Amstrad GX4000 / Plus cartridge PCB

[Chinnery]

I'm releasing my current design for the Amstrad Plus Cartridge PCB.


There is one known issue as far as I'm aware - if you have a reset switch on your Plus machine, there is about a 50/50 chance the board will not initialise properly if not using the Acid option. A power cycle ALWAYS initialises properly. I know there is a fix for this but help has not been forthcoming.


The 2 diagrams attached aught to help with component options for the Acid variant and Non-acid variant. Obviously you will always need an eprom. I have labelled on the silk screen what components go where so it should be easy to sort out.


The ZIP file contains both the eagle design schematics and the board layout file, so people who like to tinker can. There is a sub folder called "gerbers" which you can just upload to your PCB manufacturer of choice to get them produced. I've tried both JLCPCB and Seeed Studio Fusion with great results.


Enjoy!

[TotO]

Cool! Any project with it? 
Yes, the no acid design (nocash) have this known problem.

[00WReX]

Very nice of you to provide this to the community.

Excellent work.

Thank you.
Cheers,
Shane

[Chinnery]

Cool! Any project with it? 

No projects - all my project seem to reference the wrong items anyway    (Eagle noob)

Yes, the no acid design (nocash) have this known problem.

I've heard there is a workaround, but I've had no joy in finding this out. I use these for GX4000 stuff anyway, so no great harm.

[Bryce]

No projects - all my project seem to reference the wrong items anyway    (Eagle noob)
I've heard there is a workaround, but I've had no joy in finding this out. I use these for GX4000 stuff anyway, so no great harm.

The workaround would most likely involve taking the reset signal to the cartridge which wouldn't be praticle or at least mean modding the CPC.

Bryce.

[00WReX]

Thanks to @Chinnery for providing the Gerbers, I had a batch of boards made up just for fun.

I have just finished 5 boards for my own 'Arcade Classics" collection, and they work perfectly.

Cartridges made:                                                  EPROM USED:
1)Frogger - The 'Executioner' arcade port 2007          M27C2001
2)Ghosts'n Goblins GX4000 - Xifos 2018                    M27C2001
3)Space Invaders Arcade Emulator - 40Crisis 2016     M27C2001
4)Bomb Jack I - Phantomz conversion update             M27C4001
5)Scramble - PuzCPC 2019                                        M27C4001

Next plan is to get some 3D printed cases made up, and put some stickers on them to make them appear somewhat professional.

[attach=1,msg175498]

Cheers,
Shane

[Gryzor]

How long till copies of Pang and Chase HQ XIII appear on ebay?

[tjohnson]

How long till copies of Pang and Chase HQ XIII appear on ebay?

Good point, I'm sure people would like to see Chase HQ 2 appear, one day maybe one day.....

[BeaglzNZ]

Just wondering if you have managed to create a STL for a case for these cartridges? I have got a few PCBs and awaiting parts to assemble - any idea if there is enough clearance for a ZIF socket?
Cheers,
Antony

Thanks to @Chinnery for providing the Gerbers, I had a batch of boards made up just for fun.

I have just finished 5 boards for my own 'Arcade Classics" collection, and they work perfectly.

Cartridges made:                                                  EPROM USED:
1)Frogger - The 'Executioner' arcade port 2007          M27C2001
2)Ghosts'n Goblins GX4000 - Xifos 2018                    M27C2001
3)Space Invaders Arcade Emulator - 40Crisis 2016     M27C2001
4)Bomb Jack I - Phantomz conversion update             M27C4001
5)Scramble - PuzCPC 2019                                        M27C4001

Next plan is to get some 3D printed cases made up, and put some stickers on them to make them appear somewhat professional.

 

Cheers,
Shane

[Bryce]

Just wondering if you have managed to create a STL for a case for these cartridges? I have got a few PCBs and awaiting parts to assemble - any idea if there is enough clearance for a ZIF socket?
Cheers,
Antony

There's definitely not enough space for a ZIF socket. Even a standard low profile IC socket will require a hole in the lid for the IC to poke through.

Bryce.

[Philippe Lardenois]

There one STL on Thingverse for normal Cartridge.


quite easy to modify. Will have a look next week.


Done - the bottom is the same that is avalaible on Thingverse thanks to Wolfiem.
The top has 2 versions : Hole to seen the eeprom and change it without opening the cartridge / hole covered otherwise.

+

[Cwiiis]

I took it on to design a shell that's a bit closer to the original. There are some compromises due to FDM printing limitations and laziness, so it's a little bit boxier than the original and the details aren't all 100%. Biggest/most obvious changes are that the front and back sticker recesses are marked out but not recessed as that makes them near-impossible to print well, and the sides of the cart aren't tapered. Smaller differences are that the bottom edge also isn't tapered and the grips are a little different because it's impossible to print that detail with a 0.4mm nozzle.

This should be printed with the flat surface on the print bed and with supports everywhere. They print side-by-side fine and with my personal settings, it's a bit under a 4-hour print. It's designed to be printed with a 0.4mm nozzle and 0.2mm layer height. Once support is cleared, the pieces will snap together - if you're lucky, you can also pry them apart, but that may break the clasps (done it a few times here and no breakages yet, but that surprised me). Enjoy!

[Shining]

Somewhere else I've written that already.


after analyzing the oszillograms from octoate and nilquader in the past I found a solution for the reset problem with a simple rc-combination on the preset pin.
I never had any issues with a cart using this and I always use a reset-key.

[Chinnery]

I've attached my Eagle project (Sch/Brd) should anyone wish to implement Shining's reset fix. If you do and it tests well can you please repost back here? (I have no plus machine to test with)

[TotO]

Somewhere else I've written that already.

after analyzing the oszillograms from octoate and nilquader in the past I found a solution for the reset problem with a simple rc-combination on the preset pin. I never had any issues with a cart using this and I always use a reset-key.

I have tested this schematic by curiosity, and it looks to not work at all.

May be I have done something wrong...
I don't understand what does special your RC combination compared to the original schematic.

[ChrisJames]

Somebody know this cartridge

https://twitter.com/anakintf/status/1459907093780217856

[TotO]

Well... It is probably the C4CPC grandfather.
Nothing official here, despite the Amstrad logo.

[gerald]

Somebody know this cartridge

https://twitter.com/anakintf/status/1459907093780217856

Looks like a cartridge with a flash instead of eprom.
The red/black wire are used to carry WEN/OEN to the cartrige PCB (these are not wired to the cartrige connector).

So it's a nice way to avoid unplugging the flash from its socket when you need to reflash it.

[ChrisJames]

Well... It is probably the C4CPC grandfather.
Nothing official here, despite the Amstrad logo.

It looks something similar to what you want to do with the PLAY2CPC 

[abalore]

It seems easier to have a cartridge PCB with a ZIF socket for development.

[TotO]

It looks something similar to what you want to do with the PLAY2CPC 

The Play2CPC allows to write the cartridges directly from the CPC.
It is more close to what is doing the X-MEM or the good old CTC-AY project.

[BeaglzNZ]

Just wondering what thickness do you use for the PCBs? I got some done at the default thickness which is 1.6mm I think but it seems a bit too thick?
Thanks.

[Bryce]

1.6mm is the correct thickness. The card pins need to make good contact.


Bryce.

[Curlypaul]

Just wondering what thickness do you use for the PCBs? I got some done at the default thickness which is 1.6mm I think but it seems a bit too thick?
Thanks.

Are you willing to sell one? Looking to create a demo, and if I can't get a C4CPC, one of these would be the back up plan

[Shining]

I have tested this schematic by curiosity, and it looks to not work at all.

May be I have done something wrong...
I don't understand what does special your RC combination compared to the original schematic.

Perhaps my schematic was not clear, cause I missed to describe the signal on the RC-combination:


On the original oscillogram for acid reset (https://www.cpcwiki.eu/index.php/File:ACID_Init_by_Grim.png) you will notice that after startup the sin signal has to be high. Using the original schematic from nilquader, the sin signal after reset is 50% low or high cause it does not follow the reset at all. Now instead of pulling the RC to VCC, I used the CCLR signal not only for CLK+ but also for PRE at the flip flop to force the flipflop to "1" after reset. Also I lowered the time constant for the RC-combination.


I know quite some plusses where this always works and noone reported a single problem until now.