 All right. This is an experiment. I've never done this before. I have recently received my shiny new ZX Spectrum. This was made in 1982 and it suffered slightly from the decades since it was made. Before I can use it to play games like Chuck Egon, which is really what I wanted it for, and he's a bit of repair. The main thing that needs happening is these ancient capacitors have all dried up and won't work properly anymore. Due to the unique way in which the ZX Spectrum's power supply circuitry was designed, if one of the capacitors in the power supply area fails, which it's likely to do due to being so old, this can cause irretrievable damage to the memory due to the voltages not being generated properly. So I'm going to have to replace them. I've already, as you can see, taken the lid off this thing and I've gone through and I've found replacement capacitors which should fit. So I am going to try replacing them live on video, recorded on video, so that when I inevitably get everything wrong, the internet can watch and laugh. You can see here this is the RF modulator that takes the composite video signal produced by the board itself and produces a radio signal that televisions can pick up. I've already taken the lid off and modified it. The reason for this is that modern televisions don't actually cope well with the ancient RF signals that these computers made. So a minor tweak will change the socket to produce composite video instead. This works much better on modern monitors. The modification wasn't a particular success. The video signal is very bad, so there's also a change I want to make which involves replacing this wire with a capacitor which should hopefully improve it. I'm also hoping that the reason for the poor video signal is dodgy capacitors everywhere. I've already taken the heat sink off. This goes this way round and would have fitted under here. I won't bother putting that back on now. This ancient linear regulator converts the 9 volts produced by the power supply which I also have and is completely short. I'm just ignoring it completely. It converts it to the 5 volts used by the main board and the power circuitry here. The reason for the heat sink is that linear regulators get appallingly hot. I actually have a replacement modern switch mode regulator on order to replace that. I won't be firing it up today because that hasn't arrived yet. I'll record that as well so that you can see whether my modification has actually worked. I've already taken this apart. The keyboard's duff as well. There's another membrane on as well. The ZX Spectrum membranes are notoriously brittle when old and tend to fail. Some of the keys and mine don't work, including the all-important J key which you have to use to load games. I am going to start with C46 here. C46 is special. The reason it's special is that you probably won't be able to make it out on the video but the negative pole of the capacitor is actually connected here where it says plus on the motherboard. I'll just double check focus. This is due to the PCB manufacturers, PC designers making a mistake with the printing of the board. So you've got to be aware that this capacitor is actually connected backwards. If you connect it the right way around it won't work and will probably cause major damage to the board. So let's remove that first. First thing to do is to snip off the SNP capacitor off the board. That was the irrecoverable moment where my spectrum will now no longer work. Let's double check the size. This is a one microfarad capacitor. I have previously selected some capacitors which one of these is one microfarad 22. I did actually put them in order. I seem to have knocked them all over the place. There you go, one microfarad. Now the original capacitors are axial. So the wires come out each end of the capacitor. I don't have any axial capacitors so instead I'm using these radial ones where the wires come out of the same end. So we need to bend the wire back. That was actually the wrong wire. I want the long one and I can solder it in this way. The first thing I need to do is to take the old wires off the board. I now remember that I forgot to solder a great piece of kit. Now I've been warned that ZX Spectrum motherboards are not particularly great with heat. So I have my soldering iron set down low to 250 degrees centigrade which again is supposed to be the recommended temperature for doing it. So let's just see if that will melt the wire and melt the solder. Yeah, no trouble. I'm not the world's best solderer. I don't think anyone is. The great thing about these spectrums is the PCB is so huge that even my dodgy soldering skills, come on, work on it. No idea of that bit of wire went ricocheted off probably into my drink. I'm recording this using a props up cell phone. I hope the angle of view is about right and that you don't get a look at my head. Nothing ruins a good tech video and seeing somebody's hair in the way. It's an ancient Nexus 5, the only. I really hope the rest of them aren't going to be this much of a hassle. Nothing ruins a tech video more than having somebody else's hair in the way. I may have slightly buggered that. I should really have arranged for some pins and things ahead of time. Very useful for poking. Of wire through the board. Is the solder melt? Easier not to cut the plastic completely. Try that next. There we go. OK, the hole is clear. Other side, I think I made lots of these in various different kinds and they're all a bit dubious, every single one of them. My board has actually been modified twice in the factory to work around various bugs in the design. The C46 polarity one is trivial in comparison. As you can see here, this transistor has been added. There's a stray piece of wire here. It's known as the spider mod. It's to work around a bug in this custom ULA. The other modification is there's a capacitor, a monkey patch on top of the board there. I actually forget exactly what that one is. It's in the DC control circuitry. It'll be something to do with power generation. These old computers use lots of different voltages bent on this side of the board, which is why it's not coming out clearly. Old computers are much simpler by comparison. You just give them roughly five volts and they just work. Wrong wire. Don't do that. That's C27. I am going to want to remove that at some point, but not right now. I've been trying to remove this diode by mistake, so let's just push that back onto the board. It gives us something to suck when we try to sort of suck it. That was a faff. Now, what do I do with that capacitor? Microfarad cunningly bent. Here we go. Now, this is the backwards one. Positive terminal here should have the negative side. That one's negative. That wire goes in here. It's not exactly trivial with these radial capacitors. A lot of wire coming out from the one end of the radial capacitor. I need to be careful of that. Most of the capacitor lead has been used up bending over the back. So, soldered on. Is that a clean joint? That's a clean joint. There goes another one into my drink. Not the greatest. All focus, but it does seem to work. Oh, good. The video is still working. That was the dodgy one. Pick one at random, this one. Cut one side this time and see if that helps. Important safety tip. Never eat or drink while soldering. Let me just double check the polarity. Negative is pointing to this side, which is correctly labelled on the board. In order to get this wire off, if I can do a better job of it than I did last time, it needs to pull through the board, I think, but apply a little bit of tension to the back. It would just pop straight out when I apply the soldering iron. I'll do this. I'm still not clear. Heat it up when I push the capacitor through. This one should be easier because I've got a lala cap to hold it on with. This is another one microfaric capacitor. Lift it over and down. It's in that side. It's checking the size. Weird spider construction. This one is C27. I have no idea what any of these actually do. Did I mention that my soldering skills aren't great? Can you tell? Plus, this is all standing on the map, so the chances of it showing up on the camera are going to be nil. That's not working at all. OK, let's see if we can clear this joint. Some solder. Let this be a lesson to you. Always make sure the hole is clear before you put something in it. Nobody likes a blocked hole. Of course, the ones I'm not very good at. That is really incompetently bad. Joints are the ones I'm going to do first. Negative to that side. Positive to that side. A quick look to make sure there aren't any shorts. Good. Let's just double check with the multimeter. Two done. One, two, three, four, five, six, seven, eight, nine. I want to put one in there ten to go. What's next? Let's try C50, this one. This one is a 22 microfarad capacitor. If I can learn the lessons from the last time I soldered, put that straight, get some tension. There we go. No trouble. The hole is blocked. I'll double check the polarity. Positive is pointing at the correct pin on the board. I need to unblock these to get more iron in. Good. That one was a 22 microfarad. I hope it was a 22 microfarad. Yes, it was. I've got some 22s here. 22. The legs are about as long as each other, so let's use the negative one over so that we can have the legend on the top. That's a bit of a problem. It might not be big enough. It will go in just not quite long enough. I can make it pop through. It's not brilliant. These radial capacitors are not really the right choice for this, but they are what I have. I'm a little bit forced to hold the capacitor in place. Don't do that. That works because there's a little bit of solder left on the tip of the soldering iron. You're not really supposed to do that. You're supposed to always use fresh. The soldering iron should not carry solder. It's not even in straight, but whether it works electrically, that's the important bit. That's a decent joint. Pretty terrible joint. But the electrons like it, so I'll just keep it. C47. I have actually had this spectrum running briefly until people shouted at me and made me stop because of the problem with running with elderly capacitors. So I know it works. I couldn't actually play anything because the aforesaid keyboard problem, but I could type stuff on it and things happened and that sort of thing. Whether it will continue to work after my modifications is a completely different matter. We can always hope you can get spare parts for these. There are people making new rubber map membranes for the keyboard, cases, some really nice looking aluminium faceplates for the keyboard. You can do homebrew, build your own replacement motherboards using modern components, but still the same old spectrum inside. So if I do manage to fry the motherboard, I could always just replace it, build myself a new one and use the old case. I suppose if I managed to break enough of it, I'd end up with a completely new spectrum restored from scratch. But still, of course, the original one on the keyboard is just a little bit repaired, like every single component. I have the right joints. I do. When in doubt, add more solder. Big fat tracks suck away heat. So it's very hard to get the joint hot enough. I want to replace a capacitor on a modern PCI graphics card. The PCB had turned out multiple layers of copper inside the PCB to make things like grounding planes. And these formed really good heat sinks. And I ended up having to use three different soldering irons on the same joint to get enough heat into it to actually get the thing off. Right. The negative was the bottom, which is correct to causing PCB. And I did learn my lesson from last time. So let's get the holes clear. It's a very important thing which I forgot to do, which I'm going to have to do in a second. I hope there's a hole in there. There must be a hole. Oh, here's the problem. See these silver pads on the top? They are sucking up heat as a heat sink. That's why it's being so problematic. Also my soldering iron is set quite low. We'll see if that will do. OK, the problem I mentioned before is I have removed some capacitors off the board. I've got four here. I think this is the one I just removed. This is a 22 microfarad. If I replace this with a 22 microfarad capacitor, have I in fact replaced it the right one? That's another 22. We should have two ones here. One and one. Let's get rid of those. So I've got two 22s, which means this must be a 22. The last thing I want is to put the wrong one on the board. These are 22 microfarad 16 volt capacitors. This is a 22 microfarad 25 volt capacitor. That will work just fine, even though it's a fraction of the size. So that one will go. It's a well-known fact that all electronics are actually powered by magic smoke. If the magic smoke gets out of the component, then of course the component stops working. Did you know the reason why you can build stuff with the soldering iron is because solder contains extra supplies of magic smoke. You can see little bits of waste smoke coming off whenever I solder a component onto the board. Yeah, that'll do. It's not pretty. It should be working. C28 is this one up here. Dispose of old capacitor. I'm going to try something different from taking this off the board. I need a small screwdriver. Like so. Apply a little bit of tension to the board. So, apply heat and out it comes. Yeah, that's so much easier. Having something to apply leverage to rather than just a straight bit of wire makes life easier. Clarity check. Oh, that's interesting. No, right, sorry. I was misreading the labelling. I thought the negative pole was connected to the positive thing on the board, but I don't know if you can tell. The axial capacitor's negative is an arrow rather than a stripe, which it is on the radial ones. Good, this is another 22. 22. If I'd be better off doing it, I probably would, but I'm not particularly happy with having long bits of metal and these pads on the board. So, the bottom's a negative. We'll go through this hole. It really wants to go in upright. Or upside down. Everything's fine. I'm sure it's still working. If you have any real inspection or stories of seeing this, they are at the moment jumping up and down and screaming in horror. It should work. Positive that side, negative that side. I'm sure this is going to end up being a thoroughly thrilling video. Looking at where things are showing. Most of the stuff I've been doing is over here, so it's not really going to show up much. I need to be able to look down at the board. Hang on. If I ever do this again, I will try and adjust the position of things. I don't really think I want to do it now. So, which is next? Let's try this one. C25. This is almost certainly something to do with the video circuitry. It's another 22 microfarad. That's all 22s. I think there's something 722s on the board. So let's try this trick. A little bit of leverage. Place in heat. Moving. Parity check. Negative is pointing at the negative terminal. Which is what we like to see. 22. 22 is a 22. Good. That's interesting. Possibly very convenient. So if you look carefully. So there's actually a track running along the board there. Which means that this hole for the positive terminal is connected to this hole for the positive terminal. If you look right next, the negative hole. That means, I just need to double check that with the multimeter. Where did I put the multimeter? That means that we may actually be able to put. Yes, you did. Good. This means that we can put this in vertically. The way they were supposed to go. It also means that the positive hole here. Which we're going to use is in fact already clear. So I just need to clear that one, the negative hole. A solder. That visible. Not really. Down again. Some solder. Actually, do I have some solder? It is the sort of thing I would have bought. I don't see any in my drawers. Positive goes to this hole. It's not quite a match, but it will do just fine. It's a bit warm. Big fat copper tracks. Actually, I think they're big fat tin tracks. Yes, interesting fact about these boards. This ripple effect. Again, you see ripple effect over here. That's not a fault in the board. Because after the board was made, they reflowed the whole thing with a fairly thick layer of tin. For reasons. And it just leaves a layer about half a mill thick or so. It's a bit uneven hence the rippling. Actually getting there, you know. This one is going to be a pain. These ones are going to be interesting because of this one that's spider-patched. It's actually floating. It's connecting the positive end of this capacitor here to this resistor over here. And that's going to be moderately exciting to do. These are actually the opposite. Next to each other, but the opposite way round. So this one, the negative term is here. And this one is here. Both off and do them at the same time. These are 100 microvaric capacitors just for a change. It's looking at the layout of the tracks. I think we can use the same trick we did last time actually. At least for that one. We'll do the other one. Again, I have no idea what these are for. But they are probably related. Otherwise they wouldn't be so obviously related. This is the second one. And this is another 100 microvaric capacitor. Fantastic. So this one, we're very happily dropping through these two vacant holes here. Like so. Perfect. The one above is a little bit trickier. The closest two holes we have are there and there. But there is no PCB in between. So this will actually be a perfect choice. We can just drop this in. 100. If that side, negative that side. There's no stress. We like to do it. These holes are actually very useful. They're actually through holes. They are drilled through the board and then plated on the inside to make a conductive channel. They allow you to connect. They allow the board designers to connect a track on one side of the board and a track on the other. Actually looking at some of the other very dubious capacitors I've put in, we could have used the same trick. This one here. There are actually some holes in the right place to make that work. No matter. It'll probably work. Still recording. Still running out of disk. We've got three more to go. That's actually going quite well. This one. I don't want to do that one. Balanced weirdly on two legs. You can tell the designers didn't want to do it there either. There's absolutely no space to work. It's got to go. I think this one I will snip off. I'm not actually using the modulator here at all. So it could go. But it's fastened on with the monster stakes. And it also provides a rather useful phono plugs in just the right place. If I could take that off it would actually improve things quite a lot. Does it actually just sit in this little enclosure? Is it fastened down? Yeah. It doesn't want to go. I'm not moving it. This is certainly your video circuitry. It's all very dense and full of transistor something. Is it this one? Is it this one? And this is a 22 microfarad. Negative is pointing at. It wasn't labelled. That means I do not know which way around it went. Fantastic. Okay. So I did actually take a picture of the board before I started doing anything. So I'm just going to go and take a look at it now. And back. Yes. The positive end was the right-hand side. Always photograph your board before you touch it. Oh God, yes. Right. So here we have another 22 microfarad capacitor. And it's just going to have to spider in over the top and have to clear those holes. Yeah. It's just going to balance in standing on its legs. It will keep it reasonably out of the way of the other components. It shouldn't actually be too bad. But yeah, I've got to clear those two holes. I solder, remove solder. Where I work, I have a team of... I work with a team of engineers. And a little while ago I bought a IBM PC convertible. So I'm trying to clean the tip for my soldering out. There we go. And this was made in 1986. And it turned out that at least one of my team members is younger than that. And this was 1982, this ZX Spectrum. And I have a feeling that it might be older than anybody. Is it a 22? Is it a 22? Is it a 22? It is a 22. And the... It's positive on the right. I said it was cleared. I thought it was going to be clearer than that. Very much not clear. And again, I am seen to be moving off the camera. Can I actually adjust this without anything terrible happening? Let's try that. Yeah, okay. Talking to somebody in the team the other day, it turns out that this computer was a PC. How wrong is that? What should end up being thinking PC's or ordinary computers? That was me pulling my poking wires all the way through the board in the hope that it would actually clear off things a bit. Okay, let's see if this goes in now. Positive on the right-hand side. On the left. Ah, it goes in. Yeah, that was the loudspeaker. I don't actually know what happens if you connect some electrolytic capacitors right backwards. Nothing good, I'm sure. I don't know whether it's like permanent damage or it just doesn't work. All right, well. So, now I have to take this ghastly thing off. This should be easy at least. I can avoid toasting one of the other components. Now this is very, very careful to remember this. Negative right. And it's a 4.7 microfarad. You haven't had one of them before. I cheated. I looked all these up ahead of time. 4.7 microfarad. But we can't replace it until I take this other, the original capacitor off. That's another 22. So I'm just looking at the board in the various holes. Every mouse knows that two holes are better than one. And that applies to PCBs as well. A bit of tension and easy polarity check. That's the right way round. And it's labelled on the board. And that's a 22. So you can remember this. The 22 goes in here. And the 4.7 goes across there. So that's the 4.7. 22. Why don't I have some spare 100? That's quite a big 22. So that wants to go in here. But we've got all these tracks in the way. The positive goes at the top. So it looks like we may have to do our bending over thing again. There's really no space. The wires are too small. So I could easily spider it above the board slightly. There's probably room. And the wires are stiff enough to keep it clear of all these other components below. So, I mean, it's not brilliant. But that's what we're going to have to do. But the age old problem, there's stuff in my holes. It's going to be pretty heavy handed with the heat. Because you have to heat the joint all the way through. That didn't work. Otherwise the solder won't suck. There's no point just heating the surface solder and pulling that off. Positive top 22. Negative bottom. So that goes down like so. Hold the wire in. I just dropped a blob of solder onto the PCB. That's going to come off. That was easy. And this one. Now what did I say it was? Negative right. I hope it was negative right. I need to trim the leads. This is a bit terrifying because you can't trim things longer. And so that looks like it'll fit. Get all my hair. Should be done. It's going to work, of course. So the one bit left to do is I want to fix the composite mod here. And replace that with a capacitor. To do that, I need to go look up how big the capacitor should be. So be right back. All right. It is in fact a 100 microfarad capacitor. Ffripping all over my work surface. And the positive lead of the capacitor needs to go to the composite pin. And the negative one needs to go over here to where the video signal comes in. There's another modification you can do using a transistor which could improve the video signal some more. But I'm not going to do that because I don't have any transistors. So we need a 100 microfarad capacitor. 100 microfarad capacitor. Let's undo this. So this is my old modification. Lasting a little bit of wire. And let's actually remove this wire from the board completely. This was the original lead that used to connect the... It used to go through one of the little plastic holes in the sides of the modulator. So that comes off. That's left us with a non-clear hole. Of course it has. That's a thin track. The capacitor needs to go in here. The short wire needs to go through the hole in the board for the negative terminal. The positive wire needs to come through one of these plastic holes and meet the centre pin of the photoblog. There's plenty of space. It's just a little bit awkward. I'd rather like to remove the modulator completely. Go perfect. So let's solder this one on here. On here. OK. That was moderately anticlimactic. I was expecting that to be tougher. This bit was a mess. I knew that was going to be a mess. I wonder whether I can improve some of these. This one could come off the board and be replaced with a vertically up and down one. The connectors are all in place for it. Yes, let's do that. So this actually wants to come off the board. Unlike the other capacitors, we're going to want to reuse this one. So you've got these two big silver pads here and I believe they're connected up underneath. Yes, they are. So we need to connect that hole there and that hole there. So you've got one hole to clear. Very clear, but I'd love another hand. This would make this so much easier. Imagine what kind of amazing electronics humans could have produced if they'd had another arm. It's going in with one of the legs that the capacitor has bent. I really want to put this on YouTube for myself. I suppose I did promise at the beginning that I would let the internet watch and laugh. A decent joint for this leg. A decent joint for this leg. Nice and jointy, that's a dry joint. That is a great little multimeter. Well made. I don't know if it's accurate or not. Easy to use. It costs nothing. Mail order from China. Big clear display, auto scaling, everything. I normally prefer analog meters, but that is a very nice digital meter. Now this was the first one I did. That was the second one I did. C46 here was the first one I did. Now this one's a mess too, but I don't actually see any useful looking holes. So I'm just going to leave that. OK, well, I think that's done. I want to see if it works now. I can put the heatsink back on and try it with this ancient regulator. I don't have a working keyboard. I am actually going to wait until the rest of the stuff appears, but I won't post this until then. I'll just edit it all together. In a moment you will magically see in my hands here a new regulator. And as if by magic, a regulator appears. So it's now two days later, it's Wednesday and I filmed the first bit on Sunday. The regulator has arrived. Beautiful little piece of kit where this is a brute force resistive linear regulator that drops four volts by putting it through a massive resistor generating massive heat, needing a massive heatsink. This is a modern switch mode device that cheats. It's so efficient that they haven't even put anywhere to put a heatsink. So once that's on the board, the whole thing will run generally cooler, generally less stress, more reliable, and it produces better power as an indication of just how much heat this thing produced. With the heatsink installed, these capacitors next to the heatsink actually had to be rated at 150 degrees centigrade. Otherwise, you know, bad things would happen. So since the first part of the video I've actually looked at some of the footage and I've noticed a couple of things. One is that it's all horribly out of focus because I don't think there's enough light and my little cell phone which I'm using to video this with isn't capable of maintaining the focus. Let me just double check that now. It looks not too bad. A little bit better. I think I just need more light. I'm going to see if I can source another couple of lights. The other thing is that I'm practically inaudible for a lot of it. So I've moved the microphone. I seem to spend a lot of time mumbling into the board down like this. Yes, I can see some dreaded hair. So maybe this will be better. Let's see how it goes. Another thing I've noticed is I seem to have made a bit of a cockup which I'll come to later. But I'm going to have to redo some of this. We'll have to see. I know that I'm going to have to replace at least some of the capacitors but I don't know yet how many. But I'll deal with that when the time comes. So, we want to install this where this goes. 36-year-old linear regulators are not exactly sought after. I've got a drawer over here full of new ones that haven't been overloaded for four decades. So I don't really think I want to keep this. I think I'm just going to snip it off. Having three pins makes it rather awkward to remove. Noting that pin 1 is on the left that means the new regulator is going to go this way round. It looks bin. Okay. Three pins to remove. I took the entire pin out. That was good. I left the hole clear. I don't know how long this video is going to be. I'm going to spend the entire time in it talking about taking stuff out of holes. It wasn't expected. That's left a neat little solder plug. Is that visible on me? Can you see that? Let's try that here. Solder was sucked out of the hole, expanded rapidly, cooled down and produced a rather neat little cylinder. Anyway, that's not what we want. So, remove. If any position of the light actually means I am working in the shade. I have to adjust slightly. It looks okay for me. It's awfully glaring for you. Let's try over here. Working at all on that one. I should invest in some solder wick. It's a kind of braided copper fabric that you push onto the joints and you apply some heat. The fabric sucks up the solder. It's not quite as clean as it tends to smear solder about the place. I don't think this is working terribly well. Yeah, the nozzle is full of goop. Soldery goop. I'm going to have to do it in a mechanical way involving pulling, preferably without putting my finger through the loudspeaker. Slightly more hairy than I really wanted. Anyway, that's all that now. I've been left with reasonably clear holes. New regulator. Search new regulator. Beautifully. Big enough to fit. That has not actually gone flush at all with the board. I should have checked that before I soldered all three joints. It's never coming off. At least not without wrecking it. I really don't want to do that. That is going to have to do. Okay, that wasn't brilliant. The issue I mentioned earlier with the capacitors is quite simple. Let me produce a visual aid. This is the top of the case. It goes on the board. In fact, here is the bottom of the case. This goes under the board, like so. And the top of the case goes on the board. It's properly. It neatly drops down like this. And the top fits on like this. It doesn't fit on because these capacitors stick up too far. Because I mounted them vertically, because that was easier without remembering to check the board clearance first. I thought I was so clever. There's actually nine millimetres of clearance with these little standoffs, and those capacitors are ten millimetres high. This one is nine millimetres high. So it just about fits, but I'm going to take that one off as well because I want a little bit more clearance than that. I actually want a gap between the top of the capacitor and the board. The regulator. Let me grab this. That's eleven. But that is further back. This slopes up, so there's going to be slightly more space. I think I do need to get that flush. Well, I have to replace the capacitors anyway, so we'll see. Excuse me. Now, because I trimmed the wires when fastening them in, because you have to, the wires are now too short to mount in any other position, so these are now junk. I'm going to have to get rid of them completely. I'm going to do this one as well. So let's start with twenty-two. Luckily I have some spare. So this has to come off the board. First, double check the polarity and where the pins are. Positive goes up there. Yes, I'll remember this. More holes to clear. There's my shiny new capacitor. Never even got a chance to be used. Replacement. Focus check. Looking good. New position for light seems to help. No one's ever going to watch this, so who cares? This is all practice. So this wants to go through here. In fact, those are actually sort of connected together, I remember. This is a slightly different form factor, which is why it's smaller. This will actually, this will stand upright, you know. It's smaller enough. Double check that. Yeah, but let's put it flat anyway, so I need to fill some holes. Sorry, clear some holes. Spray to solder all over the board. It's a hot. I'm going to have to learn how to do this better. Again, I am drifting slowly out of camera shot. The stick of poking is officially not working. You know, it'll be awesome. This thing actually worked after I finished fafhing about with it. All that's doing is getting my fingers hot, so I'm actually going to. That was a shortcut. I'm going to actually heat the wire up and poke it through the hole at the same time. I'm going to have to actually do this properly. Heat lots and lots of heat. I want the solder liquid. That was too delayed and the solder congealed while I was waiting. I just managed to bridge two tracks with the solder. Got to be a better way to do this. Actually getting the solder out of the old holes has been by far the most difficult part of this entire proceedings. Actually, the rest of the soldering has been easy. I mean, not far from the obvious mistakes, but I need to check that continuity. I think there's still a drop of solder between the two tracks. I remember this joint from last time. It's got a whopping great copper track on both sides. That's completely off camera, isn't it? Which is why it's not actually, why it's so hard to heat up. And these plate through holes have got, of course they've got copper through the hole. So it's transmitting heat from one side of the board to the other. Maybe I would have better luck with. So I've got one spare hole there, so I don't actually need to clear the one in the big track. But I do need to clear this one next to it. I wonder if I can use the push through trick for that. So what I do is I apply a bit of force to one side of the hole against the solder plug in the hole. Melt it from the other side, and in it goes. So I want this to go through quite a long way. I think that'll do. That's standing well proud of the board. So all we do is just bend it over. But I'd like to go in a bit further. That just comes down. So now I can solder the joints in this side, and it's finished. So that's actually come through back through the board a bit, so I need to push that back in. That side's not so hot either. The rule of thumb I was taught by the electronics is they can cope with as much heat as you can. So if you hold the component in your hand while you're doing the soldering, and you stop before you actually scream in pain, you should be okay. You won't have overheated the component. A grotty joint, but it's actually connected. And I'm not going to snip that just now. I'm going to do the other ones and make sure that the lid goes on. Then I'm going to cut them all off. But it's really important to always cut off the spare leads before your power is on. All these bend, they're far too close to the polarity check, is labelled correctly. Two polarity check is in a non-standard place, so the polarity doesn't actually apply. I'm just going to have to figure it out first, principles. A tiny little bit of solder left sticking that on. Interestingly, two dead capacitors, when I was preparing for this, I actually got out too many 100 microfaric capacitors. No wait, there's a one microfaric capacitor. Good job I looked. Do I have any more 100 microfaric capacitors? This is my big box of capacitors. There's a one, 22s, and I think that's another 22. 100 microfaric capacitors. I need two of these. I'll use a spare. 10, 22, come on. I'm going to use this one. This is going to be a thoroughly exciting video watching me flail with capacitors. I take them out, and I put them back in again. I take them out, and I put them back in again. This thing will either work or not work, or I will run out of capacitors. One of the three options. Right. You have two capacitors, one of which has rather bent dodgy-looking leads. Just double-check it is a 100 microfarad. 10 volts, 25 volts. 10 volts, quite small. This was one of the reasons I didn't want to use the ones on the strip. Am I sure I don't have any more 100? 100 microfarad. Right, another fat one. I can go back in. All I need to do is get all these flipping holes cleared out, and put the capacitors back in, but this time, on their side, it's perfect. That doesn't happen often. Soak it straight out. Cleanly. By lots of heat, the board's not that fragile. I go through. And that's clean as well. It's not necessarily the right one, but it's in roughly the right place. The other capacitor is going to be a little bit more exciting because the previous two holes I used were right next to each other, so it would make sense if I was putting them in standing on end. But I'm not quite going to do that. Anyway, let's try and put this one in first. Positive on that direction. Negative on that direction. So it's here and here. Bend it over to there. Bend to your clearance. It's a little bit more bend to hold it in place. I said to hold it in place. I don't mind what I'll do. Again, this is one of those situations where I need multiple hands. So I can prop it up on that. I want a capacitor at a jointy angle. Don't push out the polarity. Positive that way, negative that way. Now I have two more holes to clear. It's right next to the edge of the board where the clearance is minimal. And I can use this hole and this hole, which are the right distance apart for the radial capacitor. That's what I did last time. So I think this is going to have to go here on its side with the leads bent like that. So it goes in that way around. And I'm going to have to cut these leads off now because they're going to get in the way. Wow, there's a solder blob and a half. That does not actually connect to three in a row. That's amazing. Positive. It needs to go through three round. I bent these the wrong way. Positive in here. Sitting on top of her. Is she going to work? Am I going to have to a different hole? Yes, I'm going to have to a different hole. I cleared the wrong one. Well, it'll technically work. It's electrically great, but I can get a much better fit by putting it through this hole instead. I bet this one will take ten minutes to get clear. Yep, that did not work. That is working. It's just being annoying because it's... Hardware is always annoying. There's always something going wrong. My miserable soldering skills are not helping. Back out a bit. Go for that. There you go. That was the last capacitor. Double check the polarity. That one's positive. That one's negative. It's a big close to the edge of the boards. Let's tweak it around a bit. The leads off. Now I want to look for solder splashes. Let's try that in the case. Let's see what happens. The case goes here. It's in. It needs to go back into my capacitor box. Does the lid go on? The regulator is sticking up too far. I'm going to have to try and get this flush, and that will be a hell of a job. It's just a couple of millimetres. I have to heat up all three joints simultaneously. My arm is about to go off into casing. That was the heatsink. It wasn't important. Wow, that actually worked. I can't believe it. That was a pretty dumb thing to do, but it actually worked. That's flush. It won't go in any further. The regulator was bought from a spectrum supplier as the recommended regulator for the board. So I assume that it will actually fit in the case. It would be pretty annoying if it didn't. Let's put this on. Oh yes, perfect. Excellent. It's electrically complete, assuming it works. I need to fit the keyboard. I actually have some visual aids over here. The way the keyboard works is there's this grotty thing. This is the keyboard membrane. These tails go through slots. This is the original one, which is 36 years old. It fits on like this. The rubber mat, which is rubber and mat-like, goes over the top. You gently massage it into place. When you press the buttons, it pushes down against the membrane which closes the circuit. It's incredibly crude and amazingly cheap. The plastic the membrane is made out of, are we still in focus? Yes! The membrane the keyboard is made out of perishes with time. To be fair, I don't think Sinclair was really intending these to work last for 36 years. So I have bought a new membrane. There are people still making them. This one feels so much better quality than the Spectrum one, than Sinclair's one. So that neatly drops in over here. The keyboard mat, which is... I washed this. Why is it covered in stuff? The keyboard mat fits over like so. The aluminium faceplate, which is this, goes on here. It's fastened down with double-sided sticky tape. I haven't put the tape on, so there's just a little bit of remnant glue, because I'm going to want to wash that mat again. But there's just enough glue to make it more or less fit. And this, these tails plug into these very, very cheap and nasty ribbon connectors. They don't actually push in at all. You just push them down and nothing moves, but then they don't come out again. They feel really dubious. And the lid goes on. And if the computer guards are generous, I now have a working ZX Spectrum. I'm going to have to give this a try. Okay. I need to do laundry. Okay, well, that was a little bit embarrassing. I did actually fire it all up live, for the first time ever on video, except I got the on and the off buttons mixed up in an open camera. And as a result, no video. Anyway, as you can see, my little Chinese reversing camera is showing sod all. So something somewhere is wrong. This is connected to composite. I had this old spiel about how great they were, which is connected to this. So something's not right. So let us do a bit of sensing and find out what. The first thing to check. Oh, yes. I mentioned earlier is my new keyboard. Not sure if you can hear the clicks. So I know that's the right noise when you press keys on the spectrum, the speaker clicks. So I know that most things work. It's just the video doesn't. And there's a variety of possible causes. One is that these pots here, which calibrate the video circuitry are incorrectly set. And my digital composite monitor doesn't realise there's a signal there because things aren't right and is refusing to turn on. It's a bit... have I put it... punctilious about its signal. So I see about 3 volts. Think that's about right. It's the right way around at any rate. OK. So I've got some instructions on how to calibrate the video using an oscilloscope. You can't see it on camera, but I have this amazing Tetronic 7603 oscilloscope weighing 13 kilos over here on the left. So it's a shame you can't see it. It's really nice. It cost me 100 francs from somebody in Zurich. I had to go to his house twice because the first time I didn't bring anything strong enough to carry it with. But there we go. It warms up. It occupies like half my workbench. So we want to connect composite out. 3 volts go there. And that to go there. That's ground to signal. And you can see it's 3 volts. So... Twiddle this one until we see that. Twiddle this one until video signal syncs. It's not happy, actually. DC AC? So I can see a signal and it's about the right size. Let me set that to ground. You know, I really need to set up the camera so you can see this. There's no actual point to me doing this on video unless you can see the oscilloscope. So give me a sec and I'll give that a try. I really hope that I've got on and off the right way around this time. All right, here is my fabulous oscilloscope. You can only see about half the buttons because seriously there are a lot of buttons. That's actually the cable for the microphone. So... Set this to AC and you can see we are seeing some signal out of the video port. That's why it's not triggering. Actually, that's right. It's set to trigger on channel 1. Here is the trigger level. You see something happens and it thinks it's triggering. Oh! Stupid me. Right. Here we go. What you can see here is the video signal. This is a frame. These lines should be flat. Not fuzzy. Let's crank this up. Ow! That's horrible. That's horrible. Okay. Yes. This means that all the adjustments are all anyhow. We need a small screwdriver. That's a small screwdriver. Very carefully. Let's try a bigger screwdriver. I need to adjust some of these pots and see what happens. This one too. That's worse. This one too is worse. Better. This one too. Just as little as possible. A slight difference but not much. This one. Apparently the top one does very little. Yeah, nothing at all. Let's keep twiddling these. You were amazing hysteresis in these things. That's me touching the pot with this tip of the screwdriver. I think that's about as good as I'm going to get. However, I do notice that the monitor has not actually lit up. It obviously thinks something's wrong. Let me power cycle the monitor. The other thing is I need to try both of the monitor ports. It's supposed to auto switch. That's interesting. If you were good enough at video you'd be able to identify all the various features of the signal. This will be the actual frame data. These will be at the top of the frame and will be, I think they call it the front porch. It's something to do with telling the monitor that there's actually a signal. If I zoom in quite a long way unless I've got this all backwards am I actually seeing individual scan lines 20 microseconds? I think I am, you know in which case the if I zoom out, it's very flickery but you can just about see there are can I make it sync? Not quite. You can see there's periodic notches that might be the top of the frame. I wonder if there's a there's a whole bunch of buttons down here to do the triggering and I think I remember seeing something about video. There may be a video mode try some of these buttons and see what happens. That one coupling AC apparently not. Internal somebody has you can just see that there. Somebody has nicely marked the enormous number of control buttons with black blobs for the sensible simple option. So before I'm looking at here as a frame I can use I can't use that. I can use this. That doesn't really want to do what I want. I'm sure I want delay triggering. Yes my wonderful oscilloscope is also quite new and it's got a lot of features I've never used before. Anyway let's have another fiddle with the pots see if I can make that a little less flatter a little more flatter. That one does nothing at all. Later versions of the spectrum didn't have these controls and they had a self-regulating system instead that was better. Right. So I think that's about as good as I'm going to get given my skill level and what I'm seeing on your oscilloscope. So one thing I've noticed is if I zoom in quite a long way you can see there's actually multiple lines. This will be because that's like 500 nanoseconds a segment. So I'm probably seeing individual pixels if there were any pixels. I noticed the default image and as I expected is a white screen with copyright messages at the bottom. Given how much I've been typing it's probably quite a lot of text there as well. Let's try power cycling the spectrum. Off. On. Yes. And you can see stuff happen as it cleared the screen. Sorry that was my multimeter telling me I haven't turned it off. The other potential option is that this monitor is just not happy with the spectrum signal. Apparently it happens. Now I know that my big TV in the living room can does cope with the spectrum signal or at least it did before I started fiddling with the pots. So I think I might give it a go on that. The trouble is I'm not going to be able to video it. So wish me luck and if these actually work I'll give you a photo. But I'm afraid that's all you're getting. Good news everyone it works. So I found I had a really nasty cheap composite to HDMI up converter and it doesn't really like the signal so it's grainy and it's watery and you can see it's flickering. It doesn't want to sync on to the spectrum's archaic composite signal. But it does work and it's showing me an actual black and white image. That'll be the reason why it's white is because I adjusted all the pots using the that was interesting. Adjusted all the pots using the oscilloscope. This will produce a much better picture on my big TV but of course you can't see that. If I can focus. Yeah I'm afraid it's just fuzzy. But like it works. I can type on it and everything. It's amazing. So let's see if I can get some colour on this if I can remember how it works. Paper. No. Paper. Symbol shift. You can't actually type keywords on this you know. Continues not the one I was looking for. Yeah that's not spectrum noise you're seeing when the screen goes glitchy like that. It's the upscaler failing to sync on to the signal. And in fact you see that white bar that's appearing. This explains what I was seeing in the oscilloscope. On the oscilloscope I was seeing sequences of strong signal with a band on one side. And I was wondering whether that was frames or scanlines. Well that band we were seeing is that white stripe so I was obviously seeing scanlines. I think I actually need to press both buttons. Yes. See the cursor has turned into an E. That means that I am in the right mode to press this button. And get L print. Typing on the spectrum. Wow I'd forgotten how grim it was. Continue. No. Do I get that? Sim shift. I tried that. So it's giving you a question mark. Both buttons to E mode followed by caps shift paper. Paper. Yes. What colour do you want? Red. Enter. Okay it did something. CLS. Enter. There you go. How's that? Actual red. On a 36 year old ZX spectrum which I repaired myself and it still works except that picture didn't be green. That'll be sync problems from this up converter. Fantastic. I'm going to sign off now because if I keep going for any longer something's going to go wrong and it's going to be embarrassing. So there you go. Fixing a ZX spectrum.