A8PicoCart - UnoCart on a Raspberry Pi Pico

[dmsc]

Hi!

3 hours ago, The Mr. Video said:

I can't seem to get mine to work at all no matter how many times I reposition it. Meanwhile, I also tried loading a regular Atari cartridge board and got those to work without having to move them too much.

 

And an obvious question, which firmware did you flash in the PICO board?   Do you see the PICO virtual disk when plugging into a PC without the boot button pressed?

 

Have Fun!

[dmsc]

Hi!

3 hours ago, The Mr. Video said:

I can't seem to get mine to work at all no matter how many times I reposition it. Meanwhile, I also tried loading a regular Atari cartridge board and got those to work without having to move them too much.

Also, from your photo, yo have a Winbond flash device, that seems strange, I tough that all of this boards had ZBIT flash media, it is a lot cheaper than winbond. Perhaps your board does not have the full 16MB?

 

Have Fun!

 

[electrotrains]

26 minutes ago, dmsc said:

Hi!

Also, from your photo, yo have a Winbond flash device, that seems strange, I tough that all of this boards had ZBIT flash media, it is a lot cheaper than winbond. Perhaps your board does not have the full 16MB?

 

Have Fun!

 

 

I've had boards with both winbond and zbit - no obvious difference in price, and they are both 16meg.

 

The firmware was initially developed on a board with winbond flash - I had to make a few changes to get it to work with zbit ones (its rated slightly slower).

But its been fully tested with both types.

 

Robin

 

[The Mr. Video]

32 minutes ago, dmsc said:

Also, from your photo, yo have a Winbond flash device, that seems strange, I tough that all of this boards had ZBIT flash media, it is a lot cheaper than winbond. Perhaps your board does not have the full 16MB?

 

Windows says that the picocart has 14.9MB of space- one megabyte shy of the needed 16. However, when I hold down the BOOTSEL button, Windows says it has 127MB. I bought mine from AliExpress, but I think I had to find a different listing because the one I found on this thread was no longer available. Where did you buy yours?

Edited November 4, 2023 by The Mr. Video

[The Mr. Video]

41 minutes ago, dmsc said:

Hi!

 

And an obvious question, which firmware did you flash in the PICO board?   Do you see the PICO virtual disk when plugging into a PC without the boot button pressed?

 

Have Fun!

I flashed it with the one from the Github page. Are there multiple versions?

[electrotrains]

3 hours ago, The Mr. Video said:

I can't seem to get mine to work at all no matter how many times I reposition it. Meanwhile, I also tried loading a regular Atari cartridge board and got those to work without having to move them too much.

(1) Did the board mount as A8-PICOCART after you flashed the uf2 firmware file?

(2) Check continuity (if you have multimeter) between the pins on the pico and the corresponding pads on the cartridge edge connector.

(3) Check you've got no shorts between adjacent pins (again with a multimeter).

(4) Check for any blobs of solder on the PCB and the pico itself.

 

Failing that, you could set pico GPIO pin to alternate between high & low in turn (you'll need to write some code) and check you can see 3.3v and 0v on the corresponding pad. Or just start again with a new pico... 😞

 

But the only one I made that didn't work first time I'd forgotten to solder a pin on the pico. I've never had to re-insert or move the cartridge to get it to work. The edge connector is one I've used for previous cart designs and I think its pretty well lined up.

[electrotrains]

11 minutes ago, The Mr. Video said:

Windows says that the picocart has 14.9MB of space- one megabyte shy of the needed 16. However, when I hold down the BOOTSEL button, Windows says it has 127MB. I bought mine from AliExpress, but I think I had to find a different listing because the one I found on this thread was no longer available. Where did you buy yours?

That's all completely normal. With the A8PicoCart firmware - there's one meg reserved for the firmware itself, and the other 15 megs allocated for the filesystem (which is what windows can see).

Edited November 4, 2023 by electrotrains

[electrotrains]

Looking at your photos - it looks possibly like you may have bridged two of the pico pins adjacent to the the word "TOP" on the PCB. The soldering looks a bit uneven in general, I suspect the problem lies there.

 

EDIT - if you can try taking some more photos top and bottom with slightly sharper focus it might help diagnose.

 

Robin

Edited November 4, 2023 by electrotrains

[The Mr. Video]

1 hour ago, electrotrains said:

(1) Did the board mount as A8-PICOCART after you flashed the uf2 firmware file?

(2) Check continuity (if you have multimeter) between the pins on the pico and the corresponding pads on the cartridge edge connector.

(3) Check you've got no shorts between adjacent pins (again with a multimeter).

(1) Yes, it did. I even put some executable files into it.

(2)  Pin 23 on the pi didn't make contact with its corresponding pad, but reflowing solder and getting continuity didn't solve the problem.

(3) Checked the pins with a multimeter, none of them are bridged together.

 

Also, here are some better pictures of the board.

Edited November 4, 2023 by The Mr. Video

[Peri Noid]

[delete, please]

Edited November 4, 2023 by Peri Noid

[bfollowell]

4 hours ago, mozzwald said:

Your solder joints aren't the best looking, maybe some are not connected. Put flux on all of the pins and reflow the joints again. Also looks like there are solder blobs stuck to the board. Clean it off with alcohol and q-tip/toothbrush after soldering. You don't want metal bits falling into your Atari.

Far be it from me to critique another's soldering. Lord knows mine isn't always top notch the older I get, but I would tend to agree with @mozzwald. Many of those look like globs stuck to the pins and it's really hard to tell if you have good, or any, connection to the solder pads underneath. I would bet that your problem has more to do with a loose solder connection or two moving around as you insert/move the card. I would heat up the old soldering iron and, at the very least, reflow every one of those solder connections and cleanup any of the solder globs around the board. There were at least a few. See how it acts for you after that. Hopefully, it will improve.

 

One thing you did say made me wonder though. You said you were able to get a regular Atari cartridge board to work without moving it around too much. I would think they'd work with no fidgeting. That makes me wonder if you have a damaged or severely worn cartridge slot. Do you have any regular old 8-bit carts in original cases that you can try? If you do, and you have to fidget around with them to get them to work, I'm thinking the problem is your cartridge slot and not the carts themselves.

 

[The Mr. Video]

7 minutes ago, bfollowell said:

One thing you did say made me wonder though. You said you were able to get a regular Atari cartridge board to work without moving it around too much. I would think they'd work with no fidgeting. That makes me wonder if you have a damaged or severely worn cartridge slot. Do you have any regular old 8-bit carts in original cases that you can try? If you do, and you have to fidget around with them to get them to work, I'm thinking the problem is your cartridge slot and not the carts themselves.

 

Cartridges work fine on my 600xl. The one I used was just kind of finicky.

[The Mr. Video]

21 hours ago, electrotrains said:

Failing that, you could set pico GPIO pin to alternate between high & low in turn (you'll need to write some code) and check you can see 3.3v and 0v on the corresponding pad.

Seeing that I got an an extra hour from daylight savings time ending, I figure I'd give it a try. I haven't gone through all of the gpio pins yet, but I noticed that pin 4 wasn't going high. I'm hoping that it's a problem with the solder joint on the board and not the board itself.

[The Mr. Video]

It lives! Turns out taking your time with soldering goes a long way. Now I just need a case for it.

 

[bfollowell]

4 hours ago, The Mr. Video said:

It lives! Turns out taking your time with soldering goes a long way. Now I just need a case for it.

 

 

This! For certain! There's nothing more important when soldering than taking you're time, making certain you get good flow and good connection between the pin and the solder pad.

 

I'm glad you got yours going, great job!

 

[manterola]

On 11/4/2023 at 12:44 AM, manterola said:

The A8PicoCart worked in my Atari 800 (OS-A)!. After pressing System Reset key I get the pink/purple menu every single time. The reset button in the cart does not work, which is not important if you use the cart with the A800 door closed. I was able to run a .CAR file.

It is not stable at all, thought. It keeps crashing while browsing the file system, or using the search feature.

But if you get to the point of selecting a ROM, it works..

 

I think that my Atari 800 with OS-A is booting up "too fast" and maybe the RP2040 is not ready yet. After pressing reset if work (kind of work). I read that something similar was going on with 2600 version when used with some TIA versions.

 

[electrotrains]

19 minutes ago, manterola said:

I think that my Atari 800 with OS-A is booting up "too fast" and maybe the RP2040 is not ready yet. After pressing reset if work (kind of work). I read that something similar was going on with 2600 version when used with some TIA versions.

 

I think that's quite likely. I seem to remember something similar was reported with one of my earlier cart designs on the 400/800 - didn't work, but did if you press RESET on the atari.

 

The menu not being very stable is likely to be due to a different clock signal on the 800 cart port (RAS? instead of PHI2). I think some kind of mod is possible to subsitute the signals, but I wouldn't want anyone to have to modify their hardware.

 

Sadly, unless someone in the UK feels like lending me their 400/800 for a bit, so I can test and debug, I'm unlikely to be able to fix. But since its open source, maybe someone else with the ability will be able to take a look?

 

Robin

 

[manterola]

12 minutes ago, electrotrains said:

The menu not being very stable is likely to be due to a different clock signal on the 800 cart port (RAS? instead of PHI2). I think some kind of mod is possible to subsitute the signals, but I wouldn't want anyone to have to modify their hardware.

I already did the mod to substitute the signals a year or so ago. I think is something to do with the menu ROM itself, because once I select a game ROM I can play that game for hours. So the cartridge emulation works fantastically.

Thank you for making this open source and address my question. With my limited assembler skills I have been playing with the menu ROM source code. However, I have not put together the whole dev environment to compile a new uf2 file. 

Thanks again, I will report back any progress.

[+David_P]

58 minutes ago, manterola said:

I think that my Atari 800 with OS-A is booting up "too fast" and maybe the RP2040 is not ready yet. After pressing reset if work (kind of work). I read that something similar was going on with 2600 version when used with some TIA versions.

 

As I recall, there's a similar issue with FujiNet; the shorter boot time on the 400/80 means the device hasn't sufficiently initialized when the computer attempts to boot.  The solution tere is an external power supply - not really an option for a cartridge.

[manterola]

On 11/15/2023 at 10:29 AM, manterola said:

I already did the mod to substitute the signals a year or so ago. I think is something to do with the menu ROM itself, because once I select a game ROM I can play that game for hours. So the cartridge emulation works fantastically.

Thank you for making this open source and address my question. With my limited assembler skills I have been playing with the menu ROM source code. However, I have not put together the whole dev environment to compile a new uf2 file. 

Thanks again, I will report back any progress.

I was reading the Altirra hardware document, and maybe the frezzes of the Cartidge in the 800 is because of the floating data bus in the range $D500-$D5FF which is the range you use to talk with the RP2040. In theory, it shouldn't be floating since there is a cart inserted, but maybe since A8PicoCart is not your typical Atari Cart, that can affect the operation. Maybe the RP2050 GPIOs are still floating...? and the Atari is reading garbage command from $D500-$D5FF?

 

As a side note, I added a few features to the A8pico Menu rom, the only important one is about jumping through pages. Do you prefer a more formal github contribution or just the code in a PM? I added more hotkeys, but those are really just really suggestions and me playing around and trying stuff from the usability POV.

 

 

[electrotrains]

On 11/18/2023 at 2:14 AM, manterola said:

As a side note, I added a few features to the A8pico Menu rom, the only important one is about jumping through pages. Do you prefer a more formal github contribution or just the code in a PM? I added more hotkeys, but those are really just really suggestions and me playing around and trying stuff from the usability POV.

 

 

Sorry for the delayed reply.

Feel free to PM me the code/changes. Might be worth posting the uf2 file so others can have a play with the proposed firmware changes.

 

Robin

[+mytek]

On 11/17/2023 at 6:14 PM, manterola said:

As a side note, I added a few features to the A8pico Menu rom, the only important one is about jumping through pages.

Be nice to also use some of the PS/2 navigation keys (Home, End, Page-up, Page-Down, ect.) in TK-II for some of this as well.

 

Here's what I did for SDrive with a crude byte swap in the original program...

 

 

Notice also that the escape key is used differently, with 'F' being used for search instead, and escape sending you back to the parent directory.

 

Alternative TK-II sdrive.atr

 

Here are the Atari key code equivalents for those PS/2 navigation keys.

[manterola]

@mytek I will consider this but I guess it is electrotrains call. I got a question, though: for example, PageUp code is 67, so is that equivalent to do Ctrl-C in a regular Atari keyboard?

 

Now the version that I put together this past week ADDED the following hotkeys to the already existent ones.

-  '/' for starting a search.

-  Backspace for going up (parent directory) in the folder structure.

-  '+' which is left-arrow without pressing Crtl, also goes to parent directory

-  Left Arrow also goes to the parent directory

-  '>'  moves one page forward

-  '<' moves one page backward

I did not update the main screen regarding the hotkeys (the hotkeys reminder in the bottom of the screen), because I was just testing these hotkeys mentioned above, and also because those are all still applicable.

 

a8_pico_cart.uf2

[+mytek]

5 hours ago, manterola said:

PageUp code is 67, so is that equivalent to do Ctrl-C in a regular Atari keyboard?

I think CTRL+C = 146     EDIT: result of PEEK (764)

 

All the PS/2 navigation keys in TK-II render key codes equivalent to shifted 1200XL F1-F4 functions, with the exception of the arrow keys which default to the Atari CTRL+Arrow equivalents (Robin's version already supports the arrow keys in both a Controlled and non-Controlled fashion). Delete and Insert are given the normal Atari equivalents to these same functions.

 

BTW, if anyone has a U1MB with @flashjazzcat's BIOS installed and runs the Side Loader with a TK-II enabled keyboard, then you can see a great example of navigating the media on the card using all these custom keys. It really has a great flow to it.

[a8isa1]

On 10/8/2023 at 5:54 AM, Bratwurst said:

 

As a general advisement to those who might not know: Bevel your card edges for your PCB to reduce wear and tear on your cartridge port, especially considering this is something you'll be apt to pull and insert several times to load different software via the USB port. The ideal time to do this would be before you solder anything to the board but you can probably ground a steel handfile to mitigate ESD.

 

 

Once again guilty of not RTFM (this thread). At least not fully. Too late. PCBWay is already fabricating my order.