Hi welcome to Tear down! Tuesday We've got the Compu Corp three to two Gs scientists will sign programmable scientific calculator dating from 1973, so yes, that makes it more than 40 years old. Fantastic! thank you very much. James from Sydney who saved it from the Sydney University dumpster and you've seen this in a previous mailbag. He sent it in and it's just a beautiful looking device.

Something sexy about it. It's keys on other absolutely horrible, some of the worst spongy keys I've ever found in my life. but I don't know it, just it just looks really quite neat. I like it and it works on four D cell batteries as we saw it I Tried to pair it up in the mailbag video, but yeah, it drew like much more than its rated current of 1.3 amps when I tried to power 7 volts through the jack there.

But anyway, it is a beautiful machine. um, 1973 vintage. It's got a gas plasma display manufactured by Burroughs I believe. So anyway, it should be real interested inside.

Let's take a look. There's going to be lots of discrete stuff in here I suspect and I can see inside the battery compartment moldy board construction so let's get into it. and here we go. It already came with these screws undone.

There were just two more in the battery compartment there so let's crack it open and hey, we're in. like Flynn There we go. Oh look, you can see the you can see the nipple there on the gas plasma display. Doesn't that look, doesn't that look nice? There we go that instantly dates it.

There we go the 48th week 72 that display was manufactured. Got a couple of I know they're nuts I thought they were trimmers there for a second night. Um, a couple of nuts. So it's Hey Yep, the whole thing just lifts out.

Look at that. There we go. Hey, not as not as discreetly as I imagined I expected. Like a whole bunch of dense TTL stuff, but it looks like it just looks like we're going to have some LSI's in there that do all the magic.

So presumably I've got a take I might after. Yeah I've got to take out little spring washers. they're off the off those and the boards will just lift out. Now what's interesting at first glance is look, they've got like the plug-in card edge connectors over here and they're not actually connected up.

There's nothing in the case or anything like that that joins them together. What? There's no way you know, bus on the end of it or you know, no backplane on the end of it that joins these boards together over here. So that must be for production testing. I would presume that they cannot.

Then you know, like plug those boards into a test jig in the production factory and test them. That would be my guess anyway. but this is how that there. This is how they join in the boards together using this stack arrangement here and they plug in.

So this is like a backplane connector. you can see the cars just slide into there like that. oops sorry I've had my camera our white balance switch to auto so the colors may not have been perfect in the previous things that was from the trade show that I went to when I go to trade shows on site I just whack automatic white balance on. Anyway, it's fixed now.

So here we go. there's the 16 digit gas plasma display that looks really quite nice and of course this whole thing just lifts out like that. We've got a flat flex going over there, it's a multiplex of course and well let's try and get all these boards out of here and see what we can get. And here you go.

we're in. There's the two main custom LSI's and I'm not sure why I sort of expected this to be chock full of like a you know TTL or other target not necessarily TTL but like chock-full of logic I don't know. Anyway, that's the back of the keyboard there keypad so that's what. Teledyne Kinetics series K patent-pending thank you very much I think your patents expired by now, but yeah, these are manufactured by Texas Instruments in their you know clearly and there we clearly custom LSI chips there we go.

aided 29th and 34th week 1972 so all the magic happens in there and not much else. We've got some UMP for ATS they're obviously driving I don't know, they are off hand and a whole bunch of that transistors. They're obviously driving the gas plasma display up here and we've got some funky looking old-school foil caps in there and well, that's about it. Double all double-sided board, just all tin plate.

nothing fancy going on there at all. a few diodes, a few resistors, and nothing obvious blowin yet that I can see that's that causing issue by the way. held our boards are tied together with just these, so there's only a couple of like us. Apart from the big gap parallel connections on the back, they've obviously got I don't know some sort of power or something going over over there.

Why they didn't put it all on the card edge connectors at the back I don't know why they still need to do that. Not a bit disappointed. Not as dense and impressive as I thought. It takes all the shine off it when all the logics inside a couple of custom LS eyes you'll never get the data on and they've obviously got a compression fit on these like hard edge connectors over here because they wouldn't budge until I loosened all that and there we go.

They all come apart I'll show you that in a second, but obviously that has got done. You know they're using all the contacts in there to get power through. They got some logic on there, but apart from that, the bottom side here. Whoo-hoo And if you're wondering why they've shorted out all these contacts here because this slides into the can, there'd be pins on the bottom there shorting out.

Well, this is just one big heat sink and see the screws there? They've got some Loctite on those and they're just these little power transistors there. And they're just heat-seeking though. So this is the power supply board and you know, not a huge amount doing here. It looks like we've got an inverter for the gas plasma display.

hence you know this. That's separate cable we saw jumping off to the main one and he is our power import here. So maybe there's something wrong around here. That could be possibly why it's not working at all when you power it from the batteries.

or it's drawing too much current when I power it from the plug packs. Oh well, apparently. so there's got to be something wrong on there. And of course, this is a battery powered calculator even if it is decels.

So they've sprung for the 74 L series here. that low power. TTL stuff. Beautiful.

And here's our memory board. Take a look at this, we've got ourselves. Rama Answer: RAM Here and obviously some sort of memory controller. ASIC Once again, all designed by well manufactured by RTI Yep, 39th week 1972 and these are obviously our program roms here.

No, they're not. He promised there because there's no window on there. They've definitely got those soldered shut so they're definitely right. once roms and check it out, we have an Intel P 2102.

We've got four of those and that's going to be the RAM for this thing I couldn't immediately I know it's an Intel SRAM but I don't know what size. So this thing has a t program words of memory so obviously it's got to have some unless something built into the process in ASIC which will see that I'm obviously the process is going to have some registers and stuff like that, so I'm not sure if this is just a storage memory or whether or not it actually used for. you know, intermediate calculations and stuff like that as well. but yeah, ad program steps so it's got to at least all that most likely are battery back there or I don't see any sort of diode kind of thing happening for the power on that, but that would be I mean they've got to keep the contents of that when you turn the thing off.

Otherwise, um, well I believe it's got non vult. I'm actually I'm only assuming I don't actually know whether or not it keeps the contents when you are turn it off. kind of be stupid after reprogramming your 80 step program every time you powered it on. but I don't know could be and this is actually our main processor board.

So those two Texas Instruments Asics we saw on the display board. they they're probably like, you know, a display control. There could be additional processing as well. But yeah, here we go.

I Mean you know, Seven, Oh six, Oh five, Oh four. Let me check cuz the other one got a three I don't remember. Ah yeah. Oh - there we go up.

Oh - and Oh eight. So this one. there we go. Oh three.

Where's Oh one? So they've got at least four chips to do this and the process in architecture I Don't know. You know whether it's a 4-bit 8-bit or whatever, a 64-bit processor? I Have no idea whatsoever. If you do, please let us trap if you've got any. especially, you know internal block diagrams.

You don't architecture information on this sort of stuff. please let us know. So this top display board here they actually call that the D scan bought. All these boards are labeled, but yeah, I Guess that's a display scan, right? So obviously they are the two word display multiplexer.

er, you know, processors. They probably of course I would contain registers to handle the display data so the processor would just offload the data and then this would handle updating the display and keeping it multiplex. So that's you know, that's that entire board. I Think it's just dedicated to that display stuff.

so it really is quite a nice and modular design here. I Mean you know they made it easy to service to test As I said, you know they've got these test card connectors on the side, so they can probably test all this jazz. So we got our display and then we've got our main processor board. I Mean look this bugger all on there.

Look at that. I mean there's no even. Oh well. they've got to some I presume that's our bypassing on the one input there, but there's like they don't even bother things.

probably so slow, it's probably running it. you know, hundreds of Kilohertz or something like that. probably. You know, not worrying about bypassing.

That's why you won't find them on any of the other chips. So that's the processor and then the memory board over here. Look, you know there's no decoupling here. It's just it's just not fast enough.

it just doesn't need it. and then a power supply board. So very nicely modular and they can plug it in and test the damn thing. So it is really quite a nice modular bit of kit.

And then there is our press together. Check that out There we go. now. press together a card edge connector, so presumably like the contacts are just shorted from one side to the other.

I'm don't know I have to measure that. So let's test that theory that the top of here goes to the bottom of this one. And yep, it does. Check it out.

Yeah, that's just not coincidence there. Yep, there you go. So that's how it's designed. So the like the top of I'm sorry, the bottom of this top board here goes to the top of the board below it and then so on.

So you'd have to design your board architecture and your pin outs to up match that sort of stuff. but that's how they get the data between boards and it's all sandwiched together and once you screw that on, it actually becomes quite a rigid structure that holds the board in place. I Don't mind that at all. That's quite novel.

So a little bit surprised at the sort of lack of density in this thing. Like you know I Expected. like all dip technology of course back then, but I expected it to be. you know, like chock-full of dip chips on each one.

like I expected? sort of. You know this total number to be sort of crammed onto one board and I don't know. maybe they used to, you know, a four layer board or something really fancy like that, but now you don't have to when you go for all these. Asics but they're certainly designed and they had to split the processor up into four separate ASIC so I'd love to know what the architecture is there that made them do that and the display processor so they couldn't fit it in one so that to use to probably this is some sort of like gate array or something like that, so you know they only had a certain amount of logic and they had to design their process using the custom gate array.

so it's probably like a gate array chips that's not actually probably not a custom. ASIC As such, what I've done here is: I've powered it up I know it draws like 2.4 amps or something 2.3 I think just irregular. So I've taken out the board, taken out the power supply board here so it's not powering anything. I still left the gas plasma high voltage display plugged in and well.

instant drawing 1.1 amps at the 6 volt nominal 6 volt battery supply. Well, 6.8 Watts just for a power supply sitting there doing. now for what something's wrong. if I have a look at the output voltage from this inverter here.

Hey, there we go. 62 volts. It's another wire there - 56 volts. Ok, now we're starting Now we're talking.

So yep, Oh so you combine those and yeah, that can be recently dangerous so you don't want to go touchy-feely around there, that's for sure. so you know I don't know what it runs out, but hey, sounds, you know it sounds reasonable. Anyway, it's a high voltage inverter so that works. but yeah, where that damn 1.1 answers coming from I got no idea.

Okay, I've disconnected that. Let's see what we get now. and ah, Bingo! look at that. So that display we're only getting on.

you can't see that, but only getting a half a watt or 0.8 I'm dead. sorry 80 milliamps. So at 6 volts. oh there you go.

That display could be cactus that? yeah. I don't know, he's it supposed to take that sort of current I don't think so because the reading on the back of the case here it says that seven volts supply at one point three amps so you know with everything. I'm almost getting that with just the plasma power supply hooked up. So ah, alright now I've disconnected this side display processor board over here from the display and we'll just power that on again and see what we get.

Yet, we're still getting an amp, so nothing to do with the gas plasma display itself. Something on that display processor board is a chewing an amp and I'll get my Bremen BM to 57 here. Very nice little meter by the way. Nice little compact meter if you're after a nice cheap one and it's got a low impedance check function so it can use that to drain the cap.

I Measured this cap here. It's got during Operation the 428 volts across it so we can actually use that to discharge now. cap on there. it's not going to hold a lot.

Sony 0.5 Mike's so you know it's not much at all that. so there we go. it's in. There's no voltage left on, it's just going into auto mode like that.

So yep, that's fine. Discharged can safely work on it. So something has failed on here, what well as drawing excess current, so something is going to be shorted or loaded down severely. Diode Test: There's a diode here.

I Mean you know, here's our input over here, so there's no. There's no connections on the bottom if you have a look at the bottom there so it's easy to see all the traces going on top. I Love tracing these. they're just, you know, so easy to do and going over to the diode there.

So first thing we're going to check is that diode that looks ok, no worries there at all. We're getting our diode drop so that's fine. So yeah, we're getting a short somewhere else. The thing: I'm going to test next that main capacitor there.

Although, as I said, I measured the voltage on that I was getting 128 volts. So switch that back to Ohms and give that a bill. As I said, we've discharged that. otherwise it's going to affect our resistance reading.

No, there we go. 7 Meg And someone's calling me. better go answer it. So that's exactly what you'd expect from a good cap.

It's just going up and up and up. I Mean we can whack it into capacitance mo. But whether or not that's going to do any good in circuit, this is a half a mic so let's have a look, but it's not shorted. I Mean we're basically looking for shorts.

So yeah, it no, doesn't like that. Maybe it's got some residual voltage on there or something like that from the charging from Eva from the from the original charger didn't discharge enough. Or maybe from the resistance test function that's obviously having a big issue with Auto ranging there so we could know this one doesn't. We can't Manual range.

Oh oh, look at that auto range only forgot about that. That's a bit annoying. I Hear I was saying this is a decent meter. Well, it still is, but yet doesn't let your manual range the capacitance.

That's crazy. Anyway, next thing I'm going to check is the other caps. I mean before we start mucking around with us semiconductors, you know these are I don't even know the pin outs? What is it you? HP 480 I Don't know and you don't know the pin outs for the custom? ASIC So I don't really know where the rails are here as such. So let's start measuring some cups and see if any error.

That's a pretty consistent 7 nanofarads there. it's probably nothing else in circuit there. Hey there we go. There's seven nano farad's a pop and it doesn't like that one.

that one's got something in circuit. We can go to our resistance range there, see if that's shorted out r2 k You know it could be something in circuits. I'm not too concerned about that. I'm looking for gross shorts at the moment, so let's measure the other caps here.

$7.00 Farad's again. Bingo. Alright, we should try and change the polarity on that one. That one is again, looks like it's low impedance.

7 nano. Farad is pretty close, so let's go back. measure this one again. Hello, hello, there's your problem.

I'm dead. It's short across that cap. so ouch ouch. There's that could be an issue we could have.

our souls are dead. Cap: They're seven Nano Farad's yes, Seven doesn't like that one either. So let's go back. No, that's one point to cake.

but that one. Oh, look at this. I Originally thought here's the cap. Here are the three caps there there and there and it's the center one that's shorted out and I thought that track was at first glance going straight across to there.

but it's not. It's cutting short. but look, that's pretty done. Close down in there, isn't it? but it's not.

But geez, yeah, there's no little bridge in there or anything. but maybe the top side I Don't know. We'll have to suck the cap out and see I know, Wait, look at the top side here. What's that little dag in there like that that could potentially be shorting out? Maybe I'm going to get the knife in there and give that a little scrape.

It doesn't look like it's touching, but geez. Alright, so I've lifted that out and there appears to be I'll show you this close up in a minute like a white residue or something on there. So yet that cap is definitely shorted, that ain't right. So yep, that well at don't know if it's going to solve the problem, but it definitely is a failure point.

So check out that residue on that capacitor and it's like all the ones next to it have that residue on there. So yep, that cap is gone. Laughs! And you can see there's evidence of some of that on the bottom of this cap over here as well. So yeah, you just changed that one as a matter of course - even though it hasn't shorted like the other one, right.

I didn't have any axial Pasteur's but yeah, I had some radial ones I'll just bend the leads I probably got some old axial ones somewhere, but probably not the right value anyway. 5600 Puff, it's going to be near enough I doubt that it makes a major difference I changed the two that were leaking I guess with that scrap white crapper losing out. so let's pair it up. Hey, that's better.

There you go. Half an amp? Awesome. So that was definitely some of the current draw right there. So with all boards plugged in now I'm getting you know, almost two amps so that is higher than the rated current on the back there of 1.3 Hmm.

And if a pair up just the power supply board on its own with the display supply disconnected, we're getting 9mm. So that's pretty much you know what you'd expect ballpark for the quiescent supply of something like this old-school power supply. So you know, no problems at all. We're basically looking for gross excess current faults.

Here's let's check the power rail here and across this main main fuel to cap up here. And Bingo. Four Point Nine volts. Presumably it's five volt logic.

and well, I'm happy with Four Point Nine. That's certainly well within spec. So um, yeah, like a 5-volt logic. They've got seven for TTL stuff in there, so you know they're going to have a 5 volt supply everywhere you got to presume.

So that's ok. So all the logic is powered up, and if we have a quick little probe around here, there's our 5 volt rail. So that's doing hunky-dory No problems whatsoever. There's no ripple on that, and we've got all these bus stuff happening over here.

so let's start probing some bus. Hello hello Mr.. signal at 10.1 Kilohertz. Okay, there we go.

Yeah, Ten Point One. Yep, oh hello. it starts working. Hello bit around in there.

Look at that 200. it's almost bang on 200 kilo. It's 203 kilohertz there, so we're getting some activity. So oh yeah, it's reasonably warm.

Yeah, they're actually getting warm to the touch. Oh, that's gone. Negative. Yeah, there we go.

Got a bit of negative stuff happening down in there? Yep, but yeah, we certainly have some. Certainly have some. clock stuff happening. There's that 10 kilohertz and 200.

So yeah, it's doing something. It's powered up and she's trying to work actually. I Just thought of something. The back of this says 7 volts at naught point 7 amps but that might it may not be able to operate from that.

That may just be to charge if you've got internal rechargeable batteries in there that may just be the charging current circuitry. So yeah, maybe this you know I mean clearly we found a problem. It was drawing excess current. We fixed that with the with the shorted cap and stuff like that.

But you know the 800 milliamps that we're drawing with all these boards in what 8 850 could actually be right? and there might actually be nothing wrong with this. It might be driving it. Maybe just the display board, all the display itself which is a cactus. And by the way, if you want to know what good a good continuity quick continuity buzzer here, this one's pretty darn quick.

You can you know if you're scanning for ground pins to find on this top board for example, yes it's powered down. we can swipe along there like that. Bingo until we got one right on the end. We got ourselves ground pin right there and if we so I just sold myself on a little ground test point there.

I've got a resistor just hanging off there. It's a bit boggy bit how you doing but it just allows us to get in here and and probe around and look. we've got some activity there so that's a win. That's line 925 Hertz There that's 5 volts T volts per division.

That's going negative by the way. so that's negative going. So there's lots of negative going stuff happening in here and that's 10 volts, 5 volts per division, 5 10 15. So like negative 15 volts and stuff like that.

So yeah, like this data get into here and I checked a few supplies on the other supply rail and year we're getting like a negative 15 supply and there was a plus 9 somewhere. And you know other stuff, but you know we're getting nothing on the display. and I presume that I don't have to, you know to see some sequence to power the damn thing up. So I don't know? Yeah, I would have expected the display just to come on, but certainly getting some activity there.

Some of those pins on that chip their negative hundred and 50 volts. That's 50 volts per division. So yeah, serious business. Well unfortunately I think I'm going to have to call it quits there for today and we we got some progress on the repair of this thing, but who knows, You know, maybe it's just not repairable.

I Don't know if anyone certainly has a schematics for it. That would be very nice. So anyway, I hope you enjoyed that teardown and a little bit of troubleshooting there of a 1973 vintage Scientific Desktop scientific calculator with a whopping 80 words of program memory and it's a bit of a beast and quite simple. And as I said, if you do have any info on the gate arrays or ASIC site used in this thing, please leave it in the comments as well.

So hope you enjoyed that And as always you can discuss it over on the EEV blog Forum Links are down below and also there's a link down below to the high-res tear down photos of this I mostly take up photos as I go along and they'll be yeah on Eevee Blog.com Catch you next time you, you.