Hi. It's part two repair time here of this uh, Compact Ibm. Well, the world's first Ibm Pc compatible machine. It's the Compact Portable uh dates from 1984 and I'll link in the video up here somewhere and down below at the end if you haven't seen it.

Anyway, we were debugging this thing. We got down to the point where the processor was actually reset. so we actually uh, traced the problem down to the basically the 8088 processor was permanently in reset and the reset comes from the 8284 uh clock driver chip. Here this is the main 14.31818 megahertz uh clock that is divided by this to give you the 4.77 meg in here as well as other stuff.

and the reset comes from here and the reset pin. That one there actually came from the power Good signal from the power supply over here which is in the bottom of the case. so that's what's stopping the processor from powering up the rails. The five volt rail 12 volt rail seemed okay, but apparently there is a power good signal coming from it was the second pin here on the power supply connector.

It was coming from in here. So let's get this power supply board out and have a squeeze. Now it uses just these. I have to take the mains input out here that's actually welded soldered shut actually that thing so it's probably just got a filter in it or something like that.

Um, hopefully no reefer caps that have blown up. But anyway, mains does get in and it does have these little clips here which then could the board can slide out, but it can only do that if you actually take off. Not sure if you're gonna be able to see it in there, right down the bottom, there in that corner. Focus your bastard there.

It is this little metal, um, springy clip that I had to undo, which, uh, like it's like a locking pin or something like that that locks the whole power supply in place. It's rather, you know it's clever and nice, but bloody annoying. Anyway, we should now be able to slide this supply out these connectors. I can't damn well get them off.

they're really hard. But anyway, that's the plan because everything else in this machine sort of slid out. Everything's designed to be modular in this machine, but they kind of like didn't implement it very well. Everything I tried to get out of this had some clutch to it and it was clunky to get out and I didn't like it.

Anyway, I'll get back to you, it does come out somehow. There we have it. There's our board still attached to the umbilicals. Uh, but let's at least we can.

I get in here and have a visual inspection because the first thing you want to do well, first thing you want to do is smell it of course. But visual inspection, make sure any of the caps aren't failed because tag tantalums like these ones here. And these ones here. These are all tag tantalums down.

Nope. down here as well. Absolutely famous for catching on fire and one of them actually did. When I switched this on, it blew up.

I see the previous video for that, but not on this power supply. did it on the uh, disc controller card. But anyway, we've got electrolytic caps and they all look, uh, really good. They've got the vents on top, none of those are bulging.

Nothing fancy going on there at all. I mean, you know it could be a ripple thing or something like that. They could have, uh, certainly dried out and have not, uh, died yet. But so they're markham jobs.

Uh, the orange ones, not sure, but the silver ones are Sprague made in Usa. they'll be fine. But no. of course, if you were actually refurbishing this thing and you were serious about, uh, you know, having this thing, uh, work for any sort of extended period of time, you would recap absolutely everything.

All the electrodes, all the tag tantalums and I didn't actually notice that before. Um, the silk screen would have been handy if they put it on the other side, but the silk screen is buried under here. Sure enough, that second pin is Pg or Power good and that's the one that's low. It's active low and that's what's causing our processor uh to reset.

but there's other rails. I've only measured the plus five and plus 12 which are fine, but we've also got uh, minus five as well and minus twelve, but they're not needed for the processor. But if the power supply of course, has any, uh, sort of you know, sensing stuff, it likely does to check the power rail. So even if the negative five volt is, uh, buggered, then it's going to just reset the processor even though the processor doesn't need the negative 5 volts.

So you know better safe than sorry. I guess it's just really annoying. I mean, we can override that. We could simply, uh, lift the pin the second pin on there and power this thing up and see if it boots.

I guess there's the back side of the board there. It's a little bit crusty, so I'll get in there with some isopropyl and uh, clean that up. But you know, you might look for, uh, dry joints. uh, you know, thermally cracked joints.

Actually, you know thermal expansion of the parts, uh as they power off and on over the years and that can cause problems, particularly on like, things like you know, the To3 power transistors and stuff like that. So you want to get in there with a little jeweler's loop or a microscope and just inspect all those. And if you're wondering what I use to have a look at things like this that I can't take over to say my mantis or my Tagano microscope, I just used my Times 10 Macro lens. This is an Optica.

this is what all my close-up shots that you see. I just screw this on the front of my camera and Bingo gives me my close-up shots like that. But you can actually just hold this near it and you can actually get look you can get like really good close-ups like that. It's you know it's It's really quite nice.

Highly recommend it. And sure enough, you can see what looks like possibly a cracked joint there. that's like on the nut that holds the To3 in place and the other one up there looks like it's maybe something similar, but I actually measured those and they seem fine. but that could be a potential intermittent source so you'd want to resolder those as a minimum so just powered it back up.

And of course, the 5 and 12 volt rails work. But sure enough, one two three, four, five pin number five or fifth ones from the end that's supposed to be uh, minus five volts and minus 12 volts is not there either. So yeah, both negative rails. Gonski.

Um, and that's why Power Good is not coming on. So we have to look at the negative rails. and here's where it gets like really annoying. Um, they're in here.

Um, so and it's like just populated with all these physically tall and dense uh parts. These are probably the output filter caps for it. The Lmr 338. There that is a positive voltage regulator, so that's almost certainly not.

uh, the negative rail there. But yeah, um, little switching jobby in there. perhaps for the, uh, negative rails? So it's just really annoying from an access point of view. And as always with Murthy, if you have a look, uh, the fifth pin down there, that's actually the negative five volts, which I'm trying to trace.

And of course it doesn't go on the net bottom does it so you can't see where it's going to it? I can barely see down in there. So actually, before I muck around on the power supply, I thought I'd actually power this board up with the bench supply and then just start change the reset pin. So I've got it hooked up to the bench supply here. I'm only generating the 5 volts and 12 volts.

In fact, probably don't even need to generate 12 volts and by looks of this, I've turned. Uh, it's on at the moment. but 12 Volts is drawing nothing. So like the -5 and stuff like that is not really needed for just getting the processor and stuff like that working I believe.

Anyway, it is drawing 1.6 amps here and if I actually, uh, reset it, put that to ground 1.519 and switch it on 1.577 so it draws a little bit more uh, power when it's uh going. But anyway, let's measure some clocks, shall we? There's our reset pin. It's low. There's our clock pin.

that's uh, that'll be four point. You're 4.77 megahertz. There it is. Yet no workers.

and uh. let's look at some bus activity, shall we? Let's just randomly. Yep, that looks like bus activity. Bus activity.

More bus activity? No, whatever that pin is. Sorry, I can't remember all the pin outs off the top of my head, but uh yeah, this looks like classic bus activity. so it's certainly doing something. So what I'm going to do now is plug in the video card in here.

because it shouldn't. Well, the video card might need the positive and negative rails. Um, Anyway, let's just plug it in and see what's what. All right there we go.

I've plugged it in, haven't hooked up monitor yet, but we're getting significantly more current. draw. Look at that. The video cards draw in huge amounts.

I mean, like we're now four amps. That's just. that's drawn a lot more than the processor is and we're drawing something on the 12 volts and I switch it on and get an absolutely nothing on the video output there. So Zippity Doodah.

Okay, we're still getting bus activity there. That's uh, processor bus activity. So it's not like it's been loaded down or anything like that. But we're certainly getting nothing on the video out.

But I don't have those other rails so that could be it. Okay, I've got uh, minus five here and minus five's drawing nothing. Zip. Okay, I got uh, minus 12 as well.

So I've got all four rails there. Just set them to nominal. like half amp, current, limit. whatever.

So whoa. Now we're going in the current limit. Negative Rail there. Oh no, No.

There we go. That's good. Maybe that was just a power on thing. So reset it because I don't have a power on reset.

So they're both drawing 0, negative 5, and negative 12. So, and I'm getting nothing on the videos, so that's still not going to help me, even if I fix my main power supply. Um, it's pointless because this thing, while the process is working, the video is not so. Once again, Like as I said, uh, I'm not sure if you have to.

maybe set some jumpers or something to get a composite video out. Maybe I can check, uh, some of the Rgb? uh, out here? perhaps? Okay, I have no idea which pins what? Hello, that looks like a sink That's definitely sinky. There we go. Yep, Oh, that's 50 Hertz.

Yeah. getting stuff there. you go. 18, 18 and a half Kilohertz That certainly looks like the horizontal.

Okay, it looks like we are getting horizontal and vertical sync there. There we go. We're getting something in there that looks maybe like some video data. Probably only one color.

Because um, that's all you need When you boot up, you're probably just getting like a text. Uh, you know, a monochrome, like one color text image. So that sort of makes sense. So it looks like that works.

So this, ah, I think this machine's working and if we actually hooked up the Crt, but I've got to put all the boards back in. I think to get all the Crt working again to be a mess to try and do that outside of the case, I think. So yeah, that's a bit tricky. I wish I could get that damn composite working.

And by the way, kids don't try this at home. I'm a professional now apparently. Uh, dip switch five and six on the main board here. These actually set the uh startup video mode and it's currently set to uh, Cga 80 column mode, so that's interesting.

I might try Mda, which is both of them off, so I will do that and repower. Okay, I'm getting the same horizontal and vertical there, so 18 and a half k, but that data does look significantly different. It was less populated before, but I'm still getting no composite output signal at all. Now here's something interesting.

I was just playing around with measuring the signals again because I've got a capture card, which we'll talk about in a minute. but I'm using the case of the Uh composite output connector here as ground and I was using that before to actually probe. But now if I actually probe these signals, let me just do a resetti let's actually probe these signals again. Look at this, see how it's just dropping down there.

It's got some 50 Hertz crap, all sorts of stuff on there. right? Look at that, you see, and then it's just. it's just dropping down and down. And here's a sink.

see. Look at that. It's just going down and down. and down and down.

Now What? When you see a signal like that and look, it's just rising up. When you see a signal like that, that's indicative that the ground is actually doing nothing. So if I disconnect the ground and measure that again, you'll see it's doing the same thing right. So our ground is.

Why is that no longer grounded? Uh, Mueller, Mueller. So if I actually measure that compared to what I know is a ground pin, here it. it's just not connected. It is not connected at all.

That is like, what? What The it? What? I, I swear it was before. That's why I used it. I would have measured it. It's actually been some quite some time since I shot this video because I just acquired a Um Cga to Vga converter card and I was just about to hook up all the signals.

I was determining where the signals were. And I do what. And sure enough, if I measure it um, this bracket and of course, which is electrically connected uh, to the ground of the composite connector and also to the shell of the Uh Cga output connector, then uh yeah. I'm getting like, um, nothing.

It's just like like 40 50 meg something like that and then it sort of like goes open compared to ground. So I have no idea how I was measuring those signals before. It has been a couple of weeks, but I like, I don't know, um scratching my head anyway. um what I did is went and ordered a uh, Cga and Ega and yuv to a Vga converter and this is actually uh, quite nice.

It's got the Vga in it's got uh, the component in as well and it should be able to do the Cga hopefully. Um. and and it's got a very came with a very nice little cable which then I've connected up to uh, pin headers here so I can just slide these on. Gotta find which pins is what? Again, I know that that one's ground is in five volt ground.

I've buzzed that one out. um so I'll just uh, re-measure that now using that as the ground pin and uh, anyway, yeah, we'll get these signals and we'll see if we can get a Cga output signal from this thing. Actually, it's really quite frustrating that there's no really easy ground points on this thing that you can just hook your scope probe up to. Um, you know, without danger of you know you can go on to maybe one of the pins of a tag tent or something like that.

You know that's why, um, you know you generally like pick a, uh, a ground connector like a a, You know, a bracket like that. So when you're designing boards like this, just make sure you include like a a nice handy ground reference loop or pin or something. Unfortunately, I'm getting zippity-doo-dah on this. no signal.

Um, so I've got to muck around with this thing. But the menu? Um, yeah, I can't make heads or tails out of that. I might need a cheat sheet so we'll see. Um, there is an auto button.

Maybe I can press and hold auto or something. I don't know I am getting. I'm getting horizontal sync, vertical sync, and I'm getting something on the green signal. So yeah, I don't know.

There's some adjustment pots in there. Maybe bugger. I just read the manual for this thing. Um, or the specs on ebay or whatever and it tells me that it's only 14.5 kilohertz to 16.5 kilohertz.

We've got 18 and a half kilohertz here. Don't win a winner chicken dinner, green screen and all. I had to change the mode of the board to Rgb S. Um, and that worked.

a treat. So there it is. Um, I don't know what these are. Okay, I assume the error codes: 301, 401, 601.

Um, you know we've got nothing. And of course there's diskette error. but this thing works. There you go it, boots.

Winner winner chicken dinner. Um, well, as you'd expect because we were seeing video signals there. So yeah, no worries. But hold on to your hat.

If I use the ground down on this board and I probe our composite output, we now have a signal. Why weren't we getting a composite signal before? look at that. That is going to work a treat. So if I hook this monitor back up to the composite out, I have no doubt that'll work.

What the whoa glitch in the matrix there? So just change this back to Av here and let's see if we. oh oh yeah, it's not as good is it? It really doesn't like that. that's yucky yucky yucky. Um, so yeah.

modern monitors. They're probably not very good at this sort of thing. Meh. who knows.

But uh yeah, that's working. A treat. And as we saw before, it doesn't take any power on the uh, minus five and minus 12 volt rails. So really, to get this thing to boot? Um, I don't even need to fix that power supply.

Uh, in this thing, all I need to do is force that or break into that line. Force it. Uh, high. And that power good line? Force it high.

and assuming that all the Crt circuitry and everything works, it'll It should just work a treat. Or at least the composite output will work. and we'll be able to see that. at least we can read it I guess.

Okay, I think the reason that the composite is working now is because I have been mucking around with these Dip Switch settings and I which changes the graphics modes and I think there's only one that seems to work. Oh okay, it just takes some time to boot. Let me let me show you that. Go like that and there's no composite output for quite some time at time.

This puppy. come on. you can do it. and it will eventually get there I believe.

or it did last time. So come on. So maybe I was bit too impatient before about checking for video signals. There we go.

Yeah, it eventually just pops up. Yep. so maybe you know it's going through memory checks or something like that before it displays any before it even enables. Like the video signal is not even enabled.

Um, until literally the text pops up so there's no sync there. There's no nothing. Okay, I've confirmed that the only mode that actually seems to work is Dip Switch uh, 5 off and 6 on. which is I believe a Cga 40 column mode and that's the only one that will give me the composite output and work with this uh, Cga to Vga converter card as well.

So anyway, I'm going to leave it set to that so that at least gives me a baseline to get this puppy up and running from now on. But uh, really, I didn't have any doubt that uh, the board would have worked because I found with these vintage computers like just you know, when you're talking about like Ttl stuffs unless you have an exploded tantalum cap like we had in the first video. something like that. really? the you know, the the silicon is actually quite robust and you know, even after all this time, the uh mask rom uh still works.

you know that's all just fine and everything's hanky, the ram in it still works, and everything's hunky-dory So um, yeah, I might have to look. I'll look up those codes. But anyway, like the thing works, the processor works, everything all the video card works. and yeah, everything's sweet.

So no real surprise, I knew I was pretty confident. You know, I was like 90 confident. When I get this board out and power it up from bench supply, it'll probably boot. And sure enough, so there you go.

I'm going to leave this video at that because I've run out of time today. I'll do the power supply troubleshooting another day, but we at least got this puppy up and running using the external Um monitor output here and I have checked There are signals on the Cj Mondo uh, Cj connector as well, so I should be able to. You know, if I assemble the thing and the Crt doesn't work and I can't get that working or I don't want to get it working, I can at least hook on an external monitor without having to, you know, hack into the internal uh header connector here. But there you go.

I hope you like that. Um, so yeah. I've got so two more unknowns as well. I definitely know the power supplies faulty, so I need to fix the power supply.

But as I said, like the minus 5 volt and minus 12 volt rails on here, draw absolutely nothing. So in terms of getting this thing working unless you want to use like a serial port or something like that, maybe where it might use that, um, and those voltages I don't you know, or the disk drives and stuff like that, uh, possibly. and also the Crt as well. They may only be for the Crt, so you know, who knows.

but uh, yeah. So to get the So, probably to test the Crt, I reckon I'm probably going to need to fix the power supply first, so that needs to be the next step. And then, uh, check to see if the Crt works and get this whole thing, uh, restored eventually. It's one of those time things.

Anyway, that's it for today. So if you liked it, please give it a big thumbs up. As always, discuss down below: catch you next time you.