Pi Pico[W] Peripheral Expansion Box Side Port Device

[visrealm]

1 minute ago, JasonACT said:

That's turned out very nicely, I'm really not a fan of the beige filament I have.

 

I'm still holding out for a 2 tone case, I live in hope

 

To be honest, my first TI was beige, which the power supply went pop within 12 months..  And TI replaced it with a black/silver model, which I used for years and years.  But all my units are now beige, and I'm not unhappy with that.  I don't feel I ever got to really know the beige model.

I've got one of each. I had a black/silver one as a kid, so it has the nostalgia value, so when I decided to buy one this year and get back into them, I had to get the shiny boy. It just sits on a shelf looking very pretty now since not long after, I picked up a beige one too which has become my testbed. The beige one had a dud mylar keyboard, so that prompted the KEYBOARD/4A-99 project. It's a PAL unit, so has the crappy RF box, so that prompted the PICO9918. If it wasn't for beige, we probably wouldn't have those projects

 

Mr fancy-pants on the shelf over there is an NTSC unit, so has decent composite output. It has a mechanical keyboard, etc. I wouldn't have been fed up enough to work on it.

[JasonACT]

16 minutes ago, visrealm said:

It's a PAL unit, so has the crappy RF box, so that prompted the PICO9918. If it wasn't for beige, we probably wouldn't have those projects

Cheers to the beige units  

[JasonACT]

On 11/19/2024 at 10:28 PM, visrealm said:

is an NTSC unit, so has decent composite output. It has a mechanical keyboard, etc. I wouldn't have been fed up enough to work on it

When I joined AtariAge, I was told NTSC stood for "Never The Same Colours"  

 

I do like my beige NTSC console though, it shall remain all original.

 

I'm currently still working on the TIBUG Debugger, and as I test it, I recall many more things I had changed "back then"..  I'm sure they ripped out quite a lot of code (from a far better product) and released a minimum working version.  I've added back the "D" disassemble current / address + range command now, and reformatted the workspace dump output - 4 Regs per line instead of 3.  Fixed the user-screen save/restore for everything except single instruction execution, which is better served by seeing each line executed without a screen restore + save on return.  Also, 32 cols & 40 cols cartridge menu options, to initialise the expected setting for debug (this is where you can see it's a cut down version of something far better, what we got is locked into 32 columns mode, but the code almost works as-is in both that and 40 cols text mode).

 

I've also added <CTRL><ALT><PAUSE> (on a USB keyboard) as a LOAD interrupt button.  This is configurable by "KBLOAD=1" in the config file, since it'll crash a console without something like a debugger properly loaded with the LOAD vector set.  I.E. Enable it, if you know what you're doing.  I'm yet to test the timing on this change though, I expect to lose 15ns or so, since every memory access takes this path.

 

Currently. it still all fits in the 8KB Cart RAM, but oh-boy I'm getting close to running out of that RAM with all these fixes.

[+chue]

Just another update on my console testing.  Previously I wrote that one of my non-working TIs actually works.  I now have tested the other 2 non-working TIs.

 

The first shows bad SRAM, which should be easily fixable:

 

 

The second unfortunately does not show any sign of life.  I powered it up using a known good power supply with correct voltages (+5, +12, -5).  Further investigation of the motherboard shows a short between the 12V line and ground (2.8 Ohms).  Shorts are tricky to diagnose, at least for me, so I will probably just use this board for parts.

[JasonACT]

6 hours ago, chue said:

The first shows bad SRAM, which should be easily fixable:

All the bit errors are in the odd (low) byte too, U609.  Nice

 

6 hours ago, chue said:

Shorts are tricky to diagnose, at least for me

Me too.  Looks like either VRAM/4116, 9904 or 9900 could be the issue, or a dead capacitor somewhere maybe.  But I guess you're going to need some spare parts now anyway  

[JasonACT]

On 11/22/2024 at 12:05 AM, JasonACT said:

I'm yet to test the timing on this change though, I expect to lose 15ns or so, since every memory access takes this path.

After testing a bunch of different code changes, I've come to a compromise, but the timing is close to the margin.  That's at 250MHz, I guess it can be made better with faster speeds if needed.  It also turns out that the 'I' command (Inspect screen location) can switch to 40 cols mode, but there's still one more message about disk errors which is unreferenced, pointing to a more complete product.  Anyway, I've made quite a few more changes, plus cleaned up the font data which has gained a bit more space in the 8KB Cart RAM.

 

The <CTRL><ALT><PAUSE> key strokes are not always processed as a LOAD interrupt either, I'm pretty sure the hit & miss nature of this is because in a tight keyboard reading loop, I don't send the signal if the keyboard CRUs are currently being read.  After a few keypresses (if needed) it does work, and it doesn't have any "bounce".  It's not ideal, but it is what it is.

 

Debugger ROMs are now part of the main firmware download.  Debugger source code changes included here also.

DebuggerC.zip

PPEB2.zip

Edited November 24 by JasonACT
EDIT: Fixed Console Tester Mode

[+dhe]

I've always puzzled over the history of TI's debugger. Through out the documentation there are hints at other things. After TI pulled out of the home computer market, an updated debugger was released to TI recognized user groups. The docs for that debugger, oddly had Navarone all over it. Then, later, Edgar D(sp?) released Super Debugger for the SuperCart. Thank you for this version, I look forward to working with it.

[JasonACT]

On 11/23/2024 at 11:15 PM, dhe said:

I've always puzzled over the history of TI's debugger. Through out the documentation there are hints at other things. After TI pulled out of the home computer market, an updated debugger was released to TI recognized user groups. The docs for that debugger, oddly had Navarone all over it. Then, later, Edgar D(sp?) released Super Debugger for the SuperCart. Thank you for this version, I look forward to working with it.

Do we still have these other debuggers?  And don't forget DiskAssembler, I had access to that way back when, but found it was so large you couldn't use it to debug a lot of software (after I removed the protection, using my modified version of TIBUG back then).

 

Anyway, I've fixed the <Ctrl><Alt><Pause> issue (it was skipping due to the <Ctrl> & <Alt> keys being held down in a tight keyboard reading loop) along with stepping through console ROM for the keyboard scan (keyboard CRU reads would trigger a 2nd LOAD interrupt, which the PPEB was doing to try and fix the read value, which I now skip so it doesn't see USB/Bluetooth keyboard keypresses when debug-stepping now).

 

I'm still looking for 10-20ns of improved timing though, and GCC isn't playing nice.  Might be time to move over to the Pico2W after this.

[+Torrax]

On 11/23/2024 at 5:15 AM, dhe said:

I've always puzzled over the history of TI's debugger. Through out the documentation there are hints at other things. After TI pulled out of the home computer market, an updated debugger was released to TI recognized user groups. The docs for that debugger, oddly had Navarone all over it. Then, later, Edgar D(sp?) released Super Debugger for the SuperCart. Thank you for this version, I look forward to working with it.

 

Here is a 7zip file of disks that I transfered to my PC back in 1997.  Super Bug (v1.0??) and Hot Bug are in the mix.

The files are in Archiver 3.03 - TI-Files format.

 

TI-2-PC.7z

[JasonACT]

On 11/26/2024 at 10:32 PM, JasonACT said:

I'm still looking for 10-20ns of improved timing though, and GCC isn't playing nice.  Might be time to move over to the Pico2W after this.

I've found it plus some, by pausing the build so I can hand-edit the generated asm..  Actually I got 30ns back by doing this, but it takes some time for rebuilds now, so I eagerly await my Pico2Ws to see how much better they are.  I think they will be very good, but I'll hand-edit the code again if needed for the original Pico's for any future changes.

 

I'm actually pretty happy with the timing on this firmware, I hope I've not stuffed anything up  

PPEB2.zip

[+chue]

18 hours ago, JasonACT said:

I'm actually pretty happy with the timing on this firmware, I hope I've not stuffed anything up

SAMs burn-in looks good, so you didn't stuff that up.

[JasonACT]

2 hours ago, chue said:

SAMs burn-in looks good, so you didn't stuff that up.

Thanks for testing!

[+Vorticon]

Finally had a chance to test out the PPEB I got from @RickyDean. I'm really mostly interested in the pcode functionality and the incorporated SAMS, and in that respect it works perfectly. My only wish would be to have a way to switch disks from the configuration screen like the TIPI so as not to have to remove the SD card each time and edit the configuration file manually.

I cannot overstate the importance of that device for those looking to use the UCSD Pascal system on the TI given how rare the pcode cards are. I really hope someone will consider offering it for sale to the TI community at large, perhaps through @arcadeshopper or @Shift838...

[JasonACT]

3 hours ago, Vorticon said:

My only wish would be to have a way to switch disks from the configuration screen like the TIPI so as not to have to remove the SD card each time and edit the configuration file manually.

While in the "CALL TIPI" config screen, pressing the SPACE-BAR takes you to the DSK/CS selection screen, for selecting a disk-image/directory from the directory currently selected from the main menu's current directory/location.  PCODE is a bit funky though, making it difficult to get into TI BASIC, and the config screen to begin with.  Not much I can do about that though.

 

Thanks to @visrealm for supplying me with the device, I'm planning to enhance the whole config screen, to be more like the FinalGROM99, but I'm not yet sure how it'll all work - but I'd like to present menu names on the right of the screen along with the filenames on the left.  Or something like that, probably when "PHM" is part of the filename, since those are the files which the FinalGROM99 make easier to use.

 

@jedimatt42 Let me know if you still want me to change the main DSR from CALL TIPI to CALL PPEB...  I will probably add "CALL PPEB" anyway, with TIPI being a legacy CALL.  But if you are totally sure you want it removed, now is the time to say so, as I think the RP2040 chip is doing about as much as it can and there won't be many more updates on that device.  I'm going to focus on the RP2350 going forward, not that I have any fixed ideas on what comes next, did I miss anything?

[+Vorticon]

6 hours ago, JasonACT said:

While in the "CALL TIPI" config screen, pressing the SPACE-BAR takes you to the DSK/CS selection screen, for selecting a disk-image/directory from the directory currently selected from the main menu's current directory/location.  PCODE is a bit funky though, making it difficult to get into TI BASIC, and the config screen to begin with.  Not much I can do about that though.

Ah got it. Wasn't evident on the config screen. Thanks. By the way, pressing H on the main pcode screen will halt the pcode card and reset the computer to the start screen at which point one can access the config screen. 

[+arcadeshopper]

14 hours ago, JasonACT said:

While in the "CALL TIPI" config screen, pressing the SPACE-BAR takes you to the DSK/CS selection screen, for selecting a disk-image/directory from the directory currently selected from the main menu's current directory/location.  PCODE is a bit funky though, making it difficult to get into TI BASIC, and the config screen to begin with.  Not much I can do about that though.

 

Thanks to @visrealm for supplying me with the device, I'm planning to enhance the whole config screen, to be more like the FinalGROM99, but I'm not yet sure how it'll all work - but I'd like to present menu names on the right of the screen along with the filenames on the left.  Or something like that, probably when "PHM" is part of the filename, since those are the files which the FinalGROM99 make easier to use.

 

@jedimatt42 Let me know if you still want me to change the main DSR from CALL TIPI to CALL PPEB...  I will probably add "CALL PPEB" anyway, with TIPI being a legacy CALL.  But if you are totally sure you want it removed, now is the time to say so, as I think the RP2040 chip is doing about as much as it can and there won't be many more updates on that device.  I'm going to focus on the RP2350 going forward, not that I have any fixed ideas on what comes next, did I miss anything?

Please remove call TIPI as it's not a TIPI   and that way it will coexist with TIPI if someone chooses.. 

 

 

[JasonACT]

1 hour ago, arcadeshopper said:

Please remove call TIPI as it's not a TIPI   and that way it will coexist with TIPI if someone chooses.. 

I'm pretty sure it won't, so that won't be the reason I remove it.

[JasonACT]

17 hours ago, JasonACT said:

I'm going to focus on the RP2350 going forward

Yeah, totally broken.  It arrived yesterday, I've spent 24 hours (minus sleep) getting it to build (Arduino Pico) and I can see almost nothing works...

No LED.  No Speech.  No Internet Time.  No Arduino EEPROM.  No Carts, and no Expansion RAM (the real XB Cart confirmed that).

 

The DEBUGLEVEL output is working, and it is reading ROM files off the SD Card (albeit slowly).  That's where I can also see no EEPROM or Internet, with output suddenly stopping when it should print the current time as a debug message.  Might be time to wait for Arduino Pico to catch up, I didn't expect this many issues.

[+dhe]

Here is how I spent Thanksgiving Day.

 

[+Vorticon]

2 hours ago, dhe said:

Here is how I spent Thanksgiving Day.

 

Nice case! Do you happen to have the STL files for it?

[+dhe]

A big thanks to @RickyDean for giving me a sacrificial board!

 

@Vorticon if Jason doesn't reply before me, I will bring home the files next week.

[RickyDean]

55 minutes ago, Vorticon said:

Nice case! Do you happen to have the STL files for it?

You can use this design on yours, but it is set up for risers on the pico which yours doesn't have, you will have to modify the access hole for the USB to sit at zero mm and not 12mm higher per @JasonACT. The risers were my idea and Jason ran with it. You can remove the pico and add risers, but it would be a bit of a job.

[+Vorticon]

1 hour ago, RickyDean said:

You can use this design on yours, but it is set up for risers on the pico which yours doesn't have, you will have to modify the access hole for the USB to sit at zero mm and not 12mm higher per @JasonACT. The risers were my idea and Jason ran with it. You can remove the pico and add risers, but it would be a bit of a job.

I guess I can design my own case. I'm pretty proficient with OpenScad

[JasonACT]

46 minutes ago, Vorticon said:

I guess I can design my own case. I'm pretty proficient with OpenScad

Or you can use my OpenScad file, and change the window height to 0:

 

On 4/2/2024 at 5:17 PM, JasonACT said:

I can't add any more to those suggestions, thanks everyone, so instead here's a small mod I made to the case:

 

 

An option to shape the top-back to be angled the same as a beige model.  (That's sparkle-gold top and copper bottom - two of the nicest PLA printer filaments I have, the photo doesn't do it justice.)

ZfinalB.zip 1.43 kB · 46 downloads

 

[+Vorticon]

2 hours ago, JasonACT said:

Or you can use my OpenScad file, and change the window height to 0:

Even better! Glad to see a fellow OpenScad user