there could be a dodgy connection inside of a connector somewhere, but I kind of doubt it.

That could be what they called "jail bars" caused by a bad cap.Did you change c446? Its in the 178 v source for the video outputs.10uf 250 v.a common problem with fly back winding derived dc power sources.Hope this helps.RonL

DITTO
Start with C446. Its an all brand common problem. May as well change it on almost any SS set. Symtoms can be very subtle.

Zeno:smoke:

I definitely changed that cap. I think I changed basically anything in the power supply, deflection area, and the RGB stuff.

Perhaps I need a better cap than the one I used, but it's a good Nichicon variety. Like I said though in my first post, I think I upped the temperature rating of most of the caps, so that might have reduced the ESR and leakage characteristics of them. I think usually higher temp resistance comes with an ESR hit.

I could try replacing it with a different one, or maybe try monkeying with the value to see if I get a different result.

If so, do you have any suggestions on what to try?

Thanks

Edit: I think it's worth mentioning that I also see the bars if I directly insert an RGB signal into the RGB mux chip on the video processing board. I did this as an experiment to see if I could inject RGB via the mux chip in place of the OSD signal and blank the screen to show a higher quality signal (I succeeded), but I did notice that I was getting the bars in the image this way too. This would leave me to believe that it's nothing about the composite input board or demodulation circuit that is effecting things at least. This set is a Toshiba, so it uses one of their optical isolator chips to get the composite video in without the need for an AC transformer to isolate the set (the optical IC has a separate ground).

If it's not coming in on the previous stages, and all the caps are good, where else could it be but on grounds is what's puzzling me.

I think I would make a quick test of bridging a cap across C446, and triple checking it's the right value, and maybe bridging it to a different ground point on the circuit.

C446 was a 10uf, 250V cap @85C. I replaced it with a 10uf, 250V cap @105C, Nichicon brand.

The ground side of the cap was solid and read as close to zero ohms as anything else on the ground plane. I removed the chassis board and made sure to measure it from multiple points on the chassis. The ground is solid.

Edit:

I tried doubling up C446, and I'm not totally sure, but it might have made a little bit of difference. It seems like the bars might be lighter. It may be possible that the original cap was a different value from the Sam's. but honestly, I've run into that too many times, and I always made a spreadsheet of each cap to make sure I put in what came out and note any discrepancies, so I kind of doubt it. Would increasing this cap's value significantly help things? I have a 100uf @250V I could put in there, but I don't want to screw anything up if that might be a mistake.

Do you think I should try sticking the 100uf cap in to see if it helps?

Also, I want to reiterate that this bars issue only shows up in the image of a certain brightness value (and therefore, I would assume voltage value). It's somewhere around 50 IRE I think. If the image is brighter than that, it won't show up, and if the image has bright areas and dark areas, it's only visible in the medium dark areas, and general only really visible in solid colors like you see in most video games.

"Also, I want to reiterate that this bars issue only shows up in the image of a certain brightness value (and therefore, I would assume voltage value). It's somewhere around 50 IRE I think. If the image is brighter than that, it won't show up, and if the image has bright areas and dark areas, it's only visible in the medium dark areas, and general only really visible in solid colors like you see in most video games."

If the cap is the problem, doubling the value should halve the contrast of the bars. Sounds like that's not happening.

Peculiar that it only shows up at medium gray level. How does it vary when you turn the brightness up and down with a blank screen? Worse at low brightness, medium at medium brightness, goes away at high brightness?

Yeah, I can't say if it's halved or not. It seems to be less, but it's hard to tell. They are definitely still there.

The bars show up when video mode is turned on, but no source is connected as well. I can induce the bars by turning up the brightness on a black screen. In fact it's most noticeable doing this, and it seems to be most obvious when the screen is grey. Turning the brightness up past a certain point makes the bars go away. They might still be there (I assume they are) but I can't see them. I guess you could also say that I can see them clearly enough all the way down to just before a black screen.

The bars are wider on the left side of the screen and get very narrow on the right side.

Sorry, no more ideas at present.

Trying to imagine/figure about it...
Definitively, is very difficult to estimate where the periodic interference are being injected. Things like already said here crosses my mind, and also about ground or reference tracks being (almost) cracked, almost invisibly, near opening (high resistance).
Or, most unlikely, but not impossible, the jungle IC with a strange internal fault.

Not impossible, but not likely. I reflowed the whole board for sure. If there is a crack I could see, I think it would be covered over. I could check again.

I have already tried replacing the jungle IC. I have three of them, I could try the third one I guess. Maybe both the first two are bad, but I sort of doubt it.

I've also replaced the flyback trying to fix it. I found the original Toshiba part NOS.

Another thought crossed my mind, is that maybe this fault was here in the original design of the set, and I didn't notice it before I recapped it. When I first got it, there were so many bad components (basically all the caps in the deflection and PSU section) that had to be replaced to get it running that I didn't have a good baseline.

If this is true, and the set normally has these bars in low grey areas, is there someway to cure it by adding components?

I tried a few things already like putting low value caps on the neck board between the Grid and the Cathodes, and I did some shielding for the flyback, but I'm not sure if what I used was enough.

Yes, it's possible this is a design flaw with this chassis, but it's so clearly visible in your photo, I wonder. I presume the bars don't move wildly, just get stronger or weaker when you change brightness?

You might try running added ground wires between boards, not following the path of the original ones, to see if it changes. (Still grasping at straws here.)

Another thing very unlikely for a SS set, but anyway: the image are compressing at the bands positions? Tube sets sometimes makes that. Just seeing another thread. But, by design, SS systems are less prone to that.

OOPS, nevermind... it varies with bright, cannot be this cause.

No, the bars don't move at all. They are completely static. Depending on the brightness, they can appear to be either lighter lines on a darker background or darker lines on a brighter background, but I believe that's normal and is just an illusion; as the brightness value gets higher, the lines seem to disappear because (I assume) the voltage level on the raster overtakes the resolution of the lines (causes it to bloom over the bars)


I could try that. I have to admit it seems like a long shot, but why not.

Yeah, also, it definitely doesn't warp a video image at all. I see no compression.

In fact, if a busy image is on the screen, I can't see the bars at all. It's only noticeable in areas of solid color or grey.

"as the brightness value gets higher, the lines seem to disappear because (I assume) the voltage level on the raster overtakes the resolution of the lines (causes it to bloom over the bars)"

Now I wonder if we're looking at the same thing. I have been assuming you have been talking about the fairly narrow dark and bright vertical bars that extend at least a fifth of a way into the picture from the left. These are far too wide for spot blooming to average out unless you are seeing the picture go terribly out of focus at high brightness. They could, however, be a power bus or ground transient coming from the power supply that reduces in amplitude as the power supply is loaded more heavily.

I see the same thing you are seeing.

They do, but then you can see there are also some narrower lines on the right side of the screen. There is one lighter thin bar towards the right. It seems like they are wide on the left, and get narrow towards the right until they are very thin. I would assume the lines are wider on the left because the energy from the flyback is greatest just after the refresh period.

That was just a guess. I was describing what happens when I start from a blank screen and slowly turn up the brightness until I see a white raster.
It made sense to me that the raster itself was what is blooming as a whole solid screen of white, not the spots surrounding the bars.

How do think I could solve that if it was true?

Thanks

To be fair, sometimes I seen (and have) some sets with some artifact. I have one Toshiba with horrible black clamping and inomogeneous near black screen. But is not much the near black artifact, and is easy to dismiss it, unless someone are playing a proper bright/cutoff adjusted Resident Evil game (even the ones from Playstation 1)... but I not dismiss it's black level unstability (very sensitive to contrast; it can be a strong CRT currrent limiting or suchlike). I did NOT solve it, since I'm using it now only for the Atari (then is indirectly solved due to usage ;-)

.

Hi - The lines on the screen line up perfectly to point to a horizontal drive pulse, or
possibly a harmonic, or ringing etc. I don't think it's injected anywhere, but it may
be on the power supply lines, any of them. Do you have a scope? You may need
one to find this. I would look at each of the voltage sources for this 1/15734
or multiple showing up on voltage source lines. All these sources are also
providing bias for the picture tube, & it's drivers, so if it's in the power
supply, you can bet it's in the tube when no picture is present to begin
with, and it's still there with picture information. So set your scope
to capture several H pulses possible set sync to the H pulse,
and look for itsy bitsy H frequency ripple in the power supply.

You see your power supply has lots of choke's in it along with
caps to filter out flyback noise at each source point.

I doubt this came from the factory like that, the Japanese
are very meticulous about how their stuff operates....

.

As Alex says, a scope might help. Are you sure the frequency is changing from left to right? I can't think of something that would do that.

Wild thought - try connecting something like a 0.1 microfarad 400 volt capacitor from green output collector to ground and see if it attenuates it. That woud tell you if it's coming through the video circuits somehow.

Design flaw ? NO WAY !
If you got a scope. Set to display a BLANK, not even snow pix.
Be sure the G-2 is set proper. Color level at zero.
Look at the @200V should be clean. Same with the 3 video outs.
There almost HAS to be crap on them.
Check the E of the 3 outs for garbage also. Trace back.
There is a little PS that runs the EXT for isolation, Check that.
There is an "IF mute". didnt trace but probably turns off IF's when
on ext..
Dont forget the blanking stage also for a look-see.

Zeno:smoke:

Yes, I have a scope. I have already scoped the power supplies and the video. Didn't see anything fishy.

Yes, I had come to this conclusion already. That's why I said the pulse is strongest right after the horizontal refresh.



They were all clean when I checked. I can see the flyback pulses, but nothing that looks like it would be happening during scan time.


It's 178V. I can see the flyback pulses in it, but they are outside of scan time. Nothing like the ripple or otherwise bars I see on the screen.

Don't see anything.

The emitter? Not sure if I did that.

Are you talking about the optical input chip's isolated ground?

That's a little bit above my head. I would need more explanation to grock it.

.

During post Vol 2 above you said -

They were all clean when I checked. I can see the flyback pulses, but nothing that looks like it would be happening during scan time.


Quote:

It's 178V. I can see the flyback pulses in it, but they are outside of scan time.

Nothing like the ripple or otherwise bars I see on the screen.

end Quote.

If there is a H pulse on it, then that's what is on the screen. That's why you have to
look up-freq. from the H-Pulse, Look for a harmonic as there are several light &
dark lines on the screen. If there is a harmonic it may occur during scan time.

Anyway if you see stuff on the 200V line, how are you sure it's not happening
during scan time? Anyway if it's happening on both sides of the screen then
it's also happening during the damper scan too. To me that points to H-Pulse
or harmonics on Power Supply lines.....

If there is for example 10 defined lines on the screen, then it's possible that
there is a noise source 10X 15, 734 on the screen....

.

Because the pulses are spaced properly and the waveform matches the service manual perfectly.

I think it's obviously the H-Pulse ringing that's getting through somehow, but I don't see it on the power supply lines because they all the waveforms look "correct" as per the manual, and they make sense in that the pulses from the flyback are spaced properly. The electrolytics for the power supply filter caps have all been replaced with new, high quality Nichicon caps

.

Happy Sunday Morning! Have you by chance scoped out the posted wave forms on
the other side of the flyback, Q402 & Q404, H-Drivers for proper size & Shape
with no out of the ordinary spikes in rising of falling sides of their pulses...?

.

A few things just in case.
Sams wave forms show 2 complete lines or frames.
The little isolation PS runs off the FBT. Check its filter cap.
Watch the lead dress, if to close to FBT it can cause this.
Remember cathodes high = dark pix, low =bright pix.
Scope should show a damped ring , refreshed (ping) at start of every hoz line.
That should show up on scope. Like I said it almost has to be there !
Also remember there are 2 or 3 grounds in this set.
AND since you re-capped check the polarity & value of every cap ! This is why
in most cases repair first then change a few caps & retest & so on.....

Zeno:smoke:

I'm not sure which one you mean. Which pin on the flyback is isolated? 162V (pin 10)?

FBT pulse goes to T501, rectified by D501, filtered by C502.
Note some of the ground symbols are different. Gotta use the right one when measuring.
MOST of the set uses one hot GND. If the set was newer it would
probably use a SMPS and that would add a GND. But with those
the chassis GND is cold.

Zeno:smoke:

To you too!

Q402 looks a bit weird. Q402, Q404, and the Video IC have all been replaced with new parts to check for any difference and none was found.

Also, the waveform on the other side of T401 looks correct. The rest of the set is not connected to Q402,404 besides the 134VDC supply, but even then only through a number of other components

All the parts in that whole section have been removed and checked for values with an LCR meter.

Ah! Okay, yes, I know about that. That's what I meant when I said the AV input. It's isolated by a optical input chip and then it uses T501 for its power I guess.

I did replace that filter cap, and I can't see any garbage in that circuit.

As far as I can tell, it's not in the video. I can inject and RGB signal into the Jungle IC and completely blank the video signal and the bars are still there

.

" Q402 looks a bit weird. " - What do you mean Gracie ?

.

The top of the square wave has an indentation and there is some flagging.

https://i.imgur.com/XdFYX2w.jpg

.

Hi So if that scope pic you posted is from the Base of Q402 then it loos like to me, and
my Squirrel companion, that there is a good bit of overshoot on the rising edge. I don't
know if'n yer schematic picture is of better clarity than what is posted, but I don't see
the same overshoot there..... But before we go there..... since we are looking at a
square wave thingie, does your scope have the test calibration post, where you
put your probe on it and then adjust the variable cap on the probe to adjust
it's LC balance so you don't have probe induced overshoot on your scope picture?

So if all that is sorted out, then you may have found your source of noise that
you are seeing on the screen..... See that decaying oscillation at the rising
edge is much higher in frequency than the horiz frequency, so that may
be making your "Jail Bars" looking stuff.....

I think, and the Squirrel Concurs, that you should poke around in that area and
see if there is a bad soder joint or something else, since you said you replaced
all that stuff over there.....

Good work, on that stuff, and finding a possible cause....

You may be able to add a very small .pf value ceramic cap
to the base of that transistor to see if you can dampen
it out a bit, and if it does, dampen it on the scope,
then see if it does it on the TV screen as well....

PS - While you are in that area, you should
check all wave forms, and check voltages
too & Please post any differences in
detail. - Thanks & Good Evening.

May the Squirrel be with you!

.

Yes, the probe is fine. It tests fine on the calibration source on the scope.


I went down that route a couple years ago (I've been working on this for awhile) and all components in that area test good and all the solder is good. I've been in there and had everything in and out multiple times, undoing and redoing the solder joints. Most components have been replaced, including the main video IC, and absolutely nothing has changed. All the scope readings are the same as they were when I started.

The only parts in that whole section I haven't replaced are T401 and the two inductors L418 and L412. I have spare parts in a donor TV set that is identical to the one I'm working on which I intended to try out, but it's in storage and I probably won't get to it for awhile.

In any case, the transformers and inductors are not likely to go bad as they are very simple components. I also tried using freeze spray on all the components in the section while scoping waveforms and I saw absolutely no movement or change from anything. The inductors were tested with a good LCR meter and read correct values, which is the best I think I can expect from an inductor.

Well, thanks for your input, and the Squirrel as well.

I hope I can figure it out, because if what you are saying is correct, I don't know how to make any difference. I can try a capacitor, but there is already a ceramic cap C425, 0.0022uf in that spot, and C405, 680pf, upstream from it on output of the horizontal signal from the video IC. I have already removed them and tested them and they are spot on.

.

Well Don't give up, that overshoot has a nice oscillation on it, to me that is an LC item
so maybe it's a LC pair that should be grounded, and is not. As mentioned earlier,
check out grounds, Find an LC Pair from the IC to that driver and be sure it's
really grounded.

If you disconnect the "1" voltage source on T401 will the signal on the base of Q402
clean up? Or disconnect the non grounded side of the secondary of T401 will the
signal at the base of Q402 clean up, or will the entire thing just not start up?

Since you removed almost all the components, you have to always keep
in mind you might have made a mistake putting stuff back, I think you
should check it all over good one more time....

Pin 59, 62, 61, & 63 on that IC have the right wave forms, & Voltage
readings? I'm not sure if a bad crystal could cause that.....

Is D401 A Zener Diode? Is it properly regulating the voltage to the
IC Chip? Looks like pin 58 Should have 7V on it....

Do you have 88V on the collector of Q402?

Did you change R411, R413? By any chance did you change it from a Carbon
to a Wire Wound, or Film type, or Visa Versa in any way?

Is that overshoot visible coming from the IC? (Pin 64)
or just at the base of the transistor?

I did not see anything, but may have missed it, what does
the notes say about the gray shaded components in the
circuit?

PS - A Long shot but while checking every last thing...
Might be good to be sure no leakage from Pri. to
Sec. of T401 you never know....

Have a Good Night!

.

.

If all else fails, you could spread out that overshoot part of the scope pic & measure
the frequency of the oscillation & make yourself a filter to ground it out.

.