Hi. Check it out. This is the Tandy 102 Portable computer. One of the world's first and most popular Uh portable computers of all time.

This is the upgrade uh to the classic Tandy 100 from 1983. this is the 102. Data about this model came out in 1986 and it's identical to the Model 100. It's just a little bit thinner and lighter.

and anyway, this was, uh, basically the most popular notebook computer in the 1980s. Had a modem building every single like, uh, you know, reporter journalist had one of these things because they could type up their stories, connect it to their acoustic coupler modem, and then dial back in and send their story back to the news desk and stuff like that, let alone countless other Uh uses. But I've done a teardown video of the Original 100, but I also have a Tandy 102. We can see some classic uh, yellowing of the Uh bromine in the plastics here.

It's a fire retardant, it causes like them to like go yellow over time and that's fairly typical. But anyway, I did a tear down of the 100. This is the 102 and it's just it used to work, but it's now got a rather unusual fault. so let's take a look at it and see if we can fix it.

Oh and by the way, the Tandy 100 is famous for being one of the last computers actually containing significant code written by Bill Gates himself. So he actually worked on the rom for this thing. So the original, uh, Tandy 100. Anyway, let's switch it on and uh, you can see it appears to actually work.

Copyright: Microsoft text, Telecom address that had like a spreadsheet, a word processor built in, and basic building and everything. But you can see up the top here. Um, there's something weird going on. It's like there's you know look.

You can see the clock up there. It's here. 54, 55, 56. But it's all sort of weedy.

Something's going on, this doesn't look correct. and well, let's go into the basic here and let's have a look. Uh yeah. Model 100 software.

Once again, this top corner up here. something weird is going on. We've got two flashing cursors. This should say you know something bites free, right? Um, I don't know how.

I can't remember how much memory is in this one. but yeah, something has gone wrong. And Microsoft there. What's all this going on? Something weird, right? Let's clear this.

Okay, so everything looks to be working just fine and dandy. But let's have a look. What happens here. Okay, what I'll do is, I'll just, uh, like one, two, three, four.

So everything seems to be working. Five, Six seven. But watch this, We go to Eight. Oops.

Once we went past eight, the cursor has wrapped back here and it's starting to overwrite. So Nine, Zero, One two, three, four, five, Six Seven Eight. and it just keeps overriding like that. And it's no coincidence that it's doing it after the eighth character.

So because you know there'll be an eight-bit bass that's shifting, whatever it might store like, you know, one column or something as a byte. or you know, whatever, it's no coincidence that the number is eight. and uh, if we go down here like this, you can see once we get actually down to the four, uh, the fifth line down here. the cursor wraps up here and it duplicates anything on the fifth line.

and then it duplicates anything on the fifth line on the first line, and then once again continues to wrap. And if we go down another line, it'll duplicate it. So the, uh, the eighth line down here will duplicate it on the fourth line isn't that fascinating. So there's obviously something to do with the how the dart is getting into the display drive.

The display drive is obviously working because the correct characters are showing up. It's like it's basically garbage in garbage out. Pretty much so, the actual Uh, column and line drivers and everything else of the Lcd seem to be just fine. so we can pretty much, uh, rule out that as an issue.

It's it's the data being fed into them. and this Lcd is going to have multiple, uh, driver chips for it. Uh, probably like there's probably going to be like one chip controlling like the first four lines here and the first eight characters. and they've probably got another chip controlling that one.

So it's probably that chip up there that's causing some sort of issue. Anyway, Luckily, we do have the schematics schematics. I need schematics for those playing along at home. This is 1986.

Uh, 10th month I presume. Um, it's like I presume, that's a lowish serial number. If anyone knows, please let us know. Product of Japan.

All the best stuff's made in Japan. Oh, and if you don't know, it has a memory protection. uh, switch because this didn't have any of that non-volatile rubbish. It was volatile.

Sram. It kept your uh programming when you switched it off. Uh, so good old Sram. And um, yeah, you could protect that or not with your batteries.

And oh, there we go. You can plug in expansion ram there, so that looks like it has an additional 62256 in there. Oh, okay. other stuff, if you haven't seen it.

Rs-232 serial, a system bus connector, a printer connector, their phone. that was your modem because it does actually have and you know, a proper isolation uh. transformer in there and your cassette tape. uh, storage.

It had a light pen. Uh, you know, good for like, inventory control for warehouses and stuff like that. Uh, that's the direction. Answer: originate.

Uh, for those who are used to your um. modem sex, you'll know all about that. All right. I think I've got this mostly off that should lift off.

We're in. like Flynn. There we go. and the keyboard on this feels beautiful by the way.

it's just. ah. it's fantastic. So there's our Lcd up there.

whether or not we can get. oh yeah, yeah, we can flippity-doo-dah You betcha we can flippity-do that. Ah, that's just beautiful so we can still use this. Let me switch it on and we should find that.

Yep, yep and picked up where we left off. Oh look, oh, that's it. I hadn't tried that. It didn't actually look.

It didn't actually refresh the contents up there. That's interesting. Ah, like. it kept everything else, but it didn't refresh that.

But when it's getting new data like it is now, I mean, if we go down a lot, there we go. No, it's updated. but yeah, it cleared it. That's fascinating.

All right. I know you want to see the rest of it. Um, I can't remember how this differs from the well yet. No, this is.

I do believe this is very different from the 100. Yep, um, this is this is the revised model. Thinner and, uh, lighter weight double-sided load. Look at that.

wow surface mount. The other one was all through-hole wasn't it? Anyway, we've got some bodge wires and some bodge caps. Um, the the bodge wires look like big ground stuff. It's an inductor or a diode.

Not sure. You can see the red glue underneath the chips there, that's to hold them on. Ah yes, I remember this from the previous teardown. This is absolutely fantastic.

This is some Pcb routing perfection Here, you'll notice that, um, there's no markings on this chip. That's because it's mounted upside down. It's mounted in a cutout in the Pcb. Why? if they put one row of them like this and the other like this? This is for Pcb routing reasons.

Just the way that the pin outs worked. If you had these chips up the other way, like these ones, um, it'd all be higgledy piggery and you need extra layers on your board to get around at all, It'd be an absolute mess. So from a Pcb layout, elegance point of view, they've mounted these chips backwards with a out in the board. It's just it.

It's simply brilliant. It really is. Hats off. Luckily, we do have the service menu, complete with all the schematics, the theory of operation.

the whole works. They don't make them like this anymore, unfortunately. Tandy 102 Custom manufacturer for Radio Shack and division of Tandy Corporation. Look at this.

Look at this. Ah, just a wet dream. Maintenance, disassembly, instructions, Theory of Operation troubleshooting. Oh beautiful.

I doubt. Though I actually haven't looked, I doubt they'll have a troubleshooting procedure for like this sort of fault in the Lcd. But yeah, you never know. Look at this.

so comprehensive. look. Look fantastic. Bobby Dazzler.

Anyway, there's the back of our Lcd board. and look at this Ta-da Here's the Lcd driver board. Look at this. We have two rows here because this is a 40 column by eight line Lcd.

And obviously each row of these chips here will be handling four uh lines. And we've got five for the columns. So 40 divided by five eight. Tada, Is it just simply this chip? Well, if I'm flipping it the right, whichever way it's you know, actually flipped.

Yeah, it could be M1 M1 up there. That could be it. So I yeah, right off the bat you would suspect uh, this driver. Here, it's a Hitachi because they did all the they still dominate Lcd drivers, don't they? I don't know.

the standard Hitachi chipset. Anyway, The Hd 44 102 for those playing along at home. So automatically, you would either suspect this chip has failed or it's getting bad drive signals coming into it. There could be a chip select.

There could be an address line. you know, whatever could be some data corruption, but you know it's obviously clocking. No, it's obviously getting the data in there correct. So it's got to be some sort of, you know, like some sort of chip select thing.

But it may not be that chip. It may be off on the main processor board which actually drives it. This specs for those playing along at home. Look at this.

four. double A batteries last like 20 hours. You can use this bad boy for. Unbelievable.

Anyway, fantastic Ada C85 processor. So we've got theory of operation here: Lcd Lcd Common Drive Lcd Segment Driver Lcd Waveform Uh, the block diagram. Yep, Okay, so an 81c55 drives into the Lcd controller and then the Lcd. But it also like it comes via the Cpu bus as well so it doesn't all come via the Pio.

So the Pio is probably only like uh, driving the chip select lines and things like that. They don't make theory of operations anymore, do they? Ah, we'll get to the there's the modem interface. I do want to do a video trying to use the modem on this thing. I think that'd be really cool.

There's the modem connector interface circuit. Um, I've done some experiments before in the past. Uh with I can't remember this 102 or the original 100. But uh yeah.

Anyway, Lcd here we are. The Lcd using Attendee 102 is composed of electrodes in a matrix arrangement. 64 common signals, 480 segment signals. This part is subdivided in three segments: the common driver, the segment driver, and the waveform.

The common Driver Here it is. That's the Hd44103 uses two common driver Ics okay, M11 and M12. Where was that thing we had before? So yeah, there's your common driver there and down there. I don't think it's going to be a common driver because well, it's common.

It's more likely to be the segment uh drivers. Now I don't think we're going to have to go into like the internal logic of the 44103 and the other uh, common driver. I don't you know? it's just it's just getting. I think it's just getting the wrong stuff clocked into it.

I think that is the problem. It could be via any number of reasons. There could be a dry joint somewhere, one of the driver chips could be failed, or something could be wrong with something. Could else could be loading down the bus, Which is, uh, going to be common.

Who knows. But uh, yeah. I don't think we're going to have to get into that sort of detail. That's sort of like the last rabbit hole we want to jump down.

Show us the schematic Dave. Okay, after the bill of materials here, it is. No. Isn't this just brilliant.

Absolutely brilliant. Fantastic. Takes a second just to render a new sharper version. but anyway, all the goodness.

it's all down here, right? So what we want? Here's our Lcd connector here. Okay, okay, we've got some Cmos logic doing some stuff here. Not worried about that yet. Let's have a look at the address and data lines here, right? So let's follow this here.

Id. So this is the uh, this is the data bus. It's not actually the address bus these chip selects are the address. So if we follow the data here, okay, this comes over here.

It's going. Yep, it's coming. It's all the work. It's coming from here.

which, uh, let's not go there. but let's follow the money up here up here. Up here here. you go there, you go We've got a 245 there and that's hooked into the 8085 processor databus right? So I don't expect there really like to be any issue with the processor addresses and data buses because everything hangs off that right so it doesn't.

Everything else works so I don't expect that to be an issue. So that's kind of like the last thing that you would look at here. so I'm not worried about the data bus there. all the data seems to be getting in, it's just getting latched wrong.

So there's uh, looks like there's all these Well, the thing is it says chip select one, two, three, four here. These ones have no label so we'll go check the Lcd schematic in a sec. but these come from Pa00 Pa1. so these come from yeah and 81c55 up here.

Okay, so that's being driven from there. so that's an Io coming from the processor so it's right into this and then four of those lines are going into driving the Lcd and then other stuff. you know. look the the read, write pin so it looks like it does.

It can potentially read back. I don't know where S1 is coming from or Si, but I wouldn't expect there like to be any issue with the reset because the Lcd is just working just fine and I don't expect it. You know there's no problem with any of the voltage drivers or any of that sort of stuff you know. So all the enable here so you know you wouldn't bother looking at that.

Uh, I don't know what this second what this chip select here does. but you know we're looking at one specific segment of the Lcd so there's nothing to see here. Move along Now to the Lcd schematic. They've got a nice looking Lcd and it's like you know it's very, it's laid out very well.

Look, you know, like M1 up here is, you know they've done it, overlaid it into the specific thing they've got. uh, information with driving the commons and the segments and the commons and stuff like this. Absolutely brilliant. So anyway so this is looking face on like we saw and as you saw we were having problems in this quadrant here.

the first four lines and the first eight characters. well as we said that's going to be controlled by M1 here. So what's coming into M1? let's have a look tilt your head with me and once again I wouldn't expect any of like the voltage things to be a problem because it's You know the segments aren't faded, they're not doing anything like that and they're all just you know common from with all the other chips. it's you know it's not a problem so it's you know data going into this sucker so it could be you know I'd be looking at the first thing I'd be probing if I got the scope out would be I'm probably going to have to.

Unless you just want to go. Well, it's probably that one. And you know, if you had like an Lcd swap or something like that to swap over, you might swap it over and you know that it's nothing on the driving side of things. It's on the Lcd module and then you might go well.

Okay, it's got to be this chip here and then you know, But even if you do that, you've still got to like probe it to see like it could be a dry joint. It could be a bad shoulder joint and one of these, uh, control pins or something like that. Once again, I don't think it's the data because the data actually gets in there. You saw it as we were able to write into these particular segments.

It's just that. Yeah, the interesting thing is, when we were driving this one here in this bottom corner of the Lcd like this. So if we were driving like line five here, line one would like duplicate this stuff on. uh, line five here.

So when we're writing data to this, it's also appearing on this and you might think, okay, you know there could be like some short between. Maybe there's some, I don't know, There's there's some tin whiskers or something happening, and there's a little short between. You know, just throwing that one out there. Uh, you keep that in the back of your mind, but this is where you've got to like.

Go back to here and look. Nine. I all these other lines, all these chip selects are being driven, so I don't know why the labels were missing. They're certainly going up here.

It's a bit hard to read, but you can see that's a Cs28 there, for example. So that's Cs2 One. Oh, they're actually out of order. Six, seven, one, two, Three.

Like. So that chip, for example, is there. And that makes sense because we need ten of them. Okay, because we've got ten, uh, pins here.

ten chip selects, and we've got ten chips total. So it's almost as if like the chip select for this one is short into the chip select for this one. Yeah, so we've got Cs21 here and that'll be this will be Cs26. Yes, So okay.

it matches the M1 M2. So yeah, we're looking at that line there and that one there. That's what's happening. Uh, so I you know, like it's not like those two lines.

I can't see how they're going to be shorted or whatever, but it could be something as simple as like an open line some they're wearing and then it's getting crosstalk. Uh, from the other lines. So yeah, I look, you know you can analyze this until their cows come home. You have a bunch of hypotheses and you just go in and check them all out.

So let's open this sucker up. and I'd just start with just probing around here for example and just seeing. Hopefully we can pro get access to the chips haven't taken apart yet, get access to the chips and probe them while we're actually using the thing. Otherwise, uh, you know you gotta like solder in mod wires to get test signals out and things like that, you know.

But first of all, open up, give it a bit of a visual inspection and then yeah, we'll get probing anyway. there's our suspect up there, M1. So what I'm going to do is, I'm going to have a get that under the microscope and have a good look at it. Do the visuals looking for any, you know, dodgy shoulder joints, any hairline cracks, any, uh, bridges, uh, caused by uh, tin whiskers? uh, for example, which you know, a growth over time.

they can grow out and short out pins and stuff like that. Um, yeah, just generally have a squeeze and you can just really see it quite uh, clearly here. how it just duplicates and go to the next line, look and shift that up. Check this out.

I wonder why they've exposed the solder mask along there. Just a little slither there there there. It's interesting they've done it on every chip. like this one at the top.

Here, it's gone on all four sides. Look at that. You'll also notice on the board here how they've got all these traces running off the edge of the board Like this. up here.

they've got up here like this. They're just buggering off over here. One of the column drivers over here just goes off the uh, even head down these signals down here. Why are they doing that? Well, that's probably for some sort of, uh, production panel testing would be.

uh, my guess. And then they, um, shear the things off that route them off, uh, later. All right, let's have a look M1 that all looks pretty smick. I've seen any tin, whiskers or shorts or anything.

Might want to go over this. see if there's any bad solder joints like as in physically not soldered. See if I can move any of these pins. they're all looking pretty good.

Nothing's loosey-goosey There really is no issue there whatsoever. I mean, it could certainly be the chip. Like, I would not rule out the chip. Hang on for a second.

I was just looking for a place to probe and I was looking at the pins I'm interested in down here and I see a little dag Hello Mr. Dag, I'm gonna. I'm actually gonna measure between these two pins and see what's what. Can't be that easy.

Surely those two pin are not 925. Ohms. My meter's beeping at 125. Ohms.

I'm going to get rid of that dag. now it's still 125. Ohms. I think that's a nothing burger, but geez, that was.

That was. coincidental. Okay, it's probing time. What I did is I just soldered a little uh, 0.1 inch header pin onto the ground pin there, which is uh, pin 5 and pin 6 is ground so that just allows me.

there was no other like convenient ground point unfortunately. and I'll include the schematic here. So what we're looking at, we want to probe the chip select pins. Um, that's like definitely the first port of call.

So we're talking pins 16 15, 14, Cs21, Pin 14. that's the culprit that we actually want right there. So Pin 12 10 8. So you know there's a whole bunch of uh, they'll be on this side of the connector over here, so it's working.

It's powered up, so Two, four, Six eight wasn't one of them. Pin Eight: I think there we go. We're getting bus activity five volts and yep, yeah, if I press the like, just pressing the enter key, uh, tens, 10's doing nothing. Okay, oh no, no sorry, I just wasn't probing it properly.

You got to get through. You might have some flux on the pins or whatever. Always good to have sharp pointed probes to get through if you ever. That's a tip like ever have.

like if your signal's missing or something like that and you think this should be a signal there, what's going on? Don't chase that red herring down a rabbit hole, just check that you're actually piercing the uh, like any uh, flux or any other contaminant on the uh joint that you've got a nice sharp pin. That's number twelve. Yeah, it's all these chip selects. They're all working now.

14. This is the bad boy. Hello, hello, hello, Number 14. Let me.

yeah, I'll push really hard on that. Yep, number 14. let's go to 16.. Yeah, 16's there.

Bingo. Winner winner chicken dinner. We found it exactly as I uh, suspected. Yep, Pin 14 there.

Cs21 which drives our M1 chip which drives the quadrant up here, which is the chip select for this chip up here. So obviously, um, it's now the chip select on this one. here. it's just flapping around in a breeze.

It's just floating, so it's picking up crosstalk from basically everywhere else. So that's why it's like latching in the same data. Um, you know it's because it's a common database among them. so they're all getting the same data.

So there's no wonder that this shift, this chip is shifting in the same data as you're typing in over here somewhere, for example. So it's yeah. the chip select is just. you know it's been enabled.

I think the Re yeah, the read write will be all the same. They'll all be the same, right? So it's right in the data. It's only the chip select that makes a difference between these chips and this sucker is floating. So that means our fault is actually on this border.

It could actually be on the ribbon cable. It could have like a broken ribbon cable or something like that. Um, yeah, it may not be an electrical fold at all. It may be a mechanical job or we could have a bad contact on the ribbon cable.

The first thing you do is just reseat the ribbon cable on both ends. Okay, those contacts look pretty shiny. No wuckers there. Let's plug it back in.

Nope. Same problem. It's not that it looks like the drive side or the other side of that connector. So unfortunately, this ribbon cable goes through the board here.

so I have to take out the entire board. Well, not. unfortunately. means we get to look at the other side.

There, you have it. there's the rest of the board. Yeah, it's still all, uh, through-hole jobs mostly. Um, although we do have some service mount, I don't think the previous version had any surface mount apart from the Lcd did it.

but I could be wrong. Anyway, Yeah, there's all our modem stuff out there. There's our isolation transformer for our modem and the 8085 made by Oki in Japan. All the best stuff made in Japan 81c55 is down here and there.

That chip contains Bill Gates himself. Copyright 1983 Microsoft. Um, so yeah, that's the the rom is soldered in. Hmm.

let's measure our backup battery there. Sorry, you can't see it. 3.42 volts. Oh, that's still pretty good.

And again, there doesn't appear to be any corrosion on the contacts at all. Nothing doing there. Plug that back in and I've reseeded that and uh, no. same old, same old.

So you know. Good thing about this is you can actually get this right out and work on the bench. I mean, look, we can. Actually, it's actually better to work on it now on the front of the board like this because we can just have our Lcd, it's just facing up.

Fantastic. Oh, that is probe in heaven. Beautiful. Beautiful.

Good thing is is that we can actually probe the top of this connector because the even numbered pins actually come out over here. So let's go Pin 14 again. 14 Tada there it is. Same problem, All right.

So it's not the cable at all. It's not the the ribbon cable or the connectors or anything to do with the Lcd. It is definitely coming from Drive on the circuitry. although I haven't inspected the solder joint.

now. hang on. Check this out. Pin Two Four 6, 8, 10, 12, 14.

that trace going through there? What is that gauge? Take. There is no way that is a coincidence. Is that some sort of like gouge chunk taken out of there? That's That's way too much of a coincidence that it's there. I'm going to get that under the microscope.

There's no way that's a coincidence. Is that being reworked? I don't know. but look, come on, that's not a coincidence. Is it that looks black? That looks like it's rotted away or something? Am I wrong? That looks dodgy ass? Let me buzz it.

Hang on. I'll see if I can get yet. Yeah, yeah. there we go.

That side, right? but it won't go over to here. Yep, there you go. We don't win a chicken dinner. Found it.

It's a track. Wow, wouldn't a bet on that. A track on the main board that somehow like corroded. Oh away.

I don't know. And these chips, you tell me, has that been I'm looking. It's like the other chips on here. They don't have the same flux residue on them.

I reckon this has probably had a repair done to it. Just look at some of the other chips out. You know these don't have any of that because this is all. um, wave, uh, soldered.

That's why it's got the glue. although that one still does have the glue under it look so I'm not sure what the deal is. Yeah, it's actually corroded away. Let me clean that up.

Not entirely sure how I I have to lift my microscope up here because, um, it's designed for like, because the board's too high so and add extreme zooms like this. But look at that. Yeah, yeah, there's some sort of corrosion that's got into that and that's eaten through that poor little track. That was it.

I found it by pretty good deduction. there. That was one of our hypothesis that, uh, the chip select was flapping around in the breeze. Maybe it was just getting crosstalk from something else and that turned out to be the case.

So yeah, all I gotta do is fix. That should be as good as new. Let's just scrape away some of the solder mask on the pad there. and this is post editing Dave here.

Yeah, my capture froze. I'm gonna like, uh, probably replace that with a Um. hardware capture solution soon. Like dedicated hardware.

Like a, uh, Black Magic? Some Atem probably. Anyway, little jump link. Yes, we have a winner winner winner chicken dinner here. It is.

No worries. Basic January 1st, 1900. Uh yeah. Y2k bug.

Um, yeah, it didn't end the world though. trust me, that was a thing. Those who remember the Y2k bug. oh jeez, I can.

Yeah, I was the Y2k engineer at uh, was it telly's or so seller. No, it was probably tallies Australia at the time. And yeah, I had to go around and certify that everything I'm talking every little piece ant project that contained a microcontroller or anything that had no real-time clock in it. I would still have to write a report certifying that it was Y2k compatible.

Anyway, Basic: Ta-da Trs-80 Model 100 Software Copyright 99 Because you want to save a few uh characters. Every character was precious in 1983, Microsoft Bill Gates himself might have written that. Um, 29 382 bytes Free Done. Fantastic.

Winner Winner chicken dinner. Hope you liked that repair video if you did, give it a big thumbs up And as always you can discuss uh down below and catch me on Odyssey and all the other platforms. You know what to do. Catch you next time you.

.