CPC6128 - new schematic capture and layout for MC0020x boards

[pelrun]

I was recently possessed by a strong need to procrastinate the need to accurately chase down signals on my 6128 while working on another project, and once again the lack of accurate information in the service manuals for my board revision (40010 gate array and 24-pin data separator) irritated me immensely.

So I finally did what I've been tempted to do on multiple occasions in the past, and properly captured everything in a KiCad project.

And it was even more of a pain to finish than I expected, but it's basically done to a reasonable level of accuracy (pcb/placement/references should be really close, parts may have incorrect values.)

https://github.com/pelrun/cpc-schematics

It also means I can export an extremely neat interactive board explorer as a website:

https://pelrun.github.io/cpc-schematics/index.html

[Gryzor]

Pretty 😌

[Rabs]

I was recently possessed by a strong need to procrastinate the need to accurately chase down signals on my 6128 while working on another project, and once again the lack of accurate information in the service manuals for my board revision (40010 gate array and 24-pin data separator) irritated me immensely.

So I finally did what I've been tempted to do on multiple occasions in the past, and properly captured everything in a KiCad project.

And it was even more of a pain to finish than I expected, but it's basically done to a reasonable level of accuracy (pcb/placement/references should be really close, parts may have incorrect values.)

https://github.com/pelrun/cpc-schematics

It also means I can export an extremely neat interactive board explorer as a website:

https://pelrun.github.io/cpc-schematics/index.html

This is excellent. If only I had this a few weeks ago. Spent hours scratching my head trying to understand the amendment to the service manual.    thanks.

[Urusergi]

wwwwwwooooowwwww 

It's amazing, incredible, wonderful, beautiful, nice, pretty, lustrous, beauteous, sparkly... (add more to the list)

It also means I can export an extremely neat interactive board explorer as a website:

Is an offline version possible?

[pelrun]

It's fully contained in a single html file, so you can just save it locally. I just pushed it to github pages because it was trivial to do so.

[ikonsgr]

And i thought i was obsessed with amstrad cpc.... 
Very nice work, and it will surely help for repairs of this motherboard version, which i think it must be the most popular too (and it's not covered by existing service manuals...)

[WacKEDmaN]

umm...theres this already...
https://www.tindie.com/products/bobsbits/amstrad-cpc6128-replica-pcb/

...edit: sorry not really on topic! ,,,but ill leave it here anyways coz i cant delete the post!

[pelrun]

Oh, I was waiting for someone to pop up with "hey, someone already did this". Although I'd consider that exactly the opposite - no design files but a physical pcb (boo). 

Although I *really* wish I'd had that top pcb photo earlier, it would have saved me a lot of work! At least I can use it now to adjust/verify what I have.

[pelrun]

Checking against Bob's clone pcb image, I found and fixed several errors in the disk circuit and a bunch of routing differences elsewhere. So thanks for not deleting your post, @WacKEDmaN , even if it was because you weren't able to

[xesrjb]

I was recently possessed by a strong need to procrastinate the need to accurately chase down signals on my 6128 while working on another project, and once again the lack of accurate information in the service manuals for my board revision (40010 gate array and 24-pin data separator) irritated me immensely.

So I finally did what I've been tempted to do on multiple occasions in the past, and properly captured everything in a KiCad project.

And it was even more of a pain to finish than I expected, but it's basically done to a reasonable level of accuracy (pcb/placement/references should be really close, parts may have incorrect values.)

https://github.com/pelrun/cpc-schematics

It also means I can export an extremely neat interactive board explorer as a website:

https://pelrun.github.io/cpc-schematics/index.html

Why a card edge connector? Centronicsport are better fmpov...,they are more stable an relyable...

xesrjb 

[pelrun]

The project has edge connectors because the original boards have edge connectors

Making new pcbs directly from these files would be pretty dumb. Mostly because if you have a gate array you almost certainly have an original pcb already, but also because nobody should be using through-hole parts on pcbs in this day and age. This is intended purely for documentation and debugging purposes, which is why I deliberately didn't bother sorting out the mounting holes or making a proper board outline.

And the centronics footprints are there too, didn't you notice?

[Kris]

Why not to integrate directly some improvements ? (RAM...)

[pelrun]

Despite quite literally answering that exact question in my previous comment, I guess I can explain my self again. 

This is supposed to be *exactly* the same as my original 6128, because I'm using it to troubleshoot that specific board, and until now there has not been *any* good reference material for it, and the service manuals are poor or actively misleading. Making random modifications would be completely counterproductive - nobody is going to make a modified CPC board in sufficient quantities to be interesting, and I already have two MC0020x CPC's, thank you. I'd rather design a peripheral to go on a real CPC instead.

If someone wants to take these files and make something else out of it, then there's nothing stopping them. It's just silly to expect *me* to do it for you.

[orzel]

This is very interesting. Especially as kicad files are available (contrary to..). Thanks a lot for sharing !

I have one question : has it been validated ? I mean, how sure we are that it would work once all proper components are soldered ? I understand gatearray and some other chips are hard to source without cannibalizing a true original cpc.

No offence intended of course, I'm really amazed by the work done. But you can't be sure with such a complex project without actually testing.

Out of curiosity, did you produced the latest version ?

[pelrun]

I'll say it again, you're not supposed to make new PCBs from this!

If you insist on doing that, go buy one of the repro PCBs that someone put on tindie, they're created from a simple scan of a real board rather than a ground-up recreation of the schematic, and so will be identical.

[orzel]

I'll say it again, you're not supposed to make new PCBs from this!

If you insist on doing that, go buy one of the repro PCBs that someone put on tindie, they're created from a simple scan of a real board rather than a ground-up recreation of the schematic, and so will be identical.

I hadn't noticed you saying that, maybe because of my poor English.

The ones on tindie, you don't have the design files, i'm not interested.

[orzel]

Hi,

Just checking your design against actual hardware. I've noticed the two following differences:

1) Just above the T of the AMSTRAD, between the power connector and ic207, there should be an electrolitic capacitor, C136.

2) At the input of the sound amplifiers, C314 is an analytic on all pictures of mainboards I can can find, but it's a small ceramic on your project. Far less important than the previous one, I agree.

greetings

[pelrun]

Yeah, looks like C136 is entirely missing! 

C314 is an oddball - it's actually non-polarised, and I didn't have a good NP footprint that fit in the space, so I just stuck in a ceramic footprint at the time and then forgot about revising it later. 

Thanks for the proofreading!