JAMMArcade.net – Page 10

Operation Wolf repair log #2

Mar 272020

Got this board set from Paulcan69 to have a look at.
The faults reported were gun shots do not register & there was no sound.
I was told before hand that the gun issue seems to be a very common problem so I focused my attention on that first.
The input from the gun comes in from the sound PCB and passed straight down to the main PCB.
First I checked the signal from the gun was actually reaching where it should do

I confirmed that on a trigger press in game that the signal reaches pin 3 of the 74F74 at IC26 and also that pin 5 output latched.
Using the scope I could also confirm the operation of the 74LS75 at IC38.

The current XY screen draw position is tracked by a set of 74LS161 counters and fed into 74LS373 transparent latches.

When the trigger is pressed the the 74LS373’s are latched and the current XY values are sent to 3 x 74LS257 data selectors.

These selectors are read by the main CPU from address $3A0000 – $3A0001.
Using the scope syncing on pin 15 of the 74LS257’s I confirmed that some of the outputs on these were always logic HIGH.
I ended up replacing IC67 & IC83 which fixed the gun problem.

Next was the sound issue.
This one had no sound at all.
All the signals looked OK from the CPU side but nothing was ever being asked to play a sound.
Given the fact that the board uses the Sony brand RAM I just went right ahead and removed it as they have been very unreliable in recent times.
It failed in an out of circuit test using my programmer

Now the music was back but the samples did not.
I already had a pretty good grasp on how the sound worked from my previous repair so started checking the same areas as before but didn’t really find anything but then I spotted this

This is part of the feedback circuit for the op-amp. Replacing this gave me some samples back but they cut out before they finished playing and some didn’t play at all

I knew exactly where to look for this and quickly found the 74LS688 at location IC41 with its output stuck LOW.

I removed and tested which obviously failed

I initially thought I would have to put this repair on hold as I needed to order some 74LS688’s but turned out I had to wait less than 24hrs for Farnell to deliver.
With the new chip fitted I had my sounds playing for the right duration but all the samples on the A channel were still missing altogether.
Using the scope I traced the sound to the output of the M5M5205 ADPCM Speech Synthesis LSI at IC50 (schematics incorrectly refer to it as IC51)

I could see the signal going into the op-amp but nothing was coming out so I replaced it.

I now had an output from the op-amp but still no sound. Tracing this through on the PCB I came across another damaged capacitor (no idea how I missed this one as well)

Replacing this brought all the sound back and that’s another one complete.

Operation Wolf repair log #1

General, PCB Repair Logs, Repair Logs 6 Responses »

Mar 262020

I recently got two Operation Wolf boards from Muddymusic to fix.
This board was not booting at all and I got a screen full of flickering garbage

I removed all four of the 27C512 program EPROM’s and verified they were good. All of them matched the “World, Set 1” version in MAME.

I confirmed continuity between all the address, data and control pins of both EPROM’s and RAM was present and correct and it was so I opted to remove the two TMM2063 work RAM chips and test those

As you can see both failed. I replaced them and tested but there was no change.
Next I started looking at the reset circuit.

Here is the main reset circuit taken from the schematics

Using the logic probe I found that signal “/TMCL” (timer clear) was floating. Shifting focus to grid A5 on the schematics we see that this signal comes from pin 13 on the 74LS138 at location IC60.

Further probing found that the A18 signal on pin 2 was also dead. This signal is the buffered MA18 (address bus pin 18) and comes from the 74LS244 at location 65. Checking the signal on the LS244 revealed that it was present so there was a break in the trace some where.
I couldn’t see anything obvious so I patched it with some Kynar wire

No on boot up I get no more resetting buyt the screen is full of garbage still

Next I started looking at the custom PC080 tilemap generators, more specifically the RAM connected to it

Checking both of these with the probe showed me there was no activity on most of these pins and some others were stuck high

I removed both of them and checked them, obviously both failed

Replacing both of these now gives me something to see.
In test mode I get this

I looked into this fault for a good hour or so before I decided the RAM I fitted was too slow. Swapping for a faster RAM type fixed both of these issues.

Now the board would successfully pass the self test in the test mode I could see that I had an issue with colours.

I started off probing the colour RAM at locations IC64 & IC65 and quickly found bit 4 of the address pins stuck low

Using the schematics I could easily trace this back to a 74LS157 at IC52

Replacing this gave me my colours back

Now onto the sound board.
The good news is all the music was present and correct. The bad new was all the voice samples were not.

Again, using the schematics I see that ultimately the addressing for the Maskrom at IC21 gets set by a bunch of 4-bit counters at IC27-IC30.
Probing these I found that the UP counter wasn’t pulsing at all

Checking the 74LS04 at IC32 revealed this inverter was bad so I replaced it and sound was restored but the sounds just carried on playing one after the other (never got a video of this).
The Z80 sets the start and stop data by writing values to address $C002 & $C003 respectively which gets latched by a couple of 74ALS574’s and fed into the B side of a 74LS688 comparator

Using the MAME debugger I could see that the data being written to the 74LS574’s was correct but probing the /A=B output pin 19 of the comparators showed that these were either never matching or they weren’t working.
I couldnt find issues up stream of these so I desoldered them and replaced them with pulls from another board as I don’t have any spares.

I now have working sounds and fully working board set ready to return.

Dangun Feveron repair log #2

General, PCB Repair Logs, Repair Logs 4 Responses »

Mar 262020

I had this Fever SOS (international version of Dangun Feveron) PCB since many Years. Here is a picture of the board with the two chips I had to replace highlighted in red:

It had flickering sprites (the problem being pretty hard to photograph, I’m not able to show screenshots of it). In other terms, every second frame of sprites was not displaying.

1) The sprites are generated by a large 240pin custom chip labeled 9838EX002 – 013 on this board. Closer inspection revealed a cross marked on it by a previous repairer as well as a mark on pin #226. That pin goes to the RAS (Row Address Strobe) pin (#14) of a DRAM (424260) at U36, which is one of the two DRAMs that deals with sprites, each of them displaying alternatively one frame after the other.

There was no signal at all on that pin (#14). Probing the equivalent pin on the adjacent DRAM at U35 showed a pulsing signal. RAS being a signal going towards the DRAM meant that the large custom chip wasn’t giving the signal needed at pin #226.

I needed to change that large custom chip. These being on Cave 68000 systems, they are unfortunately available mainly on rare and expensive games (list available in the Cave 68000 MAME driver). After a long time, I finally had the opportunity to grab a UO Poko board which has the same chip, even labeled not exactly similar: 9749EX004 – 013.

Left is the faulty chip taken off the Fever SOS board. Right is a working chip taken off a UO Poko board. Notice how the left one blistered after heating it for desoldering, while the right one didn’t change.

2) Unfortunately, swapping the chip did not resolved my sprite flickering issue but at least the signal at pin #226 was then looking good. Probing the linked DRAM at U36 showed some weird looking signals on a few I/O pins (#2, 3, 4 & 5 pins).

I swapped it with a functioning DRAM (a NEC 424260). That finally resolved the flickering sprites problem.

Board fixed !

Lightning Fighters repair log #4

PCB Repair Logs, Repair Logs Comments Off

Mar 242020

I got a Lightning Fighters PCB with GFX problems. As you can see, major colors and layers issues:

1) Probing the Konami 053251 custom chip revealed some weak signals on a few pins. This chip is a priority encoder, it deals with colors and layers and seems to be prone to failure in more and more boards of that generation. Unfortunately, you have to find a donor board of the same era to get a working one (good thing anyway is that it is used in a pretty good amount of Konami games).

Here is the result after swapping it with a good one:

2 & 3) Colors and layers are now fixed but we have some jailbars on sprites. This looks to be a sprite’s data issue. Probing the mask ROMs revealed missing signals on a few data lines for 939A05 and 939A06 ROMs. I replaced them with two burnt 27C400 EPROMs. Board is now fixed:

Here is an overview of the chips I had to replace on the board:

Konami Track & Field / Hyper Olympic Tester – BETA 1.3 & TEST 1.2

General, Technical Info 2 Responses »

Feb 212020

21/02/2020 10:59 ( AEST )

I’ve been meaning to update this page for awhile but work and life has got the better of me.

If your Track&Field/Hyper Olympic has the following layout then you have unencrypted roms and you will need this updated version of my diagnostic to run on this particular board. Note that this version will not run on boards equipped with a K1 cpu ( 42 pin custom IC ) or a game ( usually a bootleg ) with daughterboard that has a 6809 processor onboard, the daughterboard version also has some extra logic associated with it to decrypt/unscramble the opcodes.

Note: The board does not have a K1 CPU but uses a standard 6809 processor to run the game. Decryption related circuitry is also absent from this version.

I have prepared a version that will run on this board.

See TEST VERSION 1.2 – Unencrypted for download

9/02/2017

TEST 1.2 is now the current version which installs in socket 5A on Konami and Centuri versions.

Fixed detection of ROM A7 [ $9E99 ] ( Hyper olympic bootleg) which broke in TEST 1.1

See TEST VERSION 1.2 for download

21/08/2015

TEST 1.1

This is a test release. A BETA version will replace it and this post will be updated pending sufficient feedback. It has been reported by one tester that this version works on real hardware but more feedback is desired.

IMPORTANT: The diagnostic EPROM now installs @ 5A on the Centuri and Konami boards. For the Hyper Olympic bootleg install the EPROM @ A7 and for the NZ bootleg install @ 2A.

This frees up 4 sockets for verifying game EPROMs instead of only 3. Since 5A is mapped to 0xe000-0xffff and the reset vector is obtained at 0xfffe/0xffff, it made sense to move the test rom to this socket.

See TEST VERSION 1.1 for download

IMPORTANT: All versions below install in socket 1A

18/08/2015

Bullet points represent items fixed in the current version.

BETA 1.3 is current.

ROM A1 FOUND! text missing when generated check-sum verifies successfully in socket 3a & 4a.

TODO: Future versions will install in socket 5a. The main reason for this change is to ensure the test rom boots since the 6809 obtains the reset vector from EPROM A5 ( 0xfffe ), there is a chance that the reset vector could be corrupted if the game EPROM A5 is bad. Moving the code to run from socket 5a will prevent such an event and also make the first 4 sockets available for verifying an A5 EPROM or any others.

See 1.3 for download CHK: F3E9 CRC16: 1421

16/08/2015

BETA 1.2

Further checks suspended if video ram/colour ram is bad. Priority given to preserve NVRAM/high scores. ( worst case scenario – screen totally unreadable ).
Introduced audio alert. 5 beeps for bad video ram 12H, 6 beeps for bad colour ram 11H.
Provided work around to keep tiles on screen if scroll ram is detected as bad.

See 1.2 for download CHK: F3EB CRC16: 5695

Special thanks goes out to Purity on UKVAC for his help with testing and feedback.

1/08/2015

BETA 1.1

Fixed SFX menu. Speech now plays from sound index 0x80

If you believe there are missing sounds please report them in the comments.

See 1.1 for download CHK: 9909 CRC16: AB36

30/7/2015

BETA 1.0

Bootleg rom support added ( see hyprolymb from the MAME set )
Fixed watchdog reset issue during EPROM check routine
Fixed character bank bug ( was displaying extra bank of null tiles )
Added CRC of current software to version string ( calculated at runtime )

BOOTLEG – HYPER OLYMPIC ( hyprolymb )

ROM A1 $9E99
ROM A2 $35E8
ROM A4 $2FDC
ROM A5 $7DD2
ROM A7 $F932

TODO: Fix SFX test in diag mode. Speech is not playing.

See 1.0: CHK: 9F7E CRC16: 2F6F

26/07/2015

BETA 0.9

Support for Hyper Olympic roms added
MISC code improvements

See 0.9 CRC16: 70E6

See below for correct 16 bit checksums matching your set.

Pre-calculated ROM Checksums.

KONAMI – TRACK & FIELD

ROM A1 $9F6E
ROM A2 $3620
ROM A3 $2FA5
ROM A4 $7CA6
ROM A5 $FEEB

CENTURI – TRACK & FIELD

ROM A1 $9DBE
ROM A2 $3CA7
ROM A3 $17C8
ROM A4 $569E
ROM A5 $E991

KONAMI – HYPER OLYMPIC

ROM A1 $9E79
ROM A2 $35E8
ROM A3 $2ECE
ROM A4 $7DD2
ROM A5 $F932

25/07/2015

BETA 0.8

Support for Centuri roms added.

See 0.8 CRC16: 9EBE

24/7/2015

BETA 0.7

This EPROM installs on the CPU board @ 1A. This does not replace the original software in 1A but is only used as a means of testing the main-board, sound and video.

Diagnostic mode is accessible by holding down player 2 start whilst powering the game on.

In the normal mode ( power up tests ), nvram tests may be skipped to preserve high score data if so desired. This is done via DIP 8 of SW2. A ~10 second countdown is also included even if you desire to test your nvram, in this mode nvram tests may also be skipped by pressing the “player 1 button 1”.

Power up tests also include a rom utility. This currently identifies track & field software in any of the 3 available sockets ( 2A to 4A ). For example, an A1 ROM installed in any of the above available sockets will still be identified for its designated 1A socket.

Important: A5 ROM needs to remain in socket 5A, this is because the 6809 obtains its reset vectors from this ROM. If the test rom or game fails to start and the screen is full of zeros, verify the A5 rom manually.

Note: That this is a very BETA release and is currently in the testing phase, bugs are expected. Use at your own risk if you love those high scores.

See 0.7 CRC16: 8C8A

Older Entries Newer Entries