ZX Spectrum Part 1 - It's a very sick machine (and a BBC Micro update)

Hello e

very

one, welcome back to Adrienne's digital basement for today's video. I'm going to take a quick look at the Sinclair ZX Spectrum that I brought from England with me on my recent trip, if you haven't seen my last video that's where I show it. the computers I brought, you can look for a link in the description or put one at the top of the video on each side. I can't remember which one before I do it, I want to talk quickly about the BBC

micro

computers. I have both. on the workbench and are roughly assembled from the last video.

I actually reinstalled all the

part

s, although things are a little confusing between the two, but I installed plugs on both cables now where we left them. off one of the two power supplies was not putting out the minus 5 volts, this was the power supply that was in the computer with the clean keyboard and the hole in the case that I just removed when I turned it on in England. I didn't hear any sound or I could just

very

faintly hear the sound through the speaker and I found that when I put the power supply with minus 5 volts on either of the computers with the corresponding motherboard, the sound worked fine on both. computers, so you definitely pointed out minus 5 volts missing as a reason for audio not working, it makes sense, usually op amps or the amps that drive the speaker need more and less volts to work properly and since it was half missing What was probably wrong was the voltage.

There were a couple of people who commented on that in the comments section and I think you were absolutely right. This was the actual power supply that was giving me problems and lo and behold when I turned on this computer. We have completely The sound works now, in fact, just for fun. I actually burned an EEPROM here from Arkanoid and run it. We have fully functional sound and everything else works too. Now you may be wondering what I did to fix the minus 5 volts and I wish I could. I could tell you because I didn't really do anything.

I had the power supply turned off and was investigating why I wasn't getting minus 5 volts. Looking at the schematics for this power supply showed that the minus 5 volts are actually correct. Out of the transformer it has a center tap in the foreground and a +5 on one side and a minus 5 on the other and it comes out of that it goes through an inductor on just a couple of passives and that gives you the minus 5 when I started Poking around and testing Lo and behold, the minus 5 was working again, so I have a feeling there's a chance that maybe the transformer had a slight fault, maybe one of the windings on the minus 5 had a small break, or something was definitely missing. working when I was testing before, so I'm not really sure what's going on exactly, but it's working again.

Both computers run on minus 5 volts, which means they both have sound. It definitely bothers me when things just take care of themselves without me understanding what was going on. incorrect because that means this fault could appear again in the future. There was another small issue with the computer keyboard having no sound and a couple of keys,

part

icularly 2 and E, not working properly. I fixed both. Problems too, the e key works fine, as does the 2 key, all I did was remove the key cap and put some deoxy around the key switch, pushed it in and out a bunch of Sometimes, lo and behold, both work well.

In fact, it's a little funny because D West has a little lubricant on it, so now the e key feels very smooth compared to the adjacent W and our keys next to it, so when I revisit these two computers, everything It's going to focus on cleaning them up and making them look as good as possible, plus probably adding some extra features and cool new roms and mods and stuff like that, so I'm looking forward to that video, moving on to the

spectrum

. I put this aside without any proof. and today we're going to take a look at this

machine

and see if I can get it to work, as I mentioned in the last video, this

machine

is pretty dirty, a lot of dirt on the keyboard, there's a lot of shine, like seeing these little ones. white flecks I think those Oh no, okay, that was actually scratching the paint, for some reason there was a lot of shine on this, but there's also a lot of corrosion on this front panel, but that's okay, I don't care if this has something . battle scars, we'll see how this machine performs before we even decide what to do with it, so one like the BBC

micro

, which was quite expensive and commonly found in schools, the ZX Spectrum was a very popular home computer in the United Kingdom.

This pretty crappy rubber keyboard, at least in early versions like this, I think later revisions had real physical keys that were maybe a little better for typing. Well, this was built at a cost and because there are a lot of ports missing, we have a nine volt DC input, here we have an edge connector, an expansion port, basically a microphone, a headphone and a TV output, and I'm pretty sure when it says microphone this is specifically for connecting a cassette deck or you could use your smartphone these days but this would be a cassette deck for loading and saving your shows.

The TV is an RF output which I won't be able to use here, but there is a simple mod that will change this to composite, which I will have to do and then yes, it is normal DC. And that's it, if you want to say you connect a joystick, you need to use an expansion module that plugs into the back. I think Sinclair sold it for around £20 back in the day and it gives you two ports plus a cartridge. port there were a certain number of games released on cartridge, so let's open this up. I just need to remove these screws.

I think the screws have a bit of corrosion, a bit of rust, maybe it was just stored in a damp basement in the UK or maybe in someone's attic, okay so here we go umm, we kind of plug in the keyboard here, look at those keyboard connectors, they look like they should just pop out, there we go, so they seem to be in good shape. I imagine these are common failure points. there is a memory underneath that this key touches, it would be like a remote control or a calculator, so those ribbon cables can probably fail, so there's the ZX Spectrum.

Wow, look at that pretty interesting integrated drive issue, someone commented on my last video when they saw me taking this apart and thought this would be a problem: there was a red label, they were just there and they were absolutely right, they said the later numbers They were more reliable, so I thought I read that they came in two different ones. The RAM sizes are 16K and 48K and looking at this board they appear to be RAM chips, like these are 41 16 so they are 16K, these 8 chips add up to 16K, these chips maybe look like 32K chips, so they are also known as 432.

I actually don't think I was aware of the existence of this type of chip, so I think this might be the 48k machine, which is good because that It's exactly what I expected it to be, but I have to say I was worried. Based on the outside appearance, this thing was going to be rough on the inside with corrosion and all that, but look how shiny the RF modulator and heat sink is, there's really no dirt or dust here and this looks great, like this which is a problem to address and the UL a is a 5 C 1 1 - e - 3 it has what is known as the spider mod, which is the tr6 mod, it goes between two pins on the CPU and one of the pins on the UL a, it was to fix some kind of synchronization.

Problem I'm not quite sure if these machines were sold at 40K or if they all run at 16K and people added the memory later, but looking at these chips here, they are from '83 and seem to match the date codes on the most. of the other chips on this board, so it was either upgraded immediately after purchase or manufactured as a rear-facing forty-eightK board. I would definitely say that these chips were not hand soldered, which is why they were probably manufactured this way. You can correct me in the comments if you know more about the configurations of these machines when they were made, so the RF modulator is basically 5 volts and the video input cables and the one on the left is video input and essentially the mod alone implies removing. the resistor that's connected to the RCA jack here and cutting this wire and putting it through this hole and plugging it in right there, so I'm going to make that modification so we can test this, all the right modifications complete, so you see, I have the little wire soldered red on the center panel of the RCA connector there and I just cut the video input cable and ran it through the extra hole there and folded the other cable out of the way.

I could reverse this relatively easily so the first thing I need to do is test the polarity of the DC input connector, sometimes these are negative sensors, I don't want to put the polarity wrong so I'm going to touch the modulator of RF that is connected to ground and I am going to test the center pin. Wow, okay, so this is the center negative, so I'm going to give 9 volts from here. Let's prove that I'm definitely getting the negative center, so we'll put it in there. Turn on the nine volts. You should have more nine volts and we do.

I have the original block for this Sinclair and it says it puts out nine volts at 1.4 amps, so I'll start with a lower amperage than that, probably start with 500 milliamps. In this case, there is something short here. I don't want to ruin it. this is the DC regulator here 7805 that regulates up to five volts and the

spectrum

is interesting because you see here there is a coil. This is actually another power supply that takes the five volts and steps it up to plus twelve and then nine is twelve and one. minus five, the reason for this is that the 16K ram chips here actually use plus five plus 12 and minus five to run, so this little power supply actually makes that happen, so I have everything connected, we have the Sony PBM that the friend supports.

The RF mod is ready, so let's get a composite of this connector. I connected the power supply with reverse polarity, so the central bank power supply is set to 500 milliamps at nine volts and let's turn it on, oh look at that Sinclair research. Okay, it's working, but it's current limiting right now so it probably needs more power, so I'm going to turn this off and turn it up to one amp. Let's try again. Oh right, so we were in black and white before. but now the monitor says friend and I see a bit of color here, the low voltage was a problem as this doesn't really look good on camera because of the 50 Hertz.

Let's switch to LCD screen using my HDMI converter. Well, this is it. An annoying look at the output connected to my HDMI converter box basically shows up as Camera C. Very strange as my converter box doesn't work properly with the Sinclair. I'll just show you on this monitor here running the camera at 50 Hertz. so you can see it booted up, it looks pretty good, let me turn the machine off and on again to get the startup screen and then the Sinclair investigation, so it looks pretty good. I just wrote that program. It really took me five minutes to run it.

Execute it well. Well, it's working. I honestly can't imagine programming anything into this, but I guess you get used to it. I mean, it has a lot of shortcuts, things like that, but wow, let's see if there's a way to make the colors display well. no, I tried to put some color on the screen, it definitely looks a little weird, hard to see here, but the blue and red look absolutely fine, but the yellow, cyan and magenta have strange lines running through them. I see that they alternate. The colors also the white is not so white, but Wow, yes, I must say that I am absolutely happy.

This seems rough, but it actually works. I pressed all the keys and each key seems to work correctly. Still, I'll take it apart. and really clean this up good but there really isn't much to do with this machine other than cleaning it and maybe adjusting the video signal maybe a heatsink ok if you think that fits yeah yeah the ula gets pretty hot , so I'll stay. a heatsink on that I took a look at the video output on my oscilloscope and there was a pretty big DC offset so I added a hundred microfarad decoupling cap in line with it which seems to have fixed the DC offset To improve the image quality a bit I also decided to add some heatsink to UL, although I'm not too sure this will actually do anything, so due to the profile of the case I couldn't place heatsinks directly above the center of the chip which is where the die is and it gets very hot so I cut the PC slot cover and put adhesive on it so I cut the PC slot cover and used thermal adhesive to attach it to the UL a , which should conduct heat. you know all the way too and I've placed two small heat sinks here that may have some effect.

I have no idea if this will actually do anything. The whole computer doesn't run particularly hot apart from the UL a and the heatsink.heat. here for the voltage regulator i also have one of my old android phones connected to the computer so i am running tap-dancer here which i actually have tap-dancer installed and running it this is a perfect app for loading tape files on these old 8-bit microcomputers and I'm currently using this old Magnavox monochrome monitor. It actually works fine with pal or ntsc, so we'll use this now because it's small enough. I can show it on the computer at the same time.

I discovered a little problem with this computer, it's not working as well as I thought it seems to be a little

sick

, let me demonstrate it, so we'll turn on Sinclair and I have a RAM test program loaded here, we are. I'm going to load this into memory, well, all is not well with the ZX Spectrum. I was about to show you that when using my phone, the computer was having some problems loading games. If you loaded very old 16K basic programs, they usually seem to work. Every once in a while I would get maybe a random character a little bit of corruption the computer might freeze but if I tried to load something that was 48K needed it would usually say no memory or it would load and then the computer would crash so I looked I found a RAM test tape program and loaded it.

I reviewed it and tested it, but I'll get to that in a second because unfortunately what's happening now is the keyboard doesn't work anymore, this ribbon cable on this side has given up the ghost first, just some of the keys here don't They worked. like this bottom row and I finally got to the point where this whole side of the keyboard wasn't working, obviously I can't load any programs when that's the case, you see save here because I pressed s and that gave me save but typing got worse and worse, like this since the keyboard doesn't work I can't show you what I was doing on the camera, but I have some pictures I took with my smartphone that I will show here, so what was happening is when I loaded the RAM.

The test almost always the initial 16K RAM test worked fine, but when I tested the upper 32k it always had errors and showed all 8 chips as bad or all 5 chips as bad. Occasionally I had graphical corruption while testing. it was just all over the place and in fact just before I started recording I ran it through a test cycle and as soon as I started the program it did the test and when they got to 48k the computer started testing I said failed. and then the computer rebooted so I looked at some troubleshooting guides and it said that when you have problems with the UK top 30, these logic chips are what drive these chips, that these are optional chips that are required when you add the RAM, well I looked at all the rats and calves and wrote enable lines here and the address multiplexers and everything looks absolutely fine.

By the way, I don't see a problem with any of this, what's really strange is that this is a problem to address. Looking at the schematics to see the problem, the chips are not wired the same according to the schematics here from you two and just for fun I checked all three schematics for the problem and that appears to be the way this is wired. the calf shaver pins and address lines match the schematics in problem three and then they will match you too, but as I kept testing things actually seem to get a little worse and as the computer booted up you saw that it was in the startup screen without any problems, trying to run programs seems to cause random corruption and crashes, so I'm not quite sure where the problem lies.

I still think it's probably these 32k chips, maybe one of them is doing weird things on the data bus when accessed and that's causing these crashes, but as it stands now I can't do any more testing because I can't load the RAM test program without the keyboard and without it there is no way to validate it. If any of the fixes I make on this are actually helping, people have definitely commented that there are replacement membranes for the keyboard and I assume that means they are replaced with it as well, so where I am with this computer right now is that I really don't know what to do, it could easily become a money pit.

I could just start replacing parts and hopefully get this thing working, but part of me says maybe it's not worth it, it's going to take a long time to get the parts for this thing to try to fix it and it's not there anyway. in good condition, so is it worth it for me to try to fix this or should I just cut my losses and try to find another ZX Spectrum that actually works anyway? I have some advice and suggestions to give myself. I'd love to see them in the comments section below. I read every comment and you can subscribe to more videos.

Thanks for watching my videos. I apreciate it. Bye bye.