According to this Sam's Computerfact for the IBM 5151 C505 is a non-polar capacitor hooked up in series with the horizontal deflection coils, which means that large currents pass through it and therefore it must be a beefy low-ESR part (a common electrolytic would blow up nicely in its place). I'd recommend replacing it with four 4.7µF 50V+ ceramic capacitors in parallel for lower total ESR and increased reliability compared to electrolytics.

133MHz wrote:

I'd recommend replacing it with four 4.7µF 50V+ ceramic capacitors in parallel for lower total ESR and increased reliability compared to electrolytics.

Or two beefy 10uf poly-film caps likes these: http://www.digikey.com/product-detail/en/ECQ- … EF1106-ND/56416

Its worth the extra expense. High ripple and non-polarization is not exactly electrolytic's strong suit

133MHz wrote:

According to this Sam's Computerfact for the IBM 5151

THANK YOU! I've seen this manual referenced in serveral places but couldn't find it anywhere myself. This is really helpful.

133MHz wrote:

C505 is a non-polar capacitor hooked up in series with the horizontal deflection coils, which means that large currents pass through it and therefore it must be a beefy low-ESR part (a common electrolytic would blow up nicely in its place). I'd recommend replacing it with four 4.7µF 50V+ ceramic capacitors in parallel for lower total ESR and increased reliability compared to electrolytics.

I did notice it was a non-polar cap yesterday after writing that post, but didn't see it's connected to the horizontal deflection. Using an extra-low resistance part would make sense here! This would also make it prime suspect, since a "narrow" picture was the main problem I had.

How about using four of these? Or 2x5.6 + 1x6.8. I used the same types in my speaker crossovers too 🤣

I also noticed something else: the flyback is a bit "leaky". The clamp that holds the ferrite core together is corroded, and it looks like it arced over into the dust from there:

Let hope this doesn''t cause any trouble...

How about using four of these? Or 2x5.6 + 1x6.8. I used the same types in my speaker crossovers too 🤣

Those will be great.

I also noticed something else: the flyback is a bit "leaky". The clamp that holds the ferrite core together is corroded, and it looks like it arced over into the dust from there:

That is concerning. But if you are still getting raster, it is likely still okay.

"Narrow"? How narrow?

smeezekitty wrote:

That is concerning. But if you are still getting raster, it is likely still okay.

I think I'll patch it up with silicone grease for now, and first see if changing the caps has any effect. If I can fix it without changing the flyback, I think I'll still hunt down a spare one, just in case.

smeezekitty wrote:

"Narrow"? How narrow?

It only fills about 3/4 of the screen now. Vertical is still good.

This all started when suddenly the screen size and brightness increased very quickly (in less than a second). Then I got a short, very bright green flash from the entire screen, and the horizontal width and position started changing randomly and very rapidly. After a minute or two it settled down slowly and now it's stuck at about 3/4 of the total width. Now the brightness also won't turn up as far as it used to, normally there would be a point where the entire raster is illuminated but it won't even go that far anymore.
About an hour later the picture started jumping around all over the screen (without changing in size), this lasted for about 10 minutes and eventually it settled in the middle again. I've used it for several days afterward without any weird things happening, but the narrow picture and low brightness remains.

Received a box with electronics parts today, including caps. Possibly the most overkill types you can imagine. Now I just finished soldering them in.

This is how the large non-polar cap ended up, I think it's the neatest looking way I could get all these film caps to fit:

Of course, the question still remains... did we incidentally fix anything?
The answer is: no. Nothing changed, it's still narrow and rather dim, same as before:

So with that, I can conclude my little experiment: "30 year-old Taiwanese-made electrolytic capacitors should never be trusted?" --> I'd say, MYTH BUSTED! 😀

Now I can start trying to repair this thing. The width and brightness offsets are probably minor enough to correct with the internal control pots, but then that wouldn't really fix anything, would it? And these pots don't turn themselves, so something else in there must have failed. I think I'll start following the SAMS manual and see if I can find anything unusual. Any other suggestions on what may have failed would be very welcome, too!

Check if the B+ voltage is within spec, when it gets low you usually get a narrower, dimmer raster.

Yep, that's it! It's down to 11.6V (should be 14.9V)

When I desolder everything from the B+ rail it's still 11.6, so no shorts. Resistors and diode stuff seems good, and I tried replacing the big transistor with a new MJ15004 I found in my parts pile. Still 11.6. So the only remaining option is that the 7815 just.. "upgraded" itself to a 7812.

Do you think this could have affected both horizontal and vertical deflection separately? The weird part is that it did two different things spaced about an hour or two apart. I would expect to find at least two failed parts.

I'd replace C603 and C409 just in case. The regulation stage is so simple that there's very little that can fail, and using the shotgun approach might be quicker than actual troubleshooting.
As for your second question, both the horizontal and vertical stages are powered from the main B+ rail so it's likely that it was a cascading failure - some component fails in a particular stage putting additional stress on the stage behind it, which then fails, propagating backwards - much like the massive blackouts we've been seeing around the world in the last few years, which are usually caused by a small overload causing a chain reaction.

133MHz wrote:

I'd replace C603 and C409 just in case. The regulation stage is so simple that there's very little that can fail, and using the shotgun approach might be quicker than actual troubleshooting.

Yeah I've replaced all the electrolytics already, and will leave them in. I think it'll be good as new again with a new 7815, but I'll have to order one first since I don't think I have any of these on hand.

133MHz wrote:

As for your second question, both the horizontal and vertical stages are powered from the main B+ rail so it's likely that it was a cascading failure - some component fails in a particular stage putting additional stress on the stage behind it, which then fails, propagating backwards - much like the massive blackouts we've been seeing around the world in the last few years, which are usually caused by a small overload causing a chain reaction.

Would make sense, but then vertical didn't actually fail completely... It fixed itself about 10 minutes later...

Got some new toys from China 😀

That's the big non-polar 18µ cap from the horizontal drive circuit. No idea how accurate these meters are, but both say all the caps I pulled from this monitor are as good as new... 🤣