fdt-5bx5 Violently Exploding Capacitor
 

So, I got my little Watchman back with a raster, and I went about trying to set it up only to notice that the horizontal was quite shifted to the left.

I remembered that it was like this before I mothballed it a couple years ago, and there was some kind of an issue.

So I didn't know exactly what to do because there's no HOR Shift on this set, so I just set it up the best I could and kind of stretched the screen a little bit.

After it had been on for awhile, I was tweaking the purity and all of a sudden the image collapsed horizontally and a capacitor blew up inside the set.

It was C516, a 3.3uF bipolar cap, which was directly in line with the horizontal linearity coil leading to the yoke.

The service manual says the HOR yoke windings should be 0.5ohms, and I read 1ohm, so I think it's fine. I measured it and moved around the yoke and fiddled with it and the reading did not change at all, so I don't think there's a dodgy short.

Here's a picture of the cap and it's location, the part of the service manual showing the blown up cap, and a picture of the raster from before the cap blew up.

If you look at the picture, you can clearly see that the raster was pretty messed up as the circles on the right side are much smaller than the ones on the left; the horizontal linearity was all messed up, and no amount of moving the yoke helped it.



I don't think it was the capacitor's fault because I'm pretty sure it was already replaced by me. It's possible this happened before and I just forgot about it. It's been a long time since I tried to work on it.

Any thoughts?


https://i.imgur.com/o5EWdtE.jpg
https://i.imgur.com/T7pWf5S.jpg
https://i.imgur.com/YNLrG79.jpg

This cap in fact carries some quite current, so it can dry or explode ('nuff power for self heating). Can be simply a cap failure. Is worth simply to sub it for a new one. If it have some space, you can put a poly cap in place of it. Carefully watiching the image when CRT heats, and hearing if not make some strange whines.

In fact, before it explodes, maybe it's ESR are high, affecting linearity.

Well, I did actually replace this cap in the past, so it was a pretty fresh cap that blew up. I compared it to the ones I still have in stock and it's a direct match for them.

I used Panasonic BP caps. 3.3uF, 50V, same as the one that originally came out of there.

I tested the HOR output transistor, and it seems to test like a normal transistor. There aren't any other electrolytic caps in the area that look bad either.

However, perhaps the capacitor got damaged by whatever took out the 24V MOSFET? Maybe I should just try another cap

Hi to all,
Hi Vol.2,

could you please tell us where you found the schematic, i also have this model.
Elektrotanya doesn't have it.

Many Thanks! in advance,

Best Regards
jhalphen
Paris/France

Not all BP caps support continuour current flowing, so is interesting to confirm the model used; perhaps it don't support the current.
A poly cap will support it for sure.

I purchased it from Sam's a long time ago. At this point, I only have a paper copy I printed out of it, I'm not sure what happened to the pdf.

https://www.samswebsite.com/

Oh! I didn't realize that. I'm not sure how I'd be able to figure it out. The exact cap I used was this one ECE-A1HN3R3UB

https://industrial.panasonic.com/cdb...A0000C1053.pdf

For the 3.3V/50V one from this catalog, the maximum ripple is 25mA for 60Hz (120Hz). Certainly is higher for 15kHz, but is too far away...
Is good to take one with higher current classification, or to find the >10kHz ripple current. It needs to be higher than the current voltage product of the H circuit; for example, a small 5" BW draining perhaps 600mA from the 12V PSU, is good to use a cap rated at least for this current to be sure.
Some cap types have high frequency ripple specified.

Okay. Thanks for the explanation, this is new stuff for me.


I found a film capacitor, but I don't totally know what math I need to do in some cases. For example, it has a 3.3 at 100V, and it has no ripple rating. All it has is 20V/microsecond, given as a dv/dt rating.

https://i.imgur.com/NhHt0rR.png

The pulse rating if if it are absorbing pulses directly. Here is working as coupling.
Anyway, unless is a too much compact unit, poly caps in general will support the typical H sector TV current load.

Too bad that is difficult to find a nonpolar cap with full specs....

That cap was a common fail especially in Sony & Zenith B&W sets.
Wrong one will explode sooner or later. In the olden days we just
used the OEM, NP caps were hard to come by.

Zeno

Okay, that makes me feel better about it then. I am going to order up a poly cap that can take a lot more abuse, something on the order of 8A, and see if it can budge that.

Okay. I got a beefy film cap and replaced it and it fired right up.

After running it for awhile and doing some setup, the Vertical started to go, and I found that the capacitor between the pump-up pin and the output power supply on the vertical IC had blown.

This would not be a surprise to me, except that I already replaced that capacitor with a good Panasonic cap. This time I also replaced the 7830 vert IC because I have one in parts bin.

I was able to more or less dial in the purity and convergence, but I'm having some weird issues on the screen still:

1) I'm getting this effect where the top and bottom of the screen are kind of pinched inward.

Usually, if either the top or bottom is pinched in, the other side is bowed outward, indicating that the yoke is tilted. In this case though, both the top and bottom are pinched inward, so I don't know exactly what that means.

2) Horizontal linearity is still messed up. If I display a row of circles across the raster, the circles on the left side of the screen are correct, but the circles on the right side of the screen are horizontally squished. This is especially obvious in test patterns. This isn't a new issue, it was like this before the bipolar cap on the horizontal yoke blew up.

https://i.imgur.com/YNLrG79.jpeg

Anyone have an idea as to these two issues might be caused?

That type of capacitor was quite common in monitors that I repaired in the mid-1980s to the mid-2000s or so. In fact, common enough that local parts stores (when those still existed!) had them in stock, in different values.

They are/were called "high frequency non-polarized electrolytic" capacitors.

Because they are explicitly intended for operation in a horizontal sweep circuit, they are much physically larger than a "standard" non-polarized electrolytic capacitor of the same value and voltage.

Interesting. I replaced it with a beefy 3.3uf film cap. Do you think this could be causing the weird horizontal linearity issue?

For reference, it was like that when I had a bipolar electrolytic in there, before it blew up.

Based on your comment that it was the same, I doubt that the film capacitor would cause it. At lest one other monitor model I used to repair (Setchell Carlson/Audiotronics/Dotronix 23M922/23VM922) had a big film capacitor in a similar circuit location if I remember right. That one was 4.7uF/100 volts, I think, almost two inches or over 40mm long.

Okay thanks. I didn't think it would be an issue, though I don't know what could be causing it.

I replaced the electrolytics in the horizontal area already. There's also the linearity coil which I guess could be a bad value for the circuit and maybe it was always like this, but I can't imagine them thinking it's okay.

I wonder if I can improve the linearity by replacing the coil with something adjustable, or there's some other part of the circuit that I can tweak to improve it?

Maybe adding some capacitance to the big film cap?

https://i.imgur.com/T7pWf5S.jpeg

I wonder if this issue I'm seeing is caused by the magnetic core of the horizontal linearity coil losing it's magnetism over time? If that's possible, it would probably show this kind of a symptom I think.

If the reason for the magnetic core is to give the coil a non-linear response to correct for the asymmetrical energy in the flyback pulse, then losing it's magnet would nullify that correction, right?

I don't think it's feasible to replace this thing as they don't seem to make them anymore (there's a couple places that do custom jobs, but that's too expensive). I wonder if there's some other compensation I can do?

I presume you're talking about L503 and the label "MAGNET?"

This is the first time I've seen a magnetic core labeled "MAGNET," but presuming it's not a misnomer, maybe this is meant to be a saturable reactor?

I'm having trouble seeing how this sweep circuit works - what is connected off to the right?

I looked up saturable reactor because I've never heard that term before. Honestly not sure what the deal is. If what you are suggesting is true, perhaps the core is saturating and that is the reason for the distortion. If so, maybe there's some way to limit the current to the coil?

I did a bunch more tests and checked voltages, and the only thing I found that wasn't as per the schematic is the 170Vpk sine wave on the horizontal is actually about 208Vpk as I measure it on my scope. I didn't think that was any issue though as I'm sure I've noticed that voltage be similarly high in other sets.

This is what's to the right:

https://i.imgur.com/1eaI0Tk.jpg

OK, the flyback is to the right, powered from 22 volts, and the yoke is across a capacitive voltage divider. Makes some sense now.

A saturable reactor is meant to saturate. A transformer wound (two separate windings) saturable reactor was used in some motor speed controls. A DC control winding would change the core saturation and thus control the transfer of AC power on the controlled windings.I can't remember seeing a single coil meant to saturate, just trying to understand what could be going on here.

I was making a wild guess that a magnet could bias the core flux so that the inductance could decrease with heavy current flow one way and not the other, thus expanding one side of the image. Never saw a circuit like that, and really don't know if that could be what's supposed to happen here.

Is there any kind of mechanical adjustment screw on the magnet of L503?

Nothing like that. The only adjustment I get is horizontal size.

I found that there is a "Horizontal Width Adjust" transistor, Q380, that goes between the luma and some of the horizontal width stuff. I've never seen that in a circuit either. This set is filled with things that are a first for me.

It's a Japanese made Sony Watchman, but the tube is not Trinitron and was sourced from Samsung.

One thing I already tried was putting an extra .22uf capacitor across the S correction cap, but that didn't seem to fix it. I didn't notice much difference in the linearity, but maybe it was a bit more stretched out horizontally.

"Horizontal Width Width Adjust" sounds like a bad translation of something. I can't see anything adjustable connected to it, so maybe it's a blanking pulse shaper that "adjusts" (changes) the pulse width to the desired waveform (?).

Edit: I note that RV504 is labeled HORIZ SIZE, not "WIDTH."

Next thing I would try is shorting L503 to see what happens.

Also, have you checked D505 and D506 and the three caps connected to them?

I pulled both damper diodes (I think that's what they are) and they both seemed good. Test OL in one direction and .4V in the other direction on diode test.

The caps seem to be fine. They measured okay and are high quality panasonic film caps.


Is there a difference? When I change that, it changes the width of the screen.

L503 is rather large chunky magnetic core inductor with thick copper wire that is quite buried into it's structure. I don't see any way that I could shorten the wire without completely destroying it. I guess I could cut it somewhere in the middle and try to bodge it back together, but I feel somewhat wary to do that as I don't know how I'd keep it from shorting out.

Here's what it looks like:
https://i.imgur.com/Sko3Oga.png

There's also the small inductor, L502, but it's between the pincushion controls and not in the direct path of the yoke like L503 is.

1) Ok, diodes and caps are checked.

2) No difference between horizontal size and horizontal width in English. That's why I think the purpose of the transistor Q380 is a bad translation and has nothing to do with the horizontal sweep but probably the horizontal blanking pulse width.

3) L503 -- I meant shorting, not shortening, in other words, temporarily connecting a plain wire across the terminals of L503. HOWEVER, NOW THAT I SEE ITS SIZE, I WOULD NOT DO THAT AS THAT MAY ALLOW TOO MUCH HORIZONTAL SWEEP CURRENT AND INCREASE THE PULSE VOLTAGE BEYOND SAFE COMPONENT LIMITS IF YOU DO THAT. BETTER TO SEE IF YOU CAN FIND A REPLACEMENT.

The only source I know of for magnetic core inductors is a custom shop that will wind them for you, but they cost way more than the TV would ever be worth.

Not to mention that I guess at this point I can't even be sure that the coil is at fault. I'm just guess about it losing it's magnetism. In all other ways, the coil appears to be fine from the outside, so other than demagnetizing over time (which I don't even know if that could happen), I don't really have much to go on with it.

If you know of some way I could cheaply buy a replacement (ideally something adjustable) that I could try out, that would be great.

Or if there's any other obvious things to try of course.

Out of curiosity, could you test with a piece of steel to see if L503 actually is magnetized?
Again, I have never seen a permanently magnetized inductor, and that is why the notation "MAGNET" on the schematic is mysterious to me.

I found a paper on permanent magnet inductor design. The abstract says they are useful in DC applications, but L503 is being used for AC.

https://ieeexplore.ieee.org/document/5770892

Curiouser and curiouser.

Another paper talks about using a permanent magnet to cancel out DC effects in a power supply that produces both AC and DC in an inductor. Haven't found anything that talks about deliberately using a magnet in an AC inductor.

I was basing this off of the advice of an EE whom I interact with a bit over on the Vintage Computer Federation forums.

In responding to another person's question about this same kind of issue, Hugo wrote:

The post I took this from is here.

Wow - thanks for posting, and glad you found it.

[Bursting with pride for figuring this might be the case even though i have never seen one before. :D ]

I'm not a solid state TV guy, but I have a few inductors that are permanent magnet core (they pick up metal shavings and bits of component leads) that I've pulled off boards of BPC TVs I've harvested for parts) they are real and we're used in TVs.

Thanks for that info.