[Repair Log] Wild West Cowboys of Moo Mesa

[Takosuke]

Alright, let’s take a look at the Cowboys of Moo Mesa board I just got.

No video, no sound.

A quick inspection of the board shows a few things:

- The edge connector has a missing pad on the solder side, at the emplacement P where the Video Sync should be
- there is a blob of solder on the pinhole right beneath C and D where the 5V is (not sure if this one matters?)

- The sound module 054864 seems to be out, probably the cap leaking as usual for those board

- one of the legs of the Konami 054157 IC seems to be bent inside, but a quick continuity check does not show that it’s broken or shorting with anything (seems safe for now)





My first step will be to repair the edge connector.

Any advice on connecting the metal pad to the super tiny trace? Usually it’s easy to solder on a bigger trace but that’s the first time I will have to do it on something so small

 

[Takosuke]

I’ve repaired the missing edge connector using conductive copper tape (cleaned the connector good, used epoxy to fix it to the board and get the very edge of it covered to prevent it to peel off).



Now I finally have something showing up on the screen!


Unfortunately, nothing useful :/ just those jailbars, and eventually the board resets to the same bars, over and over again. Looks like the watchdog is doing his job.

I bridged JP1 to avoid it to reset while I troubleshoot.

The sound is just a scrambled mess, but that was expected, looking at the sound module. Could it be enough to create the issue?

There does not seem to be any broken trace or lifted/shorted leg on the ASICS.

What should be a good next step?

I’m about to remove the EPROMs and compare them to the Mame Roms, but I am not sure this will be the issue.

 

[Takosuke]

I’ve checked the 4 EPROMs on the board against name, and found one outlier out of 4 that did not match:

151aab02.Q6

After programming an ST 27c2001 with the file, I replaced it and... progress!

Now the jailbars are still here, and some colors are off. But the game is running and plays fine!

I assume that some ram might be faulty since the issue are with the sprites and color, need to dig deeper but still open to advice

 

[Takosuke]

Investigation continues. Looking at the initial check and the Mask rom check from the service menu, I can see 3 reported issues:

• B6 - Mask rom 8M/16M - ref HN624116P, the eprom equivalent is a 27c160-100.
• T10 - Mask rom 8M/16M - ref 151A06, I assume the eprom equivalent is the same
• D5 - 64K SRAM 150ns KM6264BL-7

I tried to program the dumps for B6 and T10 and place the eproms piggybacking on the Mask rom, but the game gives the same error during the check.

- Is piggybacking a good thing to do in that case?

I figure next step would be to try replacing the Sram, I will try to find some on ebay and see.

- Could the issue come from something else? Any advice on what I can check?

 

[ShootTheCore]

There's some repair logs for this board on JammaArcade.net written by @caius. One of the logs shows a RAM failure similar to the one you're seeing:
http://www.jammarcade.net/wild-west-c-o-w-boys-of-moo-mesa-repair-log/


So check the RAM chip at D5 for activity on all the lines with a logic probe, and try piggybacking another RAM chip on top of it if you have one.

As for the B6 and T10 ROMs, I can't speak to that specific board myself, but here's some general advice. As I understand it, I think piggybacking only works for mask ROMs if the original ROM is still storing data correctly but has failed internally on one or more pins. If the mask ROM is still active internally on all pins but is storing corrupted data, the data retrieved when the board asks for it will be scrambled with a mix of both the wrong data from the bottom mask ROM and the correct data from the piggybacked EPROM - resulting in a failed ROM test for the piggybacked ROM even if all the circuits are all working correctly.

I suggest that you use a logic probe and check each mask ROM for activity on the address and data lines - one of them may be floating (disconnected) or inactive (held high or low - note that ground is always low and Vcc is always high).

If you find a floating line, check the board for a break in the circuit trace between the floating ROM pin and its source - use the continuity setting on a multimeter to verify a good connection between the two.

If you find an inactive line, trace it to its source - it might lead to a TTL chip. Sometimes TTL chips are used to handle addressing, and a fault in the TTL chip can cause incorrect addressing and the ROM test will fail. If the inactive line traces back to a TTL, try piggybacking that TTL with a fresh one. If the piggybacked TTL works correctly, desolder the bad TTL, install a socket, and install the good TTL into the socket.

If a floating or inactive line traces back to a SMD CPU or graphics custom chip, try lightly pressing down on the chip and see if the issue resolves - the SMD may need to be reflowed.

If all the lines on the mask ROMs show proper activity, the ROM may have failed internally - very common with Konami mask ROMs in my experience. Desolder it, install a socket, and then install a programmed equivalent EPROM in its place.

Hope this helps.

 

[Takosuke]

Wow thanks a ton for the write up.
I’ll definitely check all that.

I have a logic probe but never really had to use it yet, so this will be a good test I keep track of the repair in a log here to both learn and keep me accountable so all these advice are super valuable thank you again!

I did read all the repair logs I could find including the one above, but I am still processing what exactly is being done in them

Also not having a proper test bench (yet) makes things harder to troubleshoot. But I’ll get there!

 

[Takosuke]

Some progress:

• I removed the Sound module that not only had leaky caps but was broken as well (kept a broken part with the asic from it, having a hard time to remove it from where it's soldered too though - might be useful for a replacement module later). I intent on replacing it by one from a faulty Martial Champion donor.

• I'm having trouble finding replacements for the sRam chip that is declared as faulty. I thought I could salvage one maybe from a donor board, but got no luck there
  • It's a Samsung KM6264BLS-7 (Standard SRAM, 8KX8, 70ns, CMOS, PDIP2 that seems to be 70ns instead of the 150ns written on the PCB. The only ones I found readily available are not the correct size or to be ordered by a large amount after quotes.
  • Is there any slot-in, more broadly available equivalent that I should look for?

• Also, any confirmation for a replacement eprom for the mask rom 151A06 in T10? Could not find much info about this one either or if a 27c160 would indeed work (the file for it in the Mame set is 1024 KB, as opposed to the 2018 KB of the B6 which is supposedly replaceable by the 27C160).

Thanks!

 

[buenrip]

Hi, for the last question, the correct EPROM for 1mb 42 pin are 27c800.
For the 6264 you can use some other brands, only need are the same size.
Regards.

 

[ShootTheCore]

A few suggestions:

1) You can purchase a reproduction sound module from @caius so you get a known-good part and won't have to worry about recapping the Martial Champions donor.

2) Try Ebay. I just did an Ebay search on KM6264BLS-7 and had several individual chips show up for purchase.

3) I believe a 27C400 is pin compatible for those mask ROMs. The 27C400 is a 4 megabit (2048 KB) chip, but you could double up the data using DOS's binary copy command:

copy /b 151a06.t10 + 151a06.t10 new151a06.t10

 

[Takosuke]

1) thinking about it for sure. If I manage to desolder the asic that is

2)those searches return ICs that are not the same DIPs. The chips are larger than CDIPs (I even ordered some to realize it only when I received them :/) that’s why I am looking for equivalent. I found some M6264 8x8 CDIP28 but they were 20 or 25ns. Would faster ones not create any issues?

3)I do have some 27c800, if that works then great! If not I’ll look into 27c400 thanks!

 

[ShootTheCore]

27C800 is a 42 pin EPROM while 27C400 is a 40 pin EPROM. A 27C800 won't work - sorry.

 

[Takosuke]

Yup I just realized that. Would that be the same as the infamous 27c4000 that I am trying to get Eetools to support correctly? Also have a order of 24c4100. I guess I can place one for 27c400 now ;$

 

[buenrip]

I not remind the pins now, I'm not at home. But I'm sure 27c400 are 4mbit and the size are 512 kbytes.
If he use the app to double the size, he needs 27c160.
But I repair some konami PCBs and I think 27c800 are the correct.
Regards.

 

[Takosuke]

For T10: The size is 1024 KB - 42 pins
For B6: the size is 2048 KB - 42 pins (thus 27c160)
http://mamedb.com/game/moomesa.html

So it seems 27c800 (8 Mbit) is right if the pinout is compatible with the mask rom.

 

[buenrip]

Yes, 27c800 - 8mbit 42 pin.
https://www.google.com/search?q=27c...droid-samsung&sourceid=chrome-mobile&ie=UTF-8

 

[Takosuke]

Well, I guess I can say bye bye Moo Mesa

Replaced the Ram, replace the 2 mask roms B6 and T10 after putting sockets, and replaced the sound module after getting a Caius repro from Mitsu and transplanting the ASIC.

Booted the board up by keeping it service switch down, and... Super loud screeching sound, no signal. Turned it off, removed the sound module, and instead still no signal

When I turned it on before after removing the old broken sound module, I at least had an error showing up.

RIP for now before I have more time to continue troubleshooting

 

[ShootTheCore]

Check for continuity on the replacement memory chips you soldered in. It sounds like a RAM issue.