A2Pi

[Charlie]

Found this old IDE project, if anyone's interested: http://www.qsl.net/zl1wjq/ap2ide1.htm

Thanks for that... could be a future project...

Mockingboad Update:
It would seem there are errors in the schematic... Bah! I can follow instructions, that doesn't mean I have much clue what I'm doing.

Top¬right is what I presume is the logic to choose between 6522s and therefore sound chips.
According to the schematic both 6522s have pin 25 (PHI2) labeled as /01¬BUFF but the selection logic has /01¬BUFF and 01¬BUFF running to a 7405N...
...ok, so I presume one of the 6522s is mislabled, just pick one to be /01¬BUFF and the other 01¬BUFF and disconnect the currently connected pin 25s on the 6522s. (oh, and the same for said lines feeding the two AY38193s) Assuming I'm right I guess it shouldn't mattter in the grand scheme of things. I'll just have to check channels are the right way round when wiring up the output.

Feel free to correct me if I'm wrong!
​

Of more concern is two other lines feeding the 7405N: /01 and 01... These lables appear nowhere else on the schematic as far as I can see. That rather nicely stuffs my progress.

Where should /01 and 01 be comming from? I don't know.

How many other mistakes are there that I haven't noticed?
​

I don't suppose there's a kind/knowledgeable person out there who could lend a helping hand? The schematic is on page 11 of the .pdf I posted further up this thread.

 

[Charlie]

Apple2Pi V1

Apple2Pi V1

Back on topic:

View attachment Apple2PiV1.zip
​

It seems I can't leave Eagle alone! Here's my interpretation of the Apple2Pi board. I feel rather neater than other versions, it should result in a cheaper and more robust solution with better access to the Pi's ports. It should also be compatible with all versions of the Pi with GPIO*

*Direct plugin with securing holes for 40pin GPIO devices
*26pin GPIO devices just need a shorter header soldering in place, no securing holes!
*Apple2Pi doesn't work with the Pi3

Thoughts?
Worth a production run..?

PS
I'll hopefully have a V2 out in a bit. The V1 should work just fine with a Pi Zero but I can see issues fitting the Apple's lid with a larger Pi.

- - - Updated - - -

Where were we? Oh, yes...
This one should work, and larger Pi's should still fit inside an Apple's case! oh:

View attachment Apple2PiV2.zip
​

 

[iainjh]

Bas thecorfiot is soldering me one up from that prototype board so I may need passs... depending on possible cost as always worth encouraging don't you think ?

btw the pi zero is so cheap for simplicity you could always standardise on that one?. It's now got wifi.

 

[Charlie]

Bas thecorfiot is soldering me one up from that prototype board so I may need passs... depending on possible cost as always worth encouraging don't you think ?

btw the pi zero is so cheap for simplicity you could always standardise on that one?. It's now got wifi.

Thanks for the encouragement.
As it happens my latest um, design should fit any Pi with 40pin GPIO... either soldered directly to it or with a female header plus optional securing screws. If a 26pin header is used it looks like there will be clearance for old Pi's too.
I didn't realise the new Zero has WiFi. That would indeed be a perfect fit for this project... I went with a Pi 2 plus WiFi dongle because if I can ever get on speaking terms with Linux I may bung a bunch more emulators on there. Hmm, what system will my Apple be today?
Of course if this works my Mockingchick idea may become less of a pipedream... It seems to me that I could potentially be doing everyone a favour and start banging stuff out for a fraction of the price I've seen from 'proper' retro Apple suppliers.


A bit of googling found me this PCB fab company:
https://www.seeedstudio.com/

I'm a complete NOOB when it comes to ordering 'proper' PCBs as opposed to etching my own at home so I expect to be told I've done something wrong:

I just got a quote for 20 of these boards deliverd in 5 working days @ 15.51 GBP all in!
​

Surely that can't be right..?
If that's true I'll never bother to home etch again! The main reason I don't do more stuff is I don't have the time... If this is all it takes to get a far better job than than I could ever produce with far less effort at this kind of price I'll be very happy.

C'mon, what's the catch?

Of course I have yet to work out how to be sure I have a valid design, and convert to gerber... Is that tricky..? Bas..?

 

[Charlie]

Ok, The attached .zip contains .sch .brd and gerber files for this board. Frankly the first time I've done this. I'm very tempted to put an order through to the fab company I linked to in my previous post but am nervous I've done something wrong... would be annoyng if I get back a bunch of useless lumps of PCB.

View attachment gerber a2pi spark.zip

If anyone feels like having a quick skim I'd be grateful.
​

 

[iainjh]

hi charlie

ive only got one slot left!

sure you dont want to try a mockingboard and A2Pi on one? that cost is so cheap, if correct the boards are far cheaper than slots! I do like to ask for the moon on a stick you know.. so do tell me to shove off as required

 

[Charlie]

You know mechanically I think that could work... imagine the a2pi board with the Mockingchick grafted above it. The Pi would then plug in to the back of that board, with it's I/O poking out round the edges. Neat, only one slot used up.

Thinking out loud:
While I expect the a2pi board will work (assumimg I did the gerbers right) as it's basically the same as my handwired version that I know works, the Mockingchick board I did is to some extent my own untested design. No idea if it will fly, my hope is Bas may be able to render a verdict. I would go ahead and handwire a prototype to test but I see the nice chap from eBay I got my protoboards from is out of stock.
If both boards use largely different control lines I can see mashing them together would be easy enough but if a lot is shared that gets me in to bus arbitration logic... likely a bit beyond me. (currently)

I shall think on it.


PS
I might look in to using a QFP 6522 for the Mockingchick if it proves a goer, and I think Yamaha still make a AY-3-8913 compatible IC in a very small package. (Yamaha YMZ284) The pro's being size and availability, the con being the Yamaha version has all it's outputs tied together... either end up with mono output only or stereo split across two ICs. The former would be a shame but would actually reduce the number of passives further, the latter would give stereo more like the Mockingboard 1a*

*Mockingboard 1a:
Two stereo channels, each comming from one AY-3-8913 with three(ish) voices.
*Mockingchick:
Left, right and center channels mixed from both AY-3-8913s to the stereo jack.

 

[Charlie]

a2pi mockup

a2pi mockup

Getting there?

Below is a scale mockup of my interpretation of the a2pi board 'plugged in' to a Pi 2. This should work with any Pi (except Pi 3) while leaving more than enough room for the ports. A Pi Zero is about the same size as this board, so in effect you'll have the ports peeping round the sides. Older 26pin GPIO Pis will also work, just use a shorter female header on the back.

View attachment gerber a2pi spark.zip
​

The .zip file contains .sch .brd and full set of gerber files for anyone interested. Feel free to comment. I'm going to send off an order later this evening so fingers crossed I get some functional boards back.

PS
Courtosy of Stedy kindly fixing my Mockingchick layout (thanks again) I've also learned how to do some nice ground planes which I guess is all to the good.

 

[Charlie]

Well, the order has gone to seedstudio... Here's hoping I haven't made any (silly) mistakes. I went for the cheaper, and slower, postage out of cheapness so let's see how long it takes my boards to arrive. I also went for ENIG rather than HASL for the finish - more than doubled the price but as these boards have an edge connector I hope money well spent for extra robustness and fewer issues with connectivity in the long run.

Fingers crossed.

 

[iainjh]

awesome work charlie! good luck!

 

[Charlie]

awesome work charlie! good luck!

Most kind, thank you. Here's a little something I got in the post today:

I'm now a 'fan' of Seed's 'fab' work
(ahem)
​

I must say seed did a lovely job, I shall certianly be using them again. Rather excited at this point, especially having never had some proper PCBs made before, so I just had to whip my iron out and give it a waggle. I did this one as a universal version with a 26pin header socket that should allow any Pi to fit.

Soldered up and ready for stuffing
​

Here's a couple of tasty treats to drool over... getting really rather excited indeed. Ooo, which delicious dainty shall I sample first? Nice to see the mounting holes line up perfectly so if one is plugging in to the Pi's GPIO rather than soldering direct it will still be nice 'n secure.

Pi Zero or Pi Fat, the choice is yours


​

At which point I discovered I'd accidentally overwritten my A2Pi boot disk, Doh! I don't have a serial card, so one VERY slow session with ADTPro later and...


...It didn't work.

Blast!

​

Oh, well... On to the bit I really detest when it comes to making electronic thingamys; tracking down errors. I won't bore too much with the miserable process, just my cry of anguish:

Great Steaming Elephant Turds!!!
oh: I'd gone and wired the data bus backwards oh:
(Chewbacca)


​

I believe there is a suitable quote for a time like this: "The path of my life is strewn with cowpats from the devil's own satanic herd." So there you go, ten not very useful coasters...
Actually not as bad as all that, quite salvageable actually. They just won't be as pretty as I hoped once I've cut eight traces and added a little Kynar. I have now corrected my mistake in the schematic and redone the layout for the board which I'll post shortly and I may well order some more boards... if anyone would like a 'prototype' feel free to ask. (I can be amazingly slow to answer PMs)

 

[iainjh]

Only a small error mate !

 

[Charlie]

Hello again,
Having fixed the reversed bus on one of my prototype boards, and discovered I'd also forgotten the R/W line for the 6551 *SIGH* I thought I'd give my mounting solution a go. See below:

Fixed and mounted?
​

Hoppeful that all should now be sorted I gave it another go... But no, just as before: Both Pi and Apple working fine but no communication between the two. Grrr.
So I pulled tthe ICs, put them back in my hand wired board, attached the Pi, and bunded the lot back in to the same slot in my Apple. Yep, still working like a champ. So there's still at least one more error on my board. The thing is I've buzzed it out, looked over my v1 schematic and can't see anything else wrong...
I've now done a v3 schematic and layout which I'd swear is fine but as I still haven't tracked down the remaining problem(s) with the v1 layout I'm feeling rather paranoid.

The attachment below is a .zip file containing my Eagle v3 files along with a .pdf of the original schematic my known working hand wired board was based off.

View attachment a2piv3.zip

Is there a kind soul out there who wwouldn't mind taking a look to see if v3 is indeed ok?
Many thanks.
​

 

[Charlie]

For any who may be interested I have sent off an order for my latest iteration the Apple2Pi PCB... Fingers crossed I got it right this time!

Basically as per the .zip file above, with the odd tweak, also:
-Updated silkscreen to ease any confusion on component position
-Added Apple2Pi text to the back
-HASL this time for comparison, and much cheaper!
-Added my mouse logo as a half-tone image*

*I generally do with my home etched PCBs. Also a bit of an experiment to gauge the DPI used by seedstudio.

 

[Charlie]

New PCBs back and they do look nice, well done Seeed... piccies in a bit.

Sadly it seems I killed my only 6551 so no testing and no Apple II fun until I get some more. I do hope my new boards work, buzzing it out all seems fine but I still haven't worked out why my Mk1 board (with fixes) won't work.

Apple2Pi v3
​

 

[Charlie]

Just in case anyone's interested:

View attachment Apple2PiV4.zip​

Hopefully final version on my interpretation of the Apple2Pi interface board. Not sure why but adding ground planes was messing up the signals so I decided to do without... my hand-wired board works fine without so I just deleted the offending bits.

 

[iainjh]

Goood luck chariie. Sorry for not checking in earlier. 3rd time the charm etc?

 

[Charlie]

Hello,
I don't know if the v3 works or not, waiting on ICs to arrive, but as I couldn't see a reason for the first one not to once I'd made some fixes paranoia has set in and so the v4 is a back to basics version just in case.

 

[Charlie]

Huzzah!

Well, sort of. I actually got the darn thing to work with my v3 boards so I know all is fundamentally good, including my presumed dead 6551...

The catch is it only boots right one in 5 or so, and only if moved to slot 1. I think I'm going to have to eat my words about the Apple being so slow one doesn't have to worry about cross talk...
...in an effort to keep the board as small and cheap as possible the GPIO pins run very close to any open slot lower than the one the board is plugged in to. I guess moving it to slot 1 reduces that interference a bit. I therefore assume the remaining issue with the board only starting occasionally (again done to reduce size) is that the 6511s IO lines cross the GPIO causing another source if interference. I thought I'd been really cleaver assuming the bus was too slow for cross talk to be an issue.

I'll test my theory by gluing some insulated foil to the back of the GPIO header (and same for the Pi). If that fixes the slot issue I can always hope it also fixes the remaining stability problem. If not I'll have to do a board more like the original but as that's a much bigger beast it will cost more and it does offend the eye to see so much PCB space unused.

Back in a mo'

 

[iainjh]

good luck! unfortunately some boards require slot 0, mine only has slot 7 free i think. what a pain. Darn 1mhz electronics eh