LET'S GO!!!!!!!! (Patreon)

IT LIVES ONCE AGAIN AND IT'S MY DOING (THOUGH MAYBE I MADE IT WORSE FIRST)

I'm really proud of myself. Mark the date - this is the first time I, the Cathode Ray Dude, have ever repaired a CRT. Ironic, right? And for several reasons - I've known basically how to do it for decades, watched a thousand videos, and even diagnosed others, but never actually done the work because, frankly, high voltage scares the holy hell out of me. But I got over myself and did it. In short, this means the next video SHOULD be about this device and you can take that to the bank, but for those curious about the details:

As I think I mentioned in the last post, at last report this VT had been working until it suddenly vertical-collapsed. If you don't know, that's when a CRT picture suddenly becomes a single line tall, and while it's a bummer, the nice thing is that it's pretty easy to diagnose.

This is the CRT control circuitry (usually called the "chassis board") and while it can look quite complex, from the actual circuitry perspective there's only a few components on here that're doing any major lifting. Mostly the stuff you're looking at are the diodes and transistors, of which there are less than a dozen parts - the VT100 service guide even says this:

Since the circuitry is pretty cleanly divided into horizontal- and vertical-deflection sections, a good step would have been to work through all the semiconductors in the vertical section, but I never got that far because I'd heard that these monitors are notorious for cap death. Especially a weird little guy in the dead center (the black one, above, though that's after I replaced it) with a 75uF / 6V rating, which is apparently well known for exploding, sometimes taking out the diode next to it in the process, so I needed to replace that before I did anything.

It's not actually part of the vertical circuitry and thus can't be the culprit, but since I was changing that one I just went ahead and ordered a whole set of replacements. This mostly went well, except for the part where I maybe blew up one part while trying to fix another.

Someone's going to slap their forehead before I make it through the next paragraph but oh well; what happened, happened, so here it is.

If you look at the previous pic you'll notice there are two Big Boy capacitors in the center left. One is a 1000uF / 16V cap, a pretty sizable one - and that's the smaller of the two. The bigger is this gigantic sky-blue number that, at 10uF / 25V, ought to be a tenth the size. Indeed, it ought to look just like the identically-specced one above it in this pic, which I ordered from Digikey.

Capacitor technology has improved over the decades, so it was possible that it just got easier to make one like this tiny, but that's probably not what's going on here. Notice the "HF" and "NP" nomenclature? The NP part is easy, that means it's nonpolarized, more often known these days as bipolar - you can see the "BP" on the replacement indicating that. A normal electrolytic cap has a polarity; install it backwards, and it'll likely spill its guts all over your PCB. A bipolar/nonpolar cap can be installed either way, and does the same thing in both directions. I think internally it basically consists of two normal caps back to back, but don't quote me on that.

Both caps are NP, so that's not a difference - but the HF part may be. I couldn't find much info on this, but I assume that means High Frequency, since this sits in the middle of the 15.5KHz horizontal circuitry. The new cap is not HF - but I wasn't exactly sure if that mattered. Again, cap technology has changed a lot over the years; maybe what was HF in 83 is now Normal Frequency?

To be honest, I didn't really believe that logic, but I wanted to use the cap I'd bought, so I installed it. Looked at it, saw the little tiny cap with its legs stretched out to fit the wider footprint, thought "oh, that's clearly not right," but... sigh, yeah, whatever. Put the thing back together, very gingerly fired it up, and... nada. Nothing at all, no whine, no bang, no activity whatsoever.

I immediately suspected the new cap, and also suspected I'd never needed to replace it in the first place. This part is definitely in the horiz circuitry, so it would not have caused vertical collapse, and if the horiz sweep had also failed, the previous owner would have seen an unmoving dot, not a line. So this was not the culprit behind that failure, and when I'd checked it on my multimeter it came out at 10uF, exactly as it should. That isn't a reliable test, but Really Bad Caps tend to read low, so my gut says this one was just fine all along and I didn't even need to replace it.

As a footnote, a friend who knows more about this stuff told me that the reason I couldn't find one of these is because they've largely been replaced by film caps, so I should have just gotten one of those, but under the circumstances I decided to just put the old one back in, hoping that the incorrectly-specced new cap was just throwing off the circuit so badly that it couldn't oscillate. Unfortunately that didn't fix it.

My next step was to go through the entire board checking semiconductors - they weren't bad, but maybe now they were. Daria, for her part, pointed right at the TO220 transistor in the upper left corner of the board and said "I bet it's that one."

Yeah, it was that one. Of course it was. This BU407D is the HOT - Horizontal Output Transistor - which chops the voltage to the flyback, the exact part you'd want to check in a no-sweep-no-HV situation. I was getting 43 ohms base-emitter and emitter-base, which aforementioned knowledgeable friend says is typical when these go: "they become a 50-ohm resistor."

He also suggests that, yeah, it's my fault that it's dead. This part of the circuit can do interesting things if it isn't able to oscillate correctly, so it's very possible that as soon as I turned the thing on with that wrong cap installed, this thing got itself into a State and ran too much current through the base.

Another footnote: I actually saw a dead short when I checked it first, and triumphantly announced (with fist pump) that I'd found the problem... then sheepishly discovered the short still there after removing the transistor. Turns out it wasn't a short after all, it's just that the base and emitter of this part connect to two ends of the secondary on the adjacent transformer.

I thought transformers characteristically had pretty high DC resistance due to the length of the windings but apparently I'm thinking of something else (motors? probably motors) because when I rechecked I found that it wasn't a dead short, but about 0.3 ohms, 0.1 of which was from my probe leads. So it was a total red herring which still turned out to be right.

Anyway, I was preparing to wait another few days for another delivery from Digikey, but then someone suggested I try Vetco in Bellevue. I completely forgot they sold components, and sure enough, they had a perfect matching part in stock.

Amusingly, I had to go through two layers of indirection to find it. I called, but was told their stock is all NTE (iykyk) equivalents, so I had to look up the BU407D in a cross ref; it was an NTE379, which they said they had in stock on the phone. So I show up, ask for them, and the guy brings me a pair of Motorola BU406s. Actually a better part, and certainly workable, but... why didn't they say they had that on the phone? And where's the NTE part?

Turns out, NTE went out of business four years ago, so now they're buying equivalents to the NTE equivalents and stocking them that way. And this worked out really because they really didn't have a 407 in stock, and I wouldn't have known to ask for the 406.

Anyway, I brought it home, popped it in, and this is what the first power-on looked like! Well, after I hit SET-UP, but anyway.

I'm incredibly proud of myself for overcoming my fears to stick my hands in this thing and figure out it's problems. Even if I maybe made it a little worse in the process, it now works perfectly (it's been burning-in on the table behind me for 40 minutes) and of all the tubes I own, this is the one that needed to be fixed the most.

Maybe this will embolden me enough to go work on my NEC Multisync 1... but anyway. I'm gonna post again once I get the thing booted - the previous owner included a boot disk but (shocking nobody) it's no longer working. The FDDs might need some cleaning, or the disk might need rewritten or replaced, but either way I'm sure it won't take much.

Thanks everyone!

P.S.: