What's new
0x39 is the index of the sound being played but there's probably an equivalence table of pointers in ROMs.
Above 0x39 you might be using higher part of ROMs, @Fluffy question also makes sense if the 0x39 sound is stored across lower and higher parts of the ROMs.
 
As for the music, I expect the instruments are in the sample roms and the actual sequence data is in the program rom. It is quite possible that the CPU tries to play music but because the samples are missing you don't hear anything.
 
Does 0x39 play to the end, or does it sound cut-off?
OK so i had a more in-depth look comparing to what mame does. Sound 39h is totally different and sounds messed up on the original hardware. There are also a few other ones lower down that dont play either. 25h and 26h. I'm not sure if quoting the sound ids in the test is any use if its just a lookup table, as i guess they could be from any memory address?? Damn..

Most of the sounds below 39h do play and are the same as mame including 38h, so thats good.
 
I'm not sure if quoting the sound ids in the test is any use if its just a lookup table, as i guess they could be from any memory address?? Damn..
Could very well be.
Maybe probe higher address lines while playing the sounds and try to identify a pattern.

[EDIT]
Can't you see in MAME where each sound points to?
 
OK so i had a more in-depth look comparing to what mame does. Sound 39h is totally different and sounds messed up on the original hardware. There are also a few other ones lower down that dont play either. 25h and 26h. I'm not sure if quoting the sound ids in the test is any use if its just a lookup table, as i guess they could be from any memory address?? Damn..
Most of the sounds below 39h do play and are the same as mame including 38h, so thats good.
It depends how Taito's original development environment built or re-build the soundbanks after changes, or if there are special rules how they need to be packed. Though if samples at earlier indices sound different you could listen if you can identify the sample it is taken from.
Maybe something odd with the top address lines, which means the address wraps earlier than expected, or you just get the tail end of the higher bank?
 
Some more analysis of the sample roms when test sounds are called in the service menu.

Seems that address lines A0-A7 pulse all the time. When i play a sound at 39h or lower i get activity on all address lines from A8 up to and including the top one at A19 (pin42) and activity on the data bus. When a sound fails to play in the test i get no activity on the address lines A8-A19. The data lines are only active at D0-D7 so i assume the 27c160 rom is configured for 8bit data use only. D8-D15 are dead, no activity.
 
Is it possible that one of the sound program roms is broken/corrupted? If they are 2x 8 bit for the M68k's 16 bit bus, it could be that the one containing the low address byte of the sample is working, while the high address byte returns 0?
 
Sorted it!! :)

Turns out the upper 2 pal ics at IC16 & IC17 don't need to be changed. Strange. Originally i was following what i saw online with the original rayforce pcbs, but seems these 2 ics can be left original cup final versions. I have lots of info on the differences between the pal versions in the cupfinals board and the rayforce one that i will share later, but i wanted to add a quick note now.

So the only thing need to fix the sound was the missing address line from roms to 68k/JP1 jumper board error and a recreated pal image for the new pal at IC14.

I added some quick videos on my instagram feed here.

https://www.instagram.com/p/B-pTYMHHX3O/
https://www.instagram.com/p/B-pOozqn0UE/

Please excuse the bad playing and also the colours may not look perfect as the video recorded them a bit differently.

So now i have a perfect rayforce/gunlock single layer board :)
 
Nice to hear you sorted it! Digging too deep can be trouble sometimes :)
 
Wow great work. Congrats!! @mypinballs can you please summarize what was actually needed to convert that game?
 
Here are my notes on converting a single layer (white pcb) cup finals to gun lock

Board Number K1100743A / J1100313A

ICs to swap

D66-01 IC2 Sound Samples 8-Bit Soldered in - 27c160
D66-02 IC3 Sound Samples 8-Bit Soldered in - 27c160

D66-23 IC10 Sound Prog 16-Bit Socketed - 27c020
D66-22 IC23 Sound Prog 16-Bit Socketed - 27c020
D66-09 IC14 Sound Prog Pal Socketed - GAL22V10

D66-18 IC24 Game Prog 32-bit Socketed - 27c020
D66-19 IC26 Game Prog 32-bit Socketed - 27c020
D66-24 IC35 Game Prog 32-bit Socketed - 27c020
D66-21 IC37 Game Prog 32-bit Socketed - 27c020

D66-03 IC45 Sprites 16-bit Soldered In - 27c800
D66-04 IC46 Sprites 16-bit Soldered In - 27c800
D66-05 IC47 Sprites 16-bit Soldered In - 27c800

D66-06 IC48 Tiles 16-bit Soldered In - 27c800
D66-07 IC49 TIles 16-bit Soldered In - 27c800
D66-08 IC50 Tiles 16-bit Soldered In (Mame rom is listed wrongly as IC49) - 27c800

Sound JP1 fix / A17 Sound Prog Addition

Pin1 - 5V
Pin2 - Pin31 of Sound Prog Roms instead of Pin30
Pin3 - A17 (A18-68K pin46)

Easiest to leave JP1 alone and add a jumper from IC10 & IC23 Pin30 to 68k Pin46

IMG_6399.jpg


Additional Notes

Pals installed at IC16 & IC17 should be left alone and kept as cup finals versions.
 
Last edited:
Nice work! Glad to see you were able to get it fully running.
 
great work @mypinballs !

Side note, I've gotten hold of a F3 Taito Cup Finals PCB but it is missing the PAL chip at IC28. I thought it was dumped and uploaded to jammarcade.net but I can't find the file?
 
great work @mypinballs !

Side note, I've gotten hold of a F3 Taito Cup Finals PCB but it is missing the PAL chip at IC28. I thought it was dumped and uploaded to jammarcade.net but I can't find the file?
Hi,

IC28 is labelled as D29-14 which I think originated from Arabian magic. The file is available at jamma-arcade under the hat trick hero'93 title as thats the alternative name for cup finals

You will need a GAl20v8 gal which are hard to get to use the file in its original form though

If you are stuck I can help you and supply a chip/file for use with a modern ATF22V10 gal. There is some conversion needed as the 22v10 ics are larger in size and so you can't just use the original file with the newer ics.
 
Gals are easy to find if you don't mind buying from ebay / china.
 
Back
Top