What's new
Ok, ok, ok. Found an additional write to the KeyCus when EEPROM isn't initialised:

filedropper.com/nr2mpru_3
I just tried this.
I see a starfield for a second, then the "version mismatch" message for a split second and then the board reboots, and it gets stuck in a boot loop like this.

I can actually coin it up and start a game while the star field is displayed if I'm fast enough but then it just resets. I can't get it to go into test mode though.
I get the same thing as well.
Here comes the pain of remote troubleshooting again... :cursing:
As said to @SnakeGrunger by email I had the feeling something was going on with the watchdog.

Please try those files:
http://www.filedropper.com/nr2mpru_4
They should fix service mode and possibly the watchdog issues. It's likely the game will enter service mode the first boot after the update. Just press SERVICE again to exit the menu.

Sorry for wasting your time guys...
 
Seems still stuck in a loop. I can add credits or enter the service menu for a few frames until it reboots again.
Could it be that I have not cleared the nvram ? How do you re-initialize it ?
 
if you're willing to do the work to patch these games then I'm more than happy to do the work testing them ;)
I do. I was quite annoyed to see the patch didn't work when protection seemed easy to bypass. I guess this is because a lot is going on regarding hardware part which isn't understood/needed/emulated in MAME.

Anyway, this weekend is the first with 2 days off (I've worked the past 5 Saturdays) but I'm on-call and have to support my customer, first remotely with my computer and then on site if I can't fix the issue. Being basically on my laptop the whole day I thought I could spent some time to fix this patch between 2 calls.
 
I can add credits or enter the service menu for a few frames until it reboots again.
I get the same.


Anyway, this weekend is the first with 2 days off (I've worked the past 5 Saturdays) but I'm on-call and have to support my customer, first remotely with my computer and then on site if I can't fix the issue. Being basically on my laptop the whole day I thought I could spent some time to fix this patch between 2 calls.
I'm honestly surprised you found the time to work on this at all, I thought you were taking a few months off. :)
 
Seems still stuck in a loop. I can add credits or enter the service menu for a few frames until it reboots again.
Could it be that I have not cleared the nvram ? How do you re-initialize it ?
According to what I've found in MAME I'd say it clears by itself if not properly initialised.
There might be a way to disable watchdog (jumper?) on the board.
I'm honestly surprised you found the time to work on this at all, I thought you were taking a few months off. :)
I'm stuck in front of my computer...

[EDIT]
Have you guys moved the jumpers to accommodate different ROM sizes?
 
Last edited by a moderator:
Can you guys please give me what donor game you used (MAME romset name).
The number written to RNG varies from KeyCus to KeyCus so maybe the conversion needs to be adapted for each donor.
 
most of the Jumper locations on the PCB are unpopulated... I went through and listed them all though:

JP1 - 1M-4M for SPRG EPROM
JP2 - (unpopulated) 4M-1M for SHAPE EPROM
JP3 - (unpopulated) permanently connected via trace near Audio Amp (looks like it might enable/disable power to the amp circuit)
JP4 - (unpopulated) L-R near Audio Amp (maybe mono/stereo jumper?)
JP5 - (unpopulated) /1M-1M for CHR EPROMs
JP6 - (unpopulated) 8M-/8M for CHR EPROMs
JP7 - (unpopulated) 4M-1M for DATA EPROM
JP8 - (unpopulated) 4M-1M for PRG EPROMs
JP9 - (unpopulated) CON-COFF near main CPU (related to clock speed?)
JP10- (unpopulated) 24M-12M near unpopulated OSC location (related to clock speed)
JP11- (unpopulated) 24M-28M near unpopulated OSC location (related to clock speed)
JP12- (unpopulated) 355-F32 near C355 (related to object ROMs?)
JP13- A20-BYTE for VOICE EPROM


JP1 is the only jumper that should need to be changed when going from Super World Stadium to Nebulas Ray since SWS uses a 4M ROM here and NR uses a 1M ROM.
Every test I've done, I've tried this in both positions and it doesn't seem to make a difference. Pictures of some of the other conversions I've seen have left this jumper in the 4M position despite using a 1M EPROM.

the only jumpers that seem to do anything with the CPU are 9, 10, and 11.

JP10 and JP11 are near an unpopulated 28MHz OSC position and seem to be designed to change the clock speed

JP9 is a 2 position jumper labeled "CON" and "COFF" and has a permanent trace trace connecting CON to the center pint
The right pin is tied to ground,
The center pin of this jumper goes to Pin 15 on the MC68EC020 "SCKR"
the left pin goes to a 222 Resistor network, the outputs of which go to
-pin 22 on the MC68EC020 (HRW/HRD)
-pin 19 and pin 1 on the MC68EC020 (GNDQ and SCK/SCL)
-nothing
-the unpopulated J104 connector (but populated on Point Blank for the gun board)

Here's a photo of that area:
jp9.jpg
 
Can you guys please give me what donor game you used (MAME romset name).
The number written to RNG varies from KeyCus to KeyCus so maybe the conversion needs to be adapted for each donor.
I'm using KEYCUS C434 which is from MAME ROM sws97 (Super World Stadium 97)

FWIW I've seen other conversions of this game for sale that didn't have any KEYCUS installed at all, so it seems possible a universal solution could be made.
 
it's not a matter of jumpers, the patched code is simply not good.I have two ROM sets, one requires the KEY CUSTOM IC from Point Blank and the other runs with no KEY CUSTOM installed and they both works fine.But I can't release them for obvious reasons (neither sell them so don't ask in private)
 
it's not a matter of jumpers, the patched code is simply not good.I have two ROM sets, one requires the KEY CUSTOM IC from Point Blank and the other runs with no KEY CUSTOM installed and they both works fine.But I can't release them for obvious reasons (neither sell them so don't ask in private)
I've patched all accesses to the KeyCus in MAME and it works perfectly so I think there's something going on with the hardware part.
There seems to be a link between the KeyCus and the watchdog, that would mean conversions made without any KeyCus installed rely on hardware modding.
I've done dozens of tests, disabling RNG, running patched ROMs with a different romset, disabling KeyCus area, etc. And I was never able to fault the code in MAME.
There are so many things that look weird in MAME, it just makes no sense from a hardware point of view so it's probably wrongly handled.

@twistedsymphony can you take a video with the KeyCus installed and then with no KeyCus?
 
Last edited by a moderator:
it's not a matter of jumpers, the patched code is simply not good.I have two ROM sets, one requires the KEY CUSTOM IC from Point Blank and the other runs with no KEY CUSTOM installed and they both works fine.But I can't release them for obvious reasons (neither sell them so don't ask in private)
I've patched all accesses to the KeyCus in MAME and it works perfectly so I think there's something going on with the hardware part.There seems to be a link between the KeyCus and the watchdog, that would mean conversions made without any KeyCus installed rely on hardware modding.
I've done dozens of tests, disabling RNG, running patched ROMs with a different romset, disabling KeyCus area, etc. And I was never able to fault the code in MAME.
MAME is NOT a good way to test conversions since it's only emulation which, as name tells, can't reproduce hardware 100% faithful :)
Your patched code is bugged, that's all.The one I have has been not prepared by me but it works perfectly.
There is no hardware modfication needed on board with no KEY CUSTOM installed, I still have my conversion here and can confirm I didn't do any changes on board.
 
Last edited:
MAME is NOT a good way to test conversions since it's only emulation which, as name tells, can't reproduce hardware 100% faithful :) Your patched code is bugged, that's all.The one I have has been not prepared by me but it works perfectly.
With the hardware in my hands it would have been solved quickly.
I don't think my code is bugged, however there's probably something missing, something not initially present in the code but needed when using a different KeyCus or no KeyCus at all.
 
For bugged I meant not correct.Anyway, I can confirm no hardware modifcations are needed both with original KEY CUSTOM and no KEY CUSTOM installed.
 
@twistedsymphony can you take a video with the KeyCus installed and then with no KeyCus?
I pulled the KEYCUS and it boots without restarting! I'm able to coin up and play.

for some reason though the right direction on both joysticks does not register. Also settings in test mode are not saved (for instance I enabled screen flip and as soon as I exit test it is flipped back).
 
I pulled the KEYCUS and it boots without restarting! I'm able to coin up and play.
The KEY CUSTOM must be always removed when using code that bypasses it


for some reason though the right direction on both joysticks does not register. Also settings in test mode are not saved (for instance I enabled screen flip and as soon as I exit test it is flipped back).
This is probably an issue with board, check the SMD filters near the JAMMA connector (they lose contact very easily) as well as the multiplexers.
 
@twistedsymphony can you take a video with the KeyCus installed and then with no KeyCus?
I pulled the KEYCUS and it boots without restarting! I'm able to coin up and play.
for some reason though the right direction on both joysticks does not register. Also settings in test mode are not saved (for instance I enabled screen flip and as soon as I exit test it is flipped back).
So you mean it actually works better with no KeyCus installed?
At least it's progress but it's also getting weirder, what's the link between right directions and the KeyCus? And settings?

[EDIT]
The KEY CUSTOM must be always removed when using code that bypasses it
Damn, this is useful information I wish you had posted earlier.

[EDIT 2]
Again flip screen setting does work perfectly in MAME with the patched code.
 
Last edited by a moderator:
So you mean it actually works better with no KeyCus installed?
yes. with KEYCUS it reboots, without KEYCUS it works


what's the link between right directions and the KeyCus?
I think this is just my PCB... I swapped back to original ROMs and right direction still doesn't work. I'm thinking cold joint on filter SMD as caius suggested. probably developed from handling the PCB so much as it definitely worked before.


And settings?
swapping back to original ROMs settings work. so maybe EEPROM is initializing even if it doesn't need it?
 
I think this is just my PCB... I swapped back to original ROMs and right direction still doesn't work. I'm thinking cold joint on filter SMD as caius suggested. probably developed from handling the PCB so much as it definitely worked before.
The SMD Murata filters are a common issue on these boards, they lose contact from pads simply handling/flexing the PCB.
 
Back
Top