What's new

ShootTheCore

Enlightened
Joined
Jan 20, 2016
Messages
2,049
Reaction score
3,395
Location
Logan UT
In for repair is a Namco Outfoxies PCB, running on Namco's NB-2 platform. NB-2 has a large mainboard and a ROM daughterboard. The ROMs on the daughterboard are each 16 megabit SMD chips. The owner reported that the Professor character in the wheelchair displays jailbars.

IMG_9444.jpg

While waiting for the board to arrive, I fired up MAME and deliberately zero'd out the OBJ ROMs one at a time until I isolated which ones duplicated the PCB's behavior. OBJ2L and OBJ2U each provide half of the data for each Professor sprite frame. When either one of the files is corrupted then the character renders with jailbars. If both files are corrupted then the Professor isn't drawn at all.

Once the PCB arrived, I examined the ROM board. It turns out that the Namco OG OBJ2L ROM had already been swapped out with a flash equivalent by another technician. That technician left a bunch of flux behind, and it looks like he also resoldered the OG Namco ROMs physically located above it while he was at it - likely from not masking off the other chips when applying heat to OBJ2L, causing the other chips to come loose as well.

IMG_9442.jpg

I cleaned up the leftover flux with Q-Tips and isoproyl alcohol, then examined the solder joints of the three chips under the microscope. The solder joints checked out - I didn't see any lifted pins. Continuity tests checked out for all of the pins as well. For good measure, I reflowed the solder on the OBJ2L flash ROM and re-tested the PCB, but the jailbars on the Professor remained.

This is the point in the story where my chip programmer becomes a problem. I had been using an EETools ChipMax II for almost four years as my workhorse programmer, but it failed recently with a Pull Up Resistor error, and EETools refused to repair it outside of warranty. I had previously purchased an inexpensive Embedded Computers FlashCat-XPORT for SMD chip programming, so I tried to use it here. Unfortunately, the FlashCat turned out to be untrustworthy.

First off, the PCB owner included the original Namco OBJ2L ROM with the board - the prior technician had saved it after swapping it with the flash chip. I installed the ROM into the FlashCat and ran a comparison with the equivalent MAME file. The FlashCat reported that only 44% of the data on the ROM matched the MAME file, so it seemed that the ROM was indeed bad and that swapping it out had been the right move by the last technician.

IMG_9513.jpgIMG_9512.jpg

Next, I carefully desoldered the OBJ2L flash chip from the ROM board, inserted it into the FlashCat, and tried to run the same MAME file comparison. The FlashCat wouldn't recognize the flash chip so that it could be dumped.
image1.jpeg

Alright, maybe the flash chip had gone bad then. I programmed a fresh flash chip, soldered it back to the ROM board and tested. Unfortunately, the jailbars on the Professor remained - BUT a few of his animations (like swimming) now showed up jailbar free.
image3.jpeg

That leaves the other ROM responsible for this character's sprite graphics - OBJ2U - as a potential troublemaker. Maybe a second OG Namco ROM had failed?
I carefully masked off the chips around it and desoldered it from the board.
IMG_9509.jpg

I then placed the mask ROM in the FlashCat and ran a comparison against the equivalent MAME file. It reported a 68% match - so apparently this ROM had also gone bad after all.
OBJ2U-ROM-To-MAME-Compare2.jpgIMG_9510.jpg

I programmed a fresh flash chip for OBJ2U, soldered it into place, and feeling confident - fired up the PCB for testing.....
OBJ2U-Flash-Installation.jpg

Except to discover that now the Professor sprite isn't rendering AT ALL! Not even with swimming, or any other activity.

I don't trust the FlashCat at this point to have either read the other chips or programmed the new flash chips properly. I've ordered a new programmer (Dataman 48Pro2C) and will try again with verifying the OG Namco mask ROMs and programming any required flash replacements after the new programmer arrives. To be continued...
 
Last edited:

everten

Grand Master
Joined
Jan 27, 2019
Messages
403
Reaction score
705
Location
CT, USA
Nice work, hopefully the new programmer fixes the issue. I always forget that MAME is an awesome tool for troubleshooting.

I was getting ready to pickup a FlashCat, curious if others have had issues with them.
 

ShootTheCore

Enlightened
Joined
Jan 20, 2016
Messages
2,049
Reaction score
3,395
Location
Logan UT
Time for an update on Outfoxies.

  • My new chip programmer and appropriate SMD adapter arrived. I now have two chip programmers capable of reading and programming the chips Outfoxies uses – my brand new programmer which I trust and my old programmer which I don’t.
  • There are two original Namco mask ROM chips removed from the Outfoxies board at the moment – the OBJ2L chip which was replaced by an earlier technician that the owner included in the shipping package, and the OBJ2U chip which I replaced earlier.
  • I dumped both mask ROM chips twice – once with my new programmer and once with my old programmer. The dumps don’t match between programmers, and none of the dumps match the files for Outfoxies in MAME, so it does seem that both mask ROMs are corrupted.
  • I programmed a fresh set of 29F1610 flash chips for OBJ2L and OBJ2U with my new chip programmer, and verified each chip three times before soldering them to the ROM board.
  • I also reflowed the solder on the SMD mask ROMs physically located above OBJ2L and OBJ2U since they share address and data lines with the two chips I’m working on.
  • Unfortunately the Professor character is still showing up with jailbars in the game after installing the fresh chips.
There seems to be something else going on to cause the jailbars with the Professor besides the chips, although the jailbars are worse now then they were before I started. Is the board picky about the speed and type of 29F1610 flash chips I'm using?

It seems that @caius has diagnosed and repaired a similar issue with Mach Breakers, which runs on the same motherboard as Outfoxies. I followed his procedure for substituting blank files in MAME for each OBJ chip to isolate which ROMs generated the Professor sprite data. If you have any other suggestions, please let me know.
https://www.jammarcade.net/mach-breakers-repair-log/

Also, if anyone else experencied with this family of Namco boards has suggestions, let me know. @Hammy ? @Apocalypse ?

Next, I'm planning on hooking my logic analyzer up to one of the replaced OBJ flash chips to find out what's going on with the address, data, and control lines.

IMG_9839.jpgIMG_9842.jpgIMG_9840.jpg
 

twistedsymphony

Enlightened
Staff member
Immortal
Joined
Jul 21, 2015
Messages
10,443
Reaction score
10,296
Location
NH, USA
my guess would be there is some trace damage related to those chips you replaced (perhaps introduced in the previous repair attempt). considering all of the other object ROMs are on the same buss if it were something further downstream I'd think you'd see jail-bars across all sprites.

it's tedious but if it were my board I'd go pin by pin and verify continuity from each of the suspect ROMs to the either those logic ICs or the daughter-board connector
 

ShootTheCore

Enlightened
Joined
Jan 20, 2016
Messages
2,049
Reaction score
3,395
Location
Logan UT
The traces looked fine under the microscope, but you never know - SMD stuff is definitely fragile. Testing the continuity on each pin is a good suggestion - thank you!
 

caius

Grand Master
Joined
Nov 21, 2016
Messages
572
Reaction score
859
Location
Italy
If there are jailbars then the problem is in data bus of sprites circuit.Assuming that the object ROM board is good (no broken traces, good ROMS and logics, good connection with motherboard, etc..) you should draw the sprites generation circuit to figure out what processes the data from ROMs.Most likely it's a custom ASIC (don't have any board to check).The best would be to swap in a good motherboard so you can know for sure where the fault lies.Anyway, I think you can fix it with your skills and some time.
 
Last edited:

ShootTheCore

Enlightened
Joined
Jan 20, 2016
Messages
2,049
Reaction score
3,395
Location
Logan UT
I appreciate the vote of confidence! It definitely sounds like continuity testing, logic probing and logic analysis on the two OBJ chips will be the next steps.

I don't have another Namco NB-2 board around to swap the Outfoxies ROM board with, and the cheapest I can find Mach Breakers for is $450 USD. I'll try to get by without another NB-2 test board for now... :thumbsup:
 

JoyAgain

Student
Joined
Apr 30, 2017
Messages
41
Reaction score
23
Location
Italy
I appreciate the vote of confidence! It definitely sounds like continuity testing, logic probing and logic analysis on the two OBJ chips will be the next steps.

I don't have another Namco NB-2 board around to swap the Outfoxies ROM board with, and the cheapest I can find Mach Breakers for is $450 USD. I'll try to get by without another NB-2 test board for now... :thumbsup:
I also had the same problem as you with OutFoxies.
I solved it by dump my PBC already in possession and discover that there are several revisions of the graphic roms, Obj Rot e Scr not present in mame.
If you mix the Obj roms not of the same set the graphic problem gets worse.
I have the rom file in rar, but it's too big I can't attach it now in the message.
 

ShootTheCore

Enlightened
Joined
Jan 20, 2016
Messages
2,049
Reaction score
3,395
Location
Logan UT
I also had the same problem as you with OutFoxies.
I solved it by dump my PBC already in possession and discover that there are several revisions of the graphic roms, Obj Rot e Scr not present in mame.
If you mix the Obj roms not of the same set the graphic problem gets worse.
I have the rom file in rar, but it's too big I can't attach it now in the message.
If you could get those files to me somehow then I’d really appreciate it. Maybe use mega.io or upload it to a cloud storage service like Google Drive or Microsoft OneDrive.
 

ShootTheCore

Enlightened
Joined
Jan 20, 2016
Messages
2,049
Reaction score
3,395
Location
Logan UT
Dumps have been sent to a MAME dev for further analysis.
@caius I still have the two original mask ROMs on hand that have been removed from the board. I'm pretty sure OBJ2L is bad since I can't get the same read twice. OBJ2U reads consistently but its checksum doesn't match any of the Outfoxies sets currently in MAME.

Do any of my checksums match the checksums on this new dump?

OBJ2L Checksums
C801B782
10195B40
DE11052A
EE85A1E6

OBJ2U Checksums
CE74C385
 

ShootTheCore

Enlightened
Joined
Jan 20, 2016
Messages
2,049
Reaction score
3,395
Location
Logan UT
Brother, this game NEEDS to be fixed, it's that good!

I can't help on the tech side, but would you like me to mail you my PCB to help diagnose your issue?
That’s a very generous offer, and it would be helpful to me if I could swap the A board with another game to make sure there isn’t a problem there. Let me see what the outcome is with the checksums on this new ROM dump and we’ll go from there. Thank you!
 

JoyAgain

Student
Joined
Apr 30, 2017
Messages
41
Reaction score
23
Location
Italy
If you could get those files to me somehow then I’d really appreciate it. Maybe use mega.io or upload it to a cloud storage service like Google Drive or Microsoft OneDrive.
https://drive.google.com/file/d/1RwK1xQBGE-ETQ9pDwtvP8wEFA4HiGqO8/view?usp=sharing

Here are the links let me know if you have been able to download them bye.

First of all make sure that your set matches this new set, desolder a sure good rom and compare the crc32 with one from the new set. I hope you solve it bye.
 

caius

Grand Master
Joined
Nov 21, 2016
Messages
572
Reaction score
859
Location
Italy
@caius I still have the two original mask ROMs on hand that have been removed from the board. I'm pretty sure OBJ2L is bad since I can't get the same read twice. OBJ2U reads consistently but its checksum doesn't match any of the Outfoxies sets currently in MAME.

Do any of my checksums match the checksums on this new dump?

OBJ2L Checksums
C801B782
10195B40
DE11052A
EE85A1E6

OBJ2U Checksums
CE74C385
OBJ2U CRC32 matches the new dump, OBJ2L does not.
Anyway, MAME dev said that he tried to load the new dumps under MAME and he gets a lot of GFX issue.He will compare now the dumps to see which are the differences.
 

JoyAgain

Student
Joined
Apr 30, 2017
Messages
41
Reaction score
23
Location
Italy
OBJ2U CRC32 matches the new dump, OBJ2L does not.
Anyway, MAME dev said that he tried to load the new dumps under MAME and he gets a lot of GFX issue.He will compare now the dumps to see which are the differences.
obj2u new dump crc32 ce74c385
obj2u old dump crc32 ccada5f8 ?



they are all different Caius
 

caius

Grand Master
Joined
Nov 21, 2016
Messages
572
Reaction score
859
Location
Italy
obj2u new dump crc32 ce74c385
obj2u old dump crc32 ccada5f8 ?



they are all different Caius

You misunderstood what @ShootTheCore said.He asked if CRC32 of his dumps match yours (and not if your dumps match the old MAME ones)
CRC32 of the OBJ2U dumped by him matches your dump, plaese see snap attached from archive of your dumps.
 

Attachments

  • Snap15.jpg
    Snap15.jpg
    73.1 KB · Views: 4

JoyAgain

Student
Joined
Apr 30, 2017
Messages
41
Reaction score
23
Location
Italy
You misunderstood what @ShootTheCore said.He asked if CRC32 of his dumps match yours (and not if your dumps match the old MAME ones)
CRC32 of the OBJ2U dumped by him matches your dump, plaese see snap attached from archive of your dumps.
Ok, Does this mean that his SET is identical to my dump?
Obj2l maybe it's broken.
An encouraging thing!
 
Last edited:
Top