Vertical Bands on old Monitor

[vol.2]

I have an old CRT monitor that I use with 15kHz RGB stuff, but it's got vertical banding at a particular level of brightness.

I've replaced 100% of the electrolytics in it with nice new ones, so they are not a problem.

I suspect maybe the flyback transformer could be weak because I have to crank the focus almost all the way to the end to get it to focus properly, but I'm not sure that could even cause vertical banding and I've never heard of that as a possible culprit for this before.

My question is: What should I look for that could be responsible for the banding? (knowing that all the electrolytics are good)

Thanks

Here's a photo:

 

[vwestlife]

As discussed here:

http://www.vcfed.org/forum/showthread.php?48868-Vertical-ghosting-lines-on-Iiyama-CRT

 

[vol.2]

As discussed here:

http://www.vcfed.org/forum/showthread.php?48868-Vertical-ghosting-lines-on-Iiyama-CRT

Thanks vwestlife,

Sounds like a possible "bad yoke tuning caps which also causes ringing" or a grounding issue. I'll look more carefully at that post.


However, the flyback isn't addressed in that thread. Can anyone tell me if a weak or problematic flyback transformer could be the cause of said ringing?

Thanks

 

[Hugo Holden]

There was a generalized answer to this question here:

http://www.vcfed.org/forum/showthre...-screen&highlight=1084S+with+vertical+banding


The fault is not exactly like yours but has some similarities.

As noted in the other thread, generally there are only two ways this can come about, either modulation of the CRT beam current, or modulation of the CRT beam velocity. In the former case its is best resolved by looking at the CRT's grid & cathode voltages on a scope with a x100 probe. In the later case it requires examination of the H deflection yoke current, typically adding in a low R resistor of 1/2 ohm or less in series with the H def coils and using an isolated scope if the resistor cannot be placed with one of its legs at the ground or chassis potential, depending on the design of the circuit.

Of course , before the monitor is suspect, the video signals feeding it should be checked on a scope or with another monitor.

 

[vol.2]

Thanks Hugo,


That's helps narrow it down a bit.

This banding occurs without any source connected at all. If I turn it to "video" input and turn the brightness up a touch so that it's not a black raster, the banding appears. If you turn it up a little more so that you have a raster closer to "white" the banding is no longer visible.

There are two things that are anomalous to your description about it:

1) The banding appears across the whole screen, although it's stronger on the left

2) The banding only appears at a particular level of brightness. It's only visible when a scene appears with a solid color that is close to whatever range of brightness the ripple is on.


I'm looking through those parts of the schematic. I've already pulled and tested all of the caps and resistors in the H section preceding the yoke. Every electrolytic has likewise been replaced and the brightness circuit doesn't have
and ceramics or films. In terms of the G2 and Cathode, I'm not sure about the HV because I don't have my HV meter here, I'll pick that up from storage this weekend. However, I measured the G2 and it's more or less right on. The Focus voltage is a bit low though and it has to be turned all the way up to focus, so I think I have a failing flyback resistor.

Do you think a failing flyback could be the cause of this?


One other possibly related symptom is that the H-POS and V-POS are off for this monitor (by a small amount) and they are not adjustable. I've sometimes seen ringing in monitors when this is set wrong, so it makes me think that some value related to the size could be effecting it. I thought about even hacking in a rudimentary control for the POS, but I'm not 100% confident on what I would put in there. Perhaps a variable capacitor in place of the H yoke tuning Cap?

Thanks

 

[Hugo Holden]

Thanks Hugo,


That's helps narrow it down a bit.

This banding occurs without any source connected at all. If I turn it to "video" input and turn the brightness up a touch so that it's not a black raster, the banding appears. If you turn it up a little more so that you have a raster closer to "white" the banding is no longer visible.

There are two things that are anomalous to your description about it:

1) The banding appears across the whole screen, although it's stronger on the left

2) The banding only appears at a particular level of brightness. It's only visible when a scene appears with a solid color that is close to whatever range of brightness the ripple is on.


I'm looking through those parts of the schematic. I've already pulled and tested all of the caps and resistors in the H section preceding the yoke. Every electrolytic has likewise been replaced and the brightness circuit doesn't have
and ceramics or films. In terms of the G2 and Cathode, I'm not sure about the HV because I don't have my HV meter here, I'll pick that up from storage this weekend. However, I measured the G2 and it's more or less right on. The Focus voltage is a bit low though and it has to be turned all the way up to focus, so I think I have a failing flyback resistor.

Do you think a failing flyback could be the cause of this?


One other possibly related symptom is that the H-POS and V-POS are off for this monitor (by a small amount) and they are not adjustable. I've sometimes seen ringing in monitors when this is set wrong, so it makes me think that some value related to the size could be effecting it. I thought about even hacking in a rudimentary control for the POS, but I'm not 100% confident on what I would put in there. Perhaps a variable capacitor in place of the H yoke tuning Cap?

Thanks

Well the main control of the CRT's beam current is the G1 to K voltage. But if there was enough of a ripple voltage the other gun electrodes it could have an affect, though this sort of thing seldom happens to the focus voltage and practically never with the EHT.

The thing to determine is whether it is a modulation on the G1 of k voltage. In most sets the video amps drive the cathodes, so their power supply needs to be scoped. The G1's are normally common between the guns and should have a ripple free supply, but that needs checking. Its interesting that it goes away at high brightness, it might still be there electrically though, but hard to see, but it is a clue that there could be some interference on the G1 circuit that gets minimized when the source impedance drops when the control is at one end of its travel. Also check if it seems the beam focus is changing in the banding.

With the "H yoke tuning caps" the main one is in parallel with the HOT, if that wasn't right the EHT would be way off or the HOT destroyed as it controls the flyback pulse amplitude , also your problem is in scan time, not flyback time.Sometimes there are damping R-C networks across yoke windings to prevent ringing, these are for damping not tuning. But there is no reason you couldn't manipulate that value with a variable capacitor as an experiment. Also the S correction cap does not generally cause this sort of thing.

It needs to be determined if it is a beam intensity or beam velocity modulation. Put a small rectangular grid test pattern and see if you can detect any linearity error (compression) where the banding is. If not I'd suspect some interference on the G1 voltage or the power supply voltages to the video amp, initially at least.

Can you post the schematic ?

Some diagnostic tricks could include adding extra capcitance to the G1(brightness) circuit, or additional filtering to the HT supplying the video amps. But in the case of G1 you need to look at the specific circuit because sometimes the beam blanking is introduced there and not into the video amps, so then you need to increase the filtering on the supply to the blanking amplifier instead for that experiment, or it will kill the H & V blanking pulses, confusing the result.

Also, if say it is a design where the blanking is introduced to the video amps, or the G1 for that matter, its also possible that any modulation present on the voltage level between the blanking pulses, from any blanking amplifier, would be causing it too.

 

[vol.2]

In most sets the video amps drive the cathodes, so their power supply needs to be scoped.

I've got a scope, but don't have an isolation transformer. I decided to just buy one because I should probably have one anyway, so I'll do that after I get it. I can't scope anything inside it or do high voltage until then. I can use my DMM to get the G2 and the yoke output I think. Those should be low enough and the DMM is on batteries.

Also check if it seems the beam focus is changing in the banding.

Yes, I noticed that the banding seems to "smooth out" when you lower the focus voltage. It also gets a lot less noticeable when you turn down the "sharpness" knob. Turning up the sharpness knob makes the band much more noticeable the full width of the screen and increases the number of bands. I just assumed this is happening because I'm smudging the image, so to speak and that it wasn't really important.

It needs to be determined if it is a beam intensity or beam velocity modulation. Put a small rectangular grid test pattern and see if you can detect any linearity error (compression) where the banding is. If not I'd suspect some interference on the G1 voltage or the power supply voltages to the video amp, initially at least.

Ok. I'll give that a shot. FWIW, I noticed that the vertical is a tiny bit squished towards the top, so I think I have at least a minor vertical linearity issue. In order to fill the screen image, I have to overscan a little more than I should because the linearity biases towards the top of the screen. It's on the order of a half inch too high, but the squishing in the linearity is very minor and is hard to notice on regular content.
I have a couple extra vertical amp ICs that I could try instead, but I just was assuming that the issue was in the horizontal because the banding up and down.

Can you post the schematic ?

Yes, I have the whole thing. Let me know if there is a specific thing you're after besides what I post.

Thanks for you thoughts about this, I really do appreciate it!

 

[vol.2]

I also thought of something else. When I first recapped the set, I noticed that the two big power resistors in the 5V source (RE14, RE15 18k) were super toasted and had lifted the traces from the PCB. I replaced them with new resistors and reinforced the connection on the back side with the solder. I've seen stuff like that on other sets and audio equipment just from use, but I thought I should mention it in case it's a smoking gun that I just don't see because of my lack of experience.
You can see the resistors on the bottom left of the PSU schematic. in the last picture I posted.

 

[Hugo Holden]

I also thought of something else. When I first recapped the set, I noticed that the two big power resistors in the 5V source (RE14, RE15 18k) were super toasted and had lifted the traces from the PCB. I replaced them with new resistors and reinforced the connection on the back side with the solder. I've seen stuff like that on other sets and audio equipment just from use, but I thought I should mention it in case it's a smoking gun that I just don't see because of my lack of experience.
You can see the resistors on the bottom left of the PSU schematic. in the last picture I posted.

Probably no connection with those replaced resistors, unless, possibly, the originals were carbon or metal film and got replaced with wire wound types.

At least there is no worry about anything getting into the CRT's G1, it is grounded. So the video signals & blanking all come to the cathodes via the video amplifiers. You could check the collector voltage of any of those three video out transistors, they look modest so you could get away with a standard x 10 scope probe.


You could try increasing the filter capacitance on the screen supply to help rule that out, make sure to use a 2kV rated cap or more maybe try paralleling a 0.001uf 2kv cap with C902, the 330pF cap.

 

[vol.2]

Probably no connection with those replaced resistors, unless, possibly, the originals were carbon or metal film and got replaced with wire wound types.

I removed Panasonic resistors and replaced them with the exact same ones but brand new. I guess I was more concerned about if they burned up for some reason other than time and normal heat. One of the burned ones had drifted somewhat, but they weren't shorted or open.

You could try increasing the filter capacitance on the screen supply to help rule that out, make sure to use a 2kV rated cap or more maybe try paralleling a 0.001uf 2kv cap with C902, the 330pF cap.

Ok. That sounds like a plan. Don't have that hanging around in KV rating, so I'll have to order them.

Thanks again for your help. It's very much appreciated.

 

[vol.2]

I'm still waiting on some of the parts, but I discovered that there is an inductor coil L521, which has an incorrect value from the factory. The schematic calls out 4.9, which is unusual because most of the inductors do not have a value stated at all, so that makes me think the value could be important. I am assuming that 4.9 means 4.9uh.

The value of the inductor that it's in the L521 spot is (I think) Orange-Orange-Brown, which I think is 33uh. It could be red-red-brown I suppose, but that's still 22uh. I wonder if the incorrect component could be causing this?

L521 is on the 178VDC supply and is on the neck board. It attaches directly to the collector of all three color output transistors.

 

[vol.2]

To add to that, would you think it could be a problem if the inductance is too high?

Looking at the markings of L521 again, it seems as though it may be 330uh. It's brown multiplier, and orange-orange digits. I think that means 33 times 10.

If the inductance was too high, might that create a bigger drain on the 178VDC supply?

I think maybe it's reaching to think it would be a cause of the banding, but it does seem weird.

 

[Hugo Holden]

It might be something to do with that inductor......or not.

One move with this problem where you have a disturbance that is occurring across the horizontal scan timing interval, is to sync (trigger) one channel of the scope to a pulse from the H scan circuits. Then hunt around with the other channel of the scope on the video signal circuits, including the video output stages and the power supplies too, to see if you can find the same ripple corresponding to what you are seeing on the screen across the course of the H scan time.

 

[vol.2]

It might be something to do with that inductor......or not.

Well, I'm put a 4.9uH inductor on my parts list from the electronics place. Do you think there's any harm in swapping it out to see? I got a 250VDC rated part.

One move with this problem where you have a disturbance that is occurring across the horizontal scan timing interval, is to sync (trigger) one channel of the scope to a pulse from the H scan circuits. Then hunt around with the other channel of the scope on the video signal circuits, including the video output stages and the power supplies too, to see if you can find the same ripple corresponding to what you are seeing on the screen across the course of the H scan time.

That makes sense. I don't have a 100x probe, but it sounds like I can do that without touching the high voltage. There are two areas of the circuit that prominently interact with the horizontal scan in some way. The H output transistor and amp, and the H sync detect. The H sync detect feeds the Tuner IC, the Character generation chip, and the On Screen Display IC. It's used to position the OSD text properly on the screen, and to place it correctly in time during screen time.

To be honest, I'm not 100% sure about what the schematic is telling me about where the H sync detect comes from. There are thick lines on the schematic that have "A" and "B" written on them, and there are numbers next to them. The H sync detect goes to that line and says "B2." B2 also connects to the X-Ray protect circuit and to pin 9 on the Flyback, which also seems to go to the CRT filament.
I think the B2 is some form of AC coupling, but I've not used a schematic that had things noted this way before, so I'm a bit confused.

Where would you recommend grabbing the H scan signal from in the circuit?

Thanks

 

[Hugo Holden]

Anywhere from the H sync pulse input at the input connector, to the H scan osc in the set, which locks to the incoming sync, or even the drive to the H output transistor (HOT).

Keep well away from the HOT's output as the voltage there during flyback can peak in color sets to around 1kV, so you would need a 2kV rated probe to be safe for that test, but its not needed.

Mostly for what you are doing x10 standard scope probes are fine.IF you do it at the input connector the voltage levels are low and safe to test for the sync pulses.

 

[vol.2]

Thanks for the reply.

Ok. I got a bunch of stuff. I got my isolation transformer, HV probe, new flyback, a 100x probe for the scope and a bunch of components for testing.

New Flyback made my picture focus better, which is why I wanted it, but is having no help on the ripple.

I tried paralleling the .001uf cap on c902 (screen supply) and that made zero difference.

I also scoped the horizontal output transistor on the base and the collector. The Base looks low voltage based on what the schematic says. It says 21.7V, I'm getting 16, 17 max. I can't really see what's going on with the collector because the flyback pulse is basically keeping it around 1kV. Maybe I can try to dial in the lower voltage portion, but I think it's probably being eaten up by the flyback.

In any case, it seems like that 16V on the Base is suspicious, and there's an extra droop on the waveform that doesn't appear on the schematic.

Also, I traced the Horizontal back a bit and found that the 134V supply is down at 128V. I'm not sure how bad that is, but it seems wrong, and could be why the base of the big Horizontal output is 16V instead of 21.7? Seems like that whole chain is a bit low.

BASE

What the base should be in the schematic:

Collector:

 

[vol.2]

Let me amend my previous post by saying that the 134V supply might be okay. I was scoping the supply at 10ns and I had the trigger a little high.

If I look at the 134 Source at 10microseconds and put the trigger right on it, it looks like about 132V RMS. If I back it out to 100-250usec, it says 134V. I think there's essentially a ripple in the supply that goes from about 126-134 pk-pk, but I guess it's pretty fast.

I honestly don't know if that's "good enough." If you had any insight into that I would be appreciative.

In the meantime, I'll go ahead and work back to the horizontal output of the main video IC and scope the waveform later tonight. I'm essentially suspicious of that extra droop in the waveform, and the fact that it appears to be low.

 

[Hugo Holden]

I'm not sure from this information so far as to what is going on. See if you can find a ripple pattern on the scope, that matches what you see on the CRT screen as a way of finding out where it is originating.

 

[vol.2]

Okay. So I triggered the Base of the color guns and the supply with the Base of the H output transistor. I'm getting that same droop on all the colors.

It's a 15.63kHz negative going pulse that seems like it's a repeat of the horizontal. You can see it in two of the colors here:

This is the base of the horizontal again:

I didn't even change the cursors, the pulse just shows up at that spot when the sec/div are at 10us pretty much anywhere.

 

[vol.2]

Here's an image of the 178VDC power supply on the neck board. It's got the same 64us ripple on it:

I also scoped the 134VDC supply on the Collector of the H output, and it's got a significant AC ripple on it at 64us.