Hello!
I've got what I believe to be a 100% accurate KiCad Schematic of the Apple II Revision 0 system board. To do so, I did the following:
1. Downloaded the KiCad schematic for the Apple II+ from https://github.com/baldengineer/bit-preserve/tree/main/Apple
2. Reverted the schematic back to the R0 schematic as per "The Apple II Circuit Description" by W. Gayler.
3. Imported the Gerber files for the system board as created by Mike Willegal.
4. Added footprints for the components.
5. Added the footprints to the imported gerber files to ensure accuracy.
6. Fixed various schematic issues.
I didn't upload the board file as the gerber import messed up the thickness of the traces, so it needs extensive re-work to set the proper thicknesses.
Link: https://github.com/baldengineer/bit-preserve/tree/main/Apple (
Local Archive, 2024-01-20)
Found a problem. Fixed and just need to upload.
Files fixed. :-)
Maybe I'm going blind, or did you maybe forget to post the link to where you uploaded it? :)
Well, I guess I found you on github...
https://github.com/reifsnyderb/Apple-II-R0
Having the schematics available in KiCad is going to be incredibly useful. I've wondered about this many times, when I was working with the old schematic PDFs - why we don't have a proper electronic variant, which would help a lot with searching/analyzing.
I recently found "baldengineer's" project, and was really amazed by these efforts to cover all the boards.
Maybe consider contacting him directly - and offering him your R0 version. He would certainly credit you for the work. It would be really nice to have all these schematics in one place, rather than scattered across different projects.
But in any case, nice to see KiCad projects for all the different Apple boards!
Good grief. Yeah, I forgot to post the link. Thanks for posting it.
I used the "baldengineer's" Apple II Plus as a starting point and have been in contact with him since I found the first issue. (A transistor was incorrect.) I also let him know about the Apple II R0 files I posted shortly after I posted them. (I figured it would be only fair since I used his posted files as a starting point.) I understand he got the files off of somebody else, too. Either way, I was really glad to have them available as it saved me a lot of time entering the schematics into KiCad.
The KiCad files do need a little clean-up as the component spacing and positioning could be a little nicer. Either way, having good, verified, schematics is a good starting point.
Great! Thanks for all your efforts!
My suggestion would be to also add a PDF print of your KiCad schematics to your github. This may seem useless, since they are just as static as the original Apple schematic PDFs. But it helps when someone finds your project, since the PDFs can be viewed instantly with the browser. Allows people to assess what your KiCad project really is about and what the quality is. Then, if they find it useful, they will invest the effort to clone the git project, so they can view them with KiCad...
PDF uploaded! :-)
Updated schematics have been uploaded. I have been cleaning the schematics up and making them neater, more organized, and easier to read.
To also help others find your repo, consider uploading a Readme.md that explains what the files are all about. You could then also reference both baldengineer as well as this thread if you like.
Very cool work and thanks for uploading/sharing it.
That's a good idea. Other than downloading, I've never used github before. I am slowly cleaning up the schematic and commenting it as well. I'll add a readme.md soon.
Thanks!
Brian
Very nice. Thanks for doing this.
AppleII_KiCad1.png
AppleII_KiCad2.png
AppleII_KiCad3.png
More updates:
https://github.com/reifsnyderb/Apple-II-R0
I suggest designating the capacitors with C1..Cn instead of Cap1..CapN. Using a non-standard designation causes confusion when searching the PDF for a particular capacitor and will surely cause confusion if it’s sent to a PCB manufacturer like JLCPCB to fully populate and wave-solder it.
I can see the reason, though, why the caps had to be renamed to non-standard names in the KiCad schematics: unfortunately the component identifiers in the original Apple schematics are not unique. Apple used C1..C99 for capacitors, as usual. However, for all ICs they used the grid location as the identifier. Hence, for example, C1 is either a cap - but it is also refers to the 74LS153 (in position C1). C2 is a 74LS195, etc.
That wasn't an issue with the original schematics. You would rarely confuse a ceramic cap C12 (100nF) with the IC C12 (SN74LS257). But when you make an electronic (KiCad) version, the tool requires the component identifiers to be unique.
So, you either have to change the identifiers of the ICs - or those of the caps. And sticking to the original identifiers for the ICs indeed seems a little more important for daily use... :)
KiCad lets you put as many fields as you like on the solder mask. You can have a field for the ICs called "Grid Location" for A1, B2, C3, etc. and use U1..Un in the Reference field of the ICs. The Reference field is the only one that needs to be unique. You can also hide the Reference field from the solder mask or the schematics and show some other field that doesn't need to be unique.
If I am to use JLCPCB to make this motherboard for me, I would have them wave-solder DIP sockets for the ICs, since it will take me a couple of years to find all the ICs. All the sockets having identifiers starting with different letters of the alphabet will be incredibly confusing.
My thought is that existing literature appears to refer to components by their grid location. So referring to A8, for example, as U28 would cause some serious confusion. Since I haven't seen any capacitors listed with a designation, the capacitors could be referred to my any number. I am thinking of referring to the capacitors as CP1, CP2, etc.
Take a look at the official Apple II+ schematics as an example. There they are designated C1..Cn:
Apple II Schematics.gif
You can label the ICs any way you like in the schematics or the PBC layout. For example "74LS251-H14" like Apple has done above. However the Reference field in KiCad is special and it is highly integrated with the tools of many PCB manufacturers, which if why I was suggesting to use the standard nomenclature in that particular field.
Ok. I was looking at different schematics.
I see your point. I'll get these schematics fixed and cleaned-up first, then worry about the designations.
Schematic is almost finished. I think it's good, electrically speaking. There just needs to be a little bit more clean-up and organization.
https://github.com/reifsnyderb/Apple-II-R0
Could you please update the PDF as well? Thanks!
PDF updated.
More updates have been made and the PCB has been added.
AppleII-populated.png
I love the dedication of your work. And the 3-d rendering is very pretty
But Rev-0 replica motherboards are everywhere. They're even available to buy right now on ebay.
What I would love (and I lack the time and ultimate Kicad skill to do it) is a Rev-7 motherboard design. Rev-7s (and RFI revisions) are way more useful than Rev-0, practically speaking and it would be a bonus if that Rev-7 (or RFI revision) could also be modified to run with 2716 EPROMS instead of having to find old (and incresingly scarce) genuine Apple 9116 PROMS.
Hello!
The KiCad Rev-0 replica is really the first part of a larger project. By using the Rev-0 replica, it is possible to verify the known schematics against a known working set of gerber files. I figured this is a good idea because sometimes the schematics have errors. (Oddly enough, an error was found on the gerber files....and it has been fixed.) Fortunately, I haven't been alone in this endevour as another github member has been working on the files as well...and has a similar long-term goal.
Once the Rev-0 replica is finished, a Rev-7 would be a lot easier to do. (I am not sure if my Apple II Plus is a Rev 7, it might be.) But that isn't really my ultimate goal. My thought is that it would be nice to build a complete and somewhat useable computer using modern off-the-shelf components. The computer shouldn't have any FPGAs, and should be able to be easily modified. Since an Apple I isn't realistically that useful, I figured an Apple II would be a better target because it can be built without any special LSI chips. (i.e. Atari's POKEY) An Apple II also has a lot of software available and was upgradable enough to last into the early 1990's in one form or another.
I am aware of the modern Z-80 and Commander X-16, I don't think they are as easy to modify as, say, an Apple II was.
While I decided to use some PLDs, here's the current schematic of what I believe would be an Apple II compatible computer made with components that are still in production:
https://www.applefritter.com/files/2024/02/12/AppleII.pdf
I also have the PLD pseudo code listed as well. This is so I can better plan out the PLDs.
I get what you're doing.
My comment come from the fact that it's been done already and Rev-0 boards are out there in the ether - Mike Willegal's done it, there was a kid on Facebook that did it, there's a thread going on right here at Applefritter with a guy building a Rev0 replica and several chinese and russian versions of the board have surfaced...and there are boards for sale right now.
I respect your exercise to vet the schematics, but in order to truly vet them you'll have to build it and run it, right?
But no one has done a Rev7 / RFI board, (well that's not entirely true - Henry Courbis at ReActive Micro has done it but it's only available either complete, or as a kit) and no one has done a board compatible with 2716 chips (which would require a board modification for the inverted chip-enable pins) and put it into the public domain.
Incidentally, since you mentioned readily off-the-shelf components, what are you doing about the NE558 timer chip which is long obsolete?
And the 9116 PROMs?
Another question - yor schematics include an At22V10C PAL, used as an address decoder, but it doesn't appear on the motherboard rendering.
The motherboard rendering is of the Rev-0 and woudn't have an atf22v10.
On the new design, not directly related to the Rev-0, the NE558 is replaced by 2 556 timer chips. The PROMs are replaced by a single SST39SF010. (Yes, a lot of space is wasted, but the SST39SF010 is less than $3 each.) The RAM is replaced by 48k of an AS6C1008. I think I've got the SRAM situation, and it's timing, figured out. If need be, there will easily be space on the board to make modifications as necessary. Also, all chips with low-level logic gates (i.e. AND, OR, NOT) have been replaced by PLDs. Overall, there is about a 50% chip reduction.
OK, I think I understnd what you're doing.
Still would love to see a REV-7 though.