What's new

MK2 - UB11 and UB13 failed, game stops

Asure said:
Try to set dip switches to disable the power-on test, and enable diag mode.

DIP 1 UJ1 - sw7 on - skip powerup test
DIP 2 UJ2 - sw8 on - diag mode

You should be able to enter diag mode i think.
Click to expand...

Yes did that and got into a menu, but a bit garbled graphics, but doesn't seem to be able to get it to continue the testing of the roms etc ? - is that at all possible, or will it just stop there until these errors are sorted?.


btw, anyone know how the signal from Pin 12 on the MAX691 (UH6) is supposed to be, seem like this goes to UJ11 pin 26 (CE2) port, and also is the VCC of the UJ11 supposed to be 5V or less?
 
Last edited:
Anyone by any chance ever traced the line between Vout of the MAX 691 (pin 2) and down to UJ11 VCC ? (and might have any tips on where it goes?, does it go under the battery holder by any chance ?., find it a couple of places around R138, but no been able to trace it fully....
 
The MAX691 chip controls voltages for the entire board. Once that chip is told to wake up, it turns on everything else

Generally all logic on the board will be 5V but some may need -5V as well on Midway boards

The MAX691 is not your issue or you would have a lot more chips not working/reporting in

You need to look at the schematic and check for continuity on all of the pins of the chips that are reporting as bad

If those all beep out ok then you need to check the signals to see whether the RAMs are the issue or something they are connected to

Please read some repair logs and guides on troubleshooting. This is not going to be a magic bullet solve of a random IC on the power circuit

MK2 has a full schematic available which makes it significantly easier
 
xodaraP said:
The MAX691 chip controls voltages for the entire board. Once that chip is told to wake up, it turns on everything else

Generally all logic on the board will be 5V but some may need -5V as well on Midway boards

The MAX691 is not your issue or you would have a lot more chips not working/reporting in

You need to look at the schematic and check for continuity on all of the pins of the chips that are reporting as bad

If those all beep out ok then you need to check the signals to see whether the RAMs are the issue or something they are connected to

Please read some repair logs and guides on troubleshooting. This is not going to be a magic bullet solve of a random IC on the power circuit

MK2 has a full schematic available which makes it significantly easier
Click to expand...

Yea, but I suspect the trace down to UJ11 is damaged due to the high resistance, and low voltage on the VCC pin of the UJ11 (which from the schemtic it looks like it's supposed to have a direct line from Vout (pin 2) on the MAX chip, at least the way I read it there shouldn't be anything else in between these points, was hoping that might be causing the u49 bad issue that it have - before I try to sort the video ram issue it also have...:)

thnx for the input, appreciated :)
 
The fact that you’re getting to the screen that you are means the MAX691 is not the issue

If that chip wasn’t working you would have low/no power to the entire board
 
xodaraP said:
The fact that you’re getting to the screen that you are means the MAX691 is not the issue

If that chip wasn’t working you would have low/no power to the entire board
Click to expand...

Yea, not thinking the 691 is failing (it also does have power on the Vout as expected, but from roughly 5v on the Vout, it drops down to about 1.3V on the UJ11 chip - which doesn't seem right to me - it also gives a U49 error, which I guess is expected if the voltage to UJ11 is that low?)
 
Ok I just brought up the schematic

That chip is keeping the scores in memory using the coin cell battery in addition to providing the watchdog signal. So yes that would be correct on the Vout which will be much lower while the board is running, since it is using the main power rail

It is not connected to anything else on the power rail including UJ11

It has 5 inputs

Pin 1 - Voltage in from coin cell battery
Pin 3 - Vcc in
Pin 4 - GND
Pin 11 - Watchdog input (jumper to manually stop the watchdog barking)

and 4 outputs

Pin 2 - Voltage out to keep scores/config in memory when board is not powered
Pin 12 - Enable VRAM (start the test mode/boot process)
Pin 14 - Watchdog monitor line, stops the board if there is an error and resets
Pin 15 - Reset

Please stop focusing on that chip, get the schematics and start testing traces around the RAMs that the board is literally telling you are not working correctly

tl;dr if you are seeing an image on the screen, it is not the power management IC
 
As I was going to have the UJ11 put in a socket eventually anyway I opted to remove it and do a test run of the board again, causing other errors (actually checking out the previously bad UB11 and UB13 (granted the test itself can be trusted ?).

However from what I see online the UJ23 isn't supposed to be populated on this board, but it also reports error on this test - is the reference wrong in the test or is it supposed to report bad ?. (or combined into any of the other chips?)......


There is still errors but seem to be a step further it seems....


20250930_191427.jpg
 
Does your board have the rom daughter card? If not, dips are set wrong? (set it for 8M)
The daughtercard just holds the other half of the rom(s) and the test will test only the first half of the mainboard roms when set to 4M instead of 8M.

Check here for rom layout: https://github.com/mamedev/mame/blob/master/src/mame/midway/midtunit.cpp
(Search for 'equiv' to find the table.)
Seems to me UJ19/20/22/23 should be populated w/ 27C040/4001..

Did you fix the ram errors? If not, seems to me UJ11 was hosing the address bus then.
 
Asure said:
Does your board have the rom daughter card? If not, dips are set wrong? (set it for 8M)
The daughtercard just holds the other half of the rom(s) and the test will test only the first half of the mainboard roms when set to 4M instead of 8M.

Check here for rom layout: https://github.com/mamedev/mame/blob/master/src/mame/midway/midtunit.cpp
(Search for 'equiv' to find the table.)
Seems to me UJ19/20/22/23 should be populated w/ 27C040/4001..

Did you fix the ram errors? If not, seems to me UJ11 was hosing the address bus then.
Click to expand...
Only thing I did was remove the UJ11 so after that it passed the initial ram errors, so looks like it was related to the problems relating to UJ11.

From wiring diagram, my impression is that there is supposed to be a straight connection from vOut of the 690 to VCC of UJ11 and I was expecting a low resistance reading, and pretty much vOut voltage at VCC, but on mine there is quite a high resistance and lower voltage that vOut. (anyone able to confirm how many ohms and voltage they read from vOut and to VCC?, I guess the Cap "hanging" on that line might interefere for reading resistance maybe - not sure.

This is a version without the daughter board, can't find any dip for selecting thjs? (the uf17 is a17748 as expected for 8 mb)
 
mawrick said:
Only thing I did was remove the UJ11 so after that it passed the initial ram errors, so looks like it was related to the problems relating to UJ11.

From wiring diagram, my impression is that there is supposed to be a straight connection from vOut of the 690 to VCC of UJ11 and I was expecting a low resistance reading, and pretty much vOut voltage at VCC, but on mine there is quite a high resistance and lower voltage that vOut. (anyone able to confirm how many ohms and voltage they read from vOut and to VCC?, I guess the Cap "hanging" on that line might interefere for reading resistance maybe - not sure.

This is a version without the daughter board, can't find any dip for selecting thjs? (the uf17 is a17748 as expected for 8 mb)
Click to expand...
Dip UJ 1 -#8 should be set to ON for the 8M version. The test would show 1 board then as well.
Can you show a picture of the board? Are all gfx rom sockets populated? They should be.
 
Asure said:
Dip UJ 1 -#8 should be set to ON for the 8M version. The test would show 1 board then as well.
Can you show a picture of the board? Are all gfx rom sockets populated? They should be.
Click to expand...
that seem to report a bit more correct, so guess there is some issues here, though I suspect there is something between the UH6 and UJ11 that's not realy working as it should.

gona pull the 2 roms and check em against mame roms to see if they check out. I guess these might cause corrupted graphics as well ?.

The game actually now seem to run a bit further, able to select fighter (previously it crashed on that screen), but it will either have very corrupt graphics, or just load in and crash/reboot.

20251002_194800.jpg
20251002_194630.jpg
 
Those two roms seem suspect, but it could be a data or address line that's borked causing issues too.

The game gfx are spread around those roms, if any of them is broken it's very likely to crash, the 'unzip' routine used for character bio pics and other static assets is not robust and will crash, trigger watchdog & reset if it unzips wrong gfx data.

The test marks your RAM as good and is properly programmed, so that's a plus. Same for ROM tests.
Test w/ romident (online: https://romident.coinopflorida.com/). Read as 27C801/27C080.

If they're bad, that's your reason right there.. Kung Lao and Lui Kang are in the first 1M so those sprites should be good?
(It's hard to tell looking at the source what will break for sure, i guess all backgrounds at least, these are in the last gfx area.)
 
Asure said:
Those two roms seem suspect, but it could be a data or address line that's borked causing issues too.

The game gfx are spread around those roms, if any of them is broken it's very likely to crash, the 'unzip' routine used for character bio pics and other static assets is not robust and will crash, trigger watchdog & reset if it unzips wrong gfx data.

The test marks your RAM as good and is properly programmed, so that's a plus. Same for ROM tests.
Test w/ romident (online: https://romident.coinopflorida.com/). Read as 27C801/27C080.

If they're bad, that's your reason right there.. Kung Lao and Lui Kang are in the first 1M so those sprites should be good?
(It's hard to tell looking at the source what will break for sure, i guess all backgrounds at least, these are in the last gfx area.)
Click to expand...
So I should be able to read them as 27C801/27C080 (I guess it doesn't matter what brand ?), in the schematic it seem to say 27c040-120 but I guess these maybe refer for 4Mb?

Any tips on what roms I should buy to burn new ones ?, as I guess the originals are a one time program type or (and might be damaged)?


Tried reading them as 27C801/27C080, but only get Device insertion error or damaged device (eetools burner, though I think this is mainly a check to see if it's a matching manufacturer so have tried to read them anyway), have to get samba share up and running from that xp machine before I can check the dump)
 
mawrick said:
So I should be able to read them as 27C801/27C080 (I guess it doesn't matter what brand ?), in the schematic it seem to say 27c040-120 but I guess these maybe refer for 4Mb?
Click to expand...

They should read as 27c080 any brand.
Try to read one of the roms that test good, that should work. I guess the other two are damaged. You can replace with 27c080.
The 27c040 is for the 4m with expansion board indeed. (You could try to read as 27c040 and get half the data as well.)
 
Asure said:
They should read as 27c080 any brand.
Try to read one of the roms that test good, that should work. I guess the other two are damaged. You can replace with 27c080.
The 27c040 is for the 4m with expansion board indeed. (You could try to read as 27c040 and get half the data as well.)
Click to expand...
I think maybe the error message is related to the programmer suspect the chips to be of other brand/type as I tested with a known good neo geo rom, and it show up ok when the right chip is selected but I get same error message if I choose the wrong type of chip.

But will try to read one of the chips that reads good on the board anyway to see - it should still be able to do a correct dump though :)
 
Yes, try again with a known good 27C080 from the MK2 board like i suggested.

Damaged chips won't read. Don't use detection or any auto settings, just set it to 27C080 from any brand. Reading is not affected by what you choose, and they mostly use similar writing algoritmes. You should just be able to read any of them.
 
Know of any good sources for those eproms, or only option OTP versions these days? (also what speed are they supposed to be)
 
I used to get my stuff from buyicnow.com but these days, i don't know.. anyone else? (aliexpress seems to have many fakes these days..)
 
You must log in or register to reply here.
Back
Top