SVI 328 Repair - So Many Things Wrong With This Computer!!!! 😱😱 (Part 1)
ฝัง
- เผยแพร่เมื่อ 5 ก.ย. 2024
- Support Noel's Retro Lab on Patreon: / noelsretrolab
This week I attempt to repair a Spectravideo 328, the direct ancestor of MSX computers. Unfortunately, the poor computer has all sorts of problems: shorts, blown fuses, missing pieces, broken traces, horrible soldering, and more. How did it ever get this way? And, most importantly, will we be able to revive it?
Connect with Noel's Retro Lab:
Discord ➤ / discord
Facebook ➤ / noelsretrolab
Twitter ➤ / noelsretrolab
Instagram ➤ / noelsretrolab
I've no idea why this channel appeared in my recommended feed but I'm really glad it has. Fascinating stuff.
Awesome! I'm glad you're enjoying it 😀
It popped up in my feed, but I was looking for ZX spectrum stuff. I'm fascinated, can't understand most of what's going on but lots of videos have been very useful, in answering questions I've had.
I really really like how you show "cartoon" probes on the schematic whenever you're probing the board. It makes everything so much more clear.
I very much appreciate that the computer has so many problems. That should be a very good lesson in systematic fault finding then.
Great video! Its nice how you show on the diagram what you actually do on the board. A perfect way for others (people like me who are not into electronics) to learn how to read and understand how to use schematics when tinkering with old computers 👍🏻 Great job on this MSX, and thanks for some excellent videos on your channel!
Thanks for the kind comments! Glad you enjoying the videos.
Instablaster.
resurrect the poor computer please - it's been clearly tortured... it deserves a better future!!
I'm hopeful about it. We'll see it one way or another in a couple of weeks.
Any chance you might replace those awful glued connectors with pin headers and a shiny rainbow cable? I can't bear the sight of that glue :)
I'd love to because it's so uncomfortable working with all those parts connected, but I probably won't get there. That ribbon is probably the membrane, right? I guess I could easily replace the video board, but the keyboard would be much harder I think.
@@NoelsRetroLab Yeah, that's a good point about the membrane. Stupid designers! :)
My theory held up until the end of this part. I feel that someone decided to 'upgrade' the video RAM. Thus causing a short blowing the fuse and voltage regulators. It looks like they salvaged the regulators from an old Sony hi-fi, threw them in. Probably threw fuses at it until the regulators blew out. Maybe they knocked the crystal off while 'working' on it, who knows. Very interesting.
Yes, that sounds like a very plausible theory of what happened to this poor computer.
I actually really like the design of this machine, especially the keyboard! It's a little weird having the cursor keys above the numpad, but at least they're in an inverted-T arrangement.
Shame this one was vandalised, but I'm sure Noel will fix it. =)
With the reg mesuring 14v, I thought everything would be fried. But I guess, with it being shorted on the board, the chips were saved. Cool project.
Fortunately that was only the 12V rail, which was video RAM and one other chip (don't remember, maybe the video one?). But yeah, I'm surprised things seemed to work at the end. We'll see how many more surprises there are.
What could go wrong?
FUSE SAYS PHOOOOOM! Rofl
I think the missing crystal is for the TMS9918 chip. Maybe they were trying to replace it with the TMS9928, a PAL version of the chip.
Yeah, I've seen those kinds of ugly conversion jobs in NTSC Commodore 64's that were brought to Argentina and were adapted to the weird Pal-N we use here.
That board is a bigger wreck than my marriage.
LOL "probably caused Havoc in the computer"... i laughed
love your videos Noel, keep up the good work! I'm looking forward to seeing the next part of this video to see if you get it back to life, and I'm sure you will :-)
Thanks! I'm cautiously optimistic.
That was my first computer and it died with a blue screen, so sad it was eaten by a flood. It had an amazing game library.
My mum did the bodge jobs for STC locally and her soldering was not very good.
It makes me so sad to see such a poorly computer; i really do hope that Noël can bring it back to life.
Great video, i understand all. Less part about "ULA" in center, why this is there? And how works ? ULA of z80 is no enough ?
Great use of systematic fault finding
Thanks, although if you really want systematic, make sure to check out the Inves Spectrum video! 😃
What a crazy board
very good video, It's helping fix my SVI-328 but i cannot find the schematic you show in the video. Would you help me with the link to it?
You have done well to get that far, the last person did a bit of a butcher job for sure! But stable Z80 and hopefully good main Ram. It will depend whether any over voltage made it to the TMS chip as they are a little sensitive.
Hi, I got one SVI 328 in working state , let me know if you need I take some pictures inside and check another thing for help you fix this one, cheers!
Thanks! I really appreciate it. I might get in touch if I run into trouble. Right now I'm cautiously optimistic about it.
Fantastic. Thank you!
Haha, yours needs some cleaning! I can't even see the letters in the system RAM :-)
You can quite easily replace 4116 with 4164...
The 12V regulator should not do a lot of damage, because there are probably another capacitors on 12V line on the board, so even without the one on the breakout board it should work correctly.
Yup, I've been shooting some of the adapting of 4164 into 4116 today. Lots of practice doing that from the Speccy :-)
My favourite 8 bit. Great machines.
Whoa! WTF did they do to this thing? Possibly someone trying to learn as they went? Not my project, but I am hoping you revisit and replace those 7805 and 7812 rectifiers. I don't like that "Strange Sony Satelllite" part. Great start. Looking forward to part 2.
I would like to use the original regulators as well, but I don't have a 7812 handy and things would take forever to arrive during the quarantine. I tried putting a 7805 but didn't fit in the board, even after I cleaned all the solder well. I wonder if there are 7805 models with thinner legs.
I know how that is, you need to work with what you have available. :)
These 3052V and 3122V *regulators* can apparently deliver higher output currents than the 78xx ones (a further indication that they were salvaged from a different machine). A datasheet for the 3000V-series is available online (the 05 and 12 in the middle of the name are obviously the voltages).
While this output performance is probably not needed (if the SVI indeed uses standard 78xx regulators) it could lead to a lower heat dissipation.
Also Noel measured very accurate voltages. Therefore my opinion is that these regulators are fine.
However, *if* this SVI is actually repairable (could be) and if one were to make it future proof (why not?) one would at least replace the crappy caps on these tiny boards and redo all solder joints - or better yet: drop 78xx-pin-compatible modern regulators onto the mainboard.
These 3052V and 3122V *regulators* can apparently deliver higher output currents than the 78xx ones (a further indication that they were salvaged from a different machine). A datasheet for the 3000V-series is available online (the 05 and 12 in the middle of the name are obviously the voltages).
While this output performance is probably not needed (if the SVI indeed uses standard 78xx regulators) it could lead to a lower heat dissipation.
Also Noel measured very accurate voltages. Therefore my opinion is that these regulators are fine.
However, *if* this SVI is actually repairable (could be) and if one were to make it future proof (why not?) one would at least replace the crappy caps on these tiny boards and redo all solder joints - or better yet: drop 78xx-pin-compatible modern regulators onto the mainboard.
These 3052V and 3122V *regulators* can apparently deliver higher output currents than the 78xx ones (a further indication that they were salvaged from a different machine). A datasheet for the 3000V-series is available online (the 05 and 12 in the middle of the name are obviously the voltages).
While this output performance is probably not needed (if the SVI indeed uses standard 78xx regulators) it could lead to a lower heat dissipation.
Also Noel measured very accurate voltages. Therefore my opinion is that these regulators are fine.
However, *if* this SVI is actually repairable (could be) and if one were to make it future proof (why not?) one would at least replace the crappy caps on these tiny boards and redo all solder joints - or better yet: drop 78xx-pin-compatible modern regulators onto the mainboard.
Really interesting. This is in remarkably similar (poor) condition to one that I have. Although mine is a mk1 and yours is a mk2 so the motherboards are quite different sizes.
Yes, having the MK1 and MK2 makes the service manual and schematics much more confusing. For a while I was trying to make sense of the power circuit until I realized I was looking at the MK1 version. The MK1 comes with 16KB of RAM, right? Is it any different on the outside?
Noel's Retro Lab No it still has 64k of RAM (plus 16k of VRAM). You might be thinking of the 318? On the outside, it has a slightly different badge (the mk2 has a little red square and is SVI branded I think) and the case has slightly different holes at the back, even if the ports themselves are the same (the cassette port and the monitor port share a gap in the mk2, but are separated by a bit of case in the mk1).
stopped the video to like and comment the events on 11:00 LOL
That was my Hollywood explosion editing technique :-)
@@NoelsRetroLab Yeah famous last words: "What could possibly go wrong?" hehe
I believe those voltage regulators to be original. It did not give 12V as the regulator would be oscillating without the cap connected.
There isn't much information about this computer out there, and there are fewer pictures, but the ones I saw had regular (7805 and 7812) regulators on them. Also, notice that's the same cable as the one dangling from the oscillator and the bridged trace in the back. So at least they were pulled out for some reason. And I guess I was assuming the voltage was already rectified by the time it got there, that's why it surprised me. The whole experience was super weird.
I don't think so. The cabling is identical to the other "botch wires" like for example on the missing crystal.
Also the screw positions look like could be ideal for 78xx regulators in their usual TO-220 package.
These regulators are bigger, though, and their PCBs are marked with "Sony". They look more like salvaged items from some other machine.
@@oldguy9051 Yes I agree, just seen the Sony logo on the brown PCB's .
what is the purpose of the green connector in the center of the board?
That's a cartridge slot! It's narrower than a standard MSX cartridge because, well, it wasn't quite MSX yet.
The horrors...
Espero desenlace. Good job.
This looks like a teenager who didn't know much about digital electronics and had no guidance was trying random things to "upgrade" his computer. I bet the leads to the oscillator circuit went to an external socket so he could try different crystals to see how fast he could make it run. I bet he thought the stock voltage regulators were also needlessly reducing the performance of his computer so beefier voltage regulators would give the computer more "power".
Haha, I hadn't thought about that, but I could totally see that happening! 🤣
People criticise Sinclair build quality but really they were a paragon of quality compared to some if these HK-originated machines like these and the Vtech machines. Nice specs though.
To be fair, other than going cheapo on the keyboard connectors, the rest of the board seems fine. It seems better quality than Sinclair (and that's not even looking at the keyboard). The problems seem to come mostly from people trying to fix it :-b
That is abomineering at its worst.