Modify QI console for finalgrom99 use

[dphirschler]

Not having any luck with this on my QI it my non-QI.  Fwiw, my QI has 1981 rom.

Edited May 7 by dphirschler
added pics

[dphirschler]

OK, so I put the cpld and avr firmwares on the device.  Then (on my non-QI TI), it booted it up and got a colorful menu.  I also got a Review Module Library option one time.  Then I got only TI Basic.

On my QI, I managed to get into FinalGROM99 one time and after the first menu, (shown in the previous post with a yellow block in it), it gave me a blank screen on the next menu.

I put the v1.3 firmware back on it and it's good again... at least on my non-QI.  My QI shows the blank screen with a couple glitchy lines on the right side.

[JJB]

I got it ...

 

I traced the problem to this piece of TI assembler. It is part of the browse, sort and menu code the AVR loads at CPU memory >6000:

 

wait:
       mov  *r7+, r2          ; image ready?
       jne  ready
       dec  r4
       jne  wait
       jmp  animate

 

It reads a set of ROM addresses from >6000 and checks if any data is there. While FG99 is loading that part of memory with either the next folder contents or an image, it takes the cartridge offline. Any read from that part of memory will now return >0000 on the classic TI as the GROM port data bus is pulled low during reads. 

 

However, on the QI the GROM port data bus doesn't have separate pull up/down circuitry and returns >FFFF 🙂

 

I simply changed the code to test for both >0000 and >FFFF and now both TI's! are running FG99 as it should.

 

See the attached final updates for the AVR.

 

Cheers,

 

Jochen

update.avr

[jrhodes]

So no need to update CPLD now?

I don't own a QI console yet, but this awesome progress!

[JJB]

1 hour ago, dphirschler said:

OK, so I put the cpld and avr firmwares on the device.  Then (on my non-QI TI), it booted it up and got a colorful menu.  I also got a Review Module Library option one time.  Then I got only TI Basic.

On my QI, I managed to get into FinalGROM99 one time and after the first menu, (shown in the previous post with a yellow block in it), it gave me a blank screen on the next menu.

I put the v1.3 firmware back on it and it's good again... at least on my non-QI.  My QI shows the blank screen with a couple glitchy lines on the right side.

That is really strange. Unfortunately I can't see your attached photo's? 

 

I can only suggest to install the latest AVR update (see previous post), install the same CPLD update, clean the GROM port / FG99 connectors, remove all peripherals and try again.

 

If you see glitches on your QI with the original firmware for both AVR and CPLD it is either something related to your specific QI or there may be a wider spread of timing differences between consoles than thought. If the latter that would be a bummer.

 

Would be great if we can get more members to test.

[JJB]

Just now, jrhodes said:

So no need to update CPLD now?

I don't own a QI console yet, but this awesome progress!

You will need to update both the CPLD and the AVR.

[jrhodes]

13 minutes ago, JJB said:

You will need to update both the CPLD and the AVR.

On 5/2/2022 at 9:48 PM, JJB said:

Anybody willing to test out the attached CPLD update file (JB_UPDATE.PLD)? It still has the annoying menu and folder screw-ups (see earlier in thread) but it should behave the same as the RC fix. I have attached the latest original 1.3 update as well. 

 

To install either the update or the original, rename file to UPDATE.PLD and save in root of the FG SD. Reset and the update installs automatically. Once the LED is off remove the update file from the SD.

 

BTW if you get an error trying to load an image from a folder, try power cycling that seems to help. 

 

Any observations you can share will help!

 

Thanks,

 

Jochen

JB_UPDATE.PLD 79.06 kB · 6 downloads ORG13_UPDATE.PLD 79.06 kB · 6 downloads

JB_UPDATE.PLD from this post is the correct one to use with this final update, right?

[JJB]

6 minutes ago, jrhodes said:

JB_UPDATE.PLD from this post is the correct one to use with this final update, right?

Yes that's the one. For convenience I have re-attached latest updates + originals to this post.

ORG13_UPDATE.PLD FG99ORG_update.avr JB_UPDATE.PLD JB_update.avr

[JJB]

just replying further to avoid possible update problems:

 

- make sure SD card is either FAT or FAT32 formatted

- save update CPLD file as UPDATE.PLD file in / of the SD card

- stick SD card in FG99 and press reload button

- LED lights up for a while

- after LED turns off remove SD card

- delete UPDATE.PLD file

- reset FG99

- save update AVR file as UPDATE.AVR file in / of the SD card

- stick SD card in FG99 and press reload button

- LED lights up shortly

- after LED turns off remove SD card

- delete UPDATE.AVR file

- stick SD card back in FG99

- reset FG99

 

... this should work. If you want to revert to standard firmware repeat the above with the original PLD and AVR files.

 

Ralph Benzinger has mentioned on his FG99 website that some SD cards don't seem to work properly; hence you might need to try a different one.

[+OLD CS1]

Success!  My FG99 now works properly with both my QI and non-QI consoles.  I am going to swap my QI console back in to be my main desk system and I will do more trialing over the next couple of days.

 

9 hours ago, JJB said:

Ralph Benzinger has mentioned on his FG99 website that some SD cards don't seem to work properly; hence you might need to try a different one.

I have an 8GB SanDisk Ultra which requires I reset the FG99 before it will load.  None of my 1GB or smaller SanDisk SDs, including microSD in adapters, have this problem.

[dphirschler]

Mines does not work on the non-QI.  Didn't try the QI.

[JJB]

14 minutes ago, dphirschler said:

Mines does not work on the non-QI.  Didn't try the QI.

Are you able to try different SD's at all? A friend of mine testing just had similar problems like you and another SD card did the trick. 

 

Also make sure to update CPLD and AVR separately (see earlier post for steps)

 

Hope it will work for you as well. 

[JJB]

26 minutes ago, OLD CS1 said:

Success!  My FG99 now works properly with both my QI and non-QI consoles.  I am going to swap my QI console back in to be my main desk system and I will do more trialing over the next couple of days.

 

I have an 8GB SanDisk Ultra which requires I reset the FG99 before it will load.  None of my 1GB or smaller SanDisk SDs, including microSD in adapters, have this problem.

Awesome! 

[+acadiel]

I’ll have to try this in my /4 as well.  As I recall, FG acted a bit strange in the /4.

 

[+dhe]

6 hours ago, acadiel said:

I’ll have to try this in my /4 as well.  As I recall, FG acted a bit strange in the /4.

Now we just need to see how the finalgrom works in a 99/8! 😃

[+OLD CS1]

I had an hour or so to play around last night, and aside from some corrupt images on my SD, all worked well on both my QI and standard non-QI consoles.  Unfortunately, I found my QI console does not play well with my WHT-SCSI card so the standard console has to go back into play.  Bummer.

[ti99iuc]

On 5/8/2022 at 12:45 PM, dhe said:

Now we just need to see how the finalgrom works in a 99/8! 😃

I only tested it with FlashRom99 time ago
Would be lovely to have it working also with FG99! 

[JJB]

3 hours ago, OLD CS1 said:

I had an hour or so to play around last night, and aside from some corrupt images on my SD, all worked well on both my QI and standard non-QI consoles.  Unfortunately, I found my QI console does not play well with my WHT-SCSI card so the standard console has to go back into play.  Bummer.

Any idea why that is? Does it point to other incompatibilities between the 2 consoles?

[+OLD CS1]

5 hours ago, JJB said:

Any idea why that is? Does it point to other incompatibilities between the 2 consoles?

No ideas.  Just is, and I am concerned about my QI console as part of this test.  The only other QI console I have is currently in storage, mounted on MDF for VCF exhibition.

[JJB]

Hopefully a few other non-QI and QI users start testing; although I'm rather confident the solution works, it is a rather small sample at the moment.

 

Anyway until I hear otherwise I consider this problem solved for the moment.