Froztbite repair logs

[froztbite]

Hi, i'm still a humble novice in learning so I try to absorb a lot from other people repair logs. I have a background in software but not in hardware side so a lot of this is learning as I go. I however like to spread my fixes to as many place as possible so a fellow enthusiast might find a solution or idea to their own fixes. I have posted some of these for example in arcade-musem forums but not everyone occupies every forum or google search picks up on everything.

I will try to contain single fix into a single post with photos. I'm happy to discuss on the fixes

 

[froztbite]

Game PCB: New Dynablaster - Global Quest / New Atomic Punk - Global Quest / Bomber Man World (different name for each region)


Symptom: No volume and low volume image is ok. Adjusting the volume slighly higher started to show pulsating image and when volume was high enough the image would stay fuzzy all the time.


Steps to find solution:

• Found a few other repair logs through google where all had one or multiple capacitor gone bad on this PCB. One repair log in particular had similar symptoms description and they replaced the largest one in the audio section.
• Desoldered the suspected bad capacitor and replaced with a new one.

Repair: Replaced a capacitor gone bad in sound circuit which was affecting image when volume was high enough.

Disclaimer: I replaced every capacitor just in case after I verified the fix. My tester at least reported one of the smaller ones to be faulty as well, but didn't notice any difference in the game behaviour after replacing it. This is not visible in the pictures.

 

[froztbite]

Game PCB: Sega ST-V Motherboard


Symptom: Sound was a mess which I can best describe that it had similar effect that volume would be turned all the way up and the speaker can't handle it but the same problem was with no matter what the volume was.


Steps:

• Found a repair log with similar issue description and it was resolved with replacing the TDA1386T (DAC). -> https://www.jammarcade.net/sega-st-v-bad-audio-samples/
• Used Audio Probe to check the output on the TDA1386T. The left channel output ( 18 ) was crystal clear but right channel output (21) had the same bad audio that I was hearing through the jamma connector.
• Desoldered the suspected bad chip and replaced with a new one.

Repair: Replaced a TDA1386T (DAC) partially gone bad in sound circuit which was ruining the fun.

Picture of the replacement done

 

[froztbite]

Game PCB: Golfing Greats


Symptom: Broken Konami 051550 custom chip was causing the game to not boot and it just reset constantly. Shipping company was not kind to my package and it had sustain a nasty bump.

Steps:

• Bought a replacement chip made by caius arcade from tindie -> https://www.tindie.com/products/caiusar ... placement/
• Removed the broken chip and replaced with the reproduction

Repair: Replaced a broken Konami 051550 custom chip with a reproduction replacement.


Bought this game in a lot of untested PCB from a fellow that had them stored away for around 20 years. Board looked to be in great shape and in the listing pictures the chip was still intact but shipping internationally has its risks.

 

[froztbite]

Game PCB: Crude Buster / Two Crude

Symptom: Bought this as with graphic glitches. Most of the graphics were working fine but a few sections had some weirdness.

Steps:

• Checked the Mame rom files documentation that the top part of the board were the graphics (GFX) roms
• Started to check if the roms were fine in the GFX section
• While the rom check was giving ok, the placement it returned didn't match the board layout
• Rearranged the top 4 GFX roms since they wrere in the wrong order

Repair: The cause of the graphic glitches were the wrong order of the GFX roms.

Following pictures are so that left is how I received it and right is after I fixed it. Actually most of the game looked correct and I was only able to spot a few difference on the first level. Probably more would have been found if I played it further. Sorry if the contrast or color is a bit off, all of the images have been captured with my phone camera.

First of the board how it arrived and how I rearranged it

 

[froztbite]

Game PCB: Dark Seal

Symptom: P1 movement were working but actions were not (Attack and Magic)

Steps:

1. Traced the malfunctioning controls to Data East resistor array looking chip RCDM-l105
2. The voltage drop when pushing buttons left it in around 3v.
3. Futher inspection showed that the action buttons had almost zero resistance when compared to the common ground.
4. Checked with a working RCDM-l105 that the resistance was 1K between the ground and the other pins.
5. Replaced with a "10pins 1Kohms Bussed" resistor array
6. Not working tried to find the difference between the two component and found a really good breakdown from a bit newer similar Data East Component RCDM-l107: http://www.dragonslairfans.com/smfor/index.php?topic=7526.0
7. Seems like every other pin is connected by a smaller resistor instead of the common ground. So for example 2. pin is connected to the jamma edge for attack button and 3. pin is connected to the board logic for the P1 attack and those two are connected by a smaller resitor.
8. Cut every other leg of the off the shelf resistor array and connected with smaller resistors.

Repair: Replaced custom Data East RCDM-l105 with off the shelf "10pins 1Kohms Bussed" with some modification. According the other repair log (much smarter person than me) there was also a diode inside the custom component but it was more for a fail safe and not required for the component to function. As far as I can see the RCDM-l105 and RCDM-l107 do not have much difference so the might even be work the same in functionality.

As I understood from the other repair log that failure of the RCDM-l105 is most likely connecting the jamma connector the wrong way. I bought this PCB on the cheap from Ebay where the seller claimed the PCB was working fully. While the game and P2 was working fully, but with P1 you could move around and coin in the game it would have been unplayable if you played with longer than the starting area.

Attached picture of the final repair on the pcb and how I understood the wiring is now.

 

[tiff_lee]

Nice logs mate, good stuff.

Someone must of been messing around with that Crude Busters and had a dose of amnesia mid repair or whatever it was they were doing.

 

[froztbite]

The Crude Buster was weird for sure since all of the ROMs look to be from factory that they haven't been replaced or anything. My best theory was that a friend of the seller pulled prank and forgot to tell them . It is a decent game and fairly inexpensive. I got it for 50€ with the GFX issue disclaimer but have seen it go well above 100€ as working.

 

[froztbite]

Game PCB: Tekken (1)

Symptom: Player 4 inputs (player 2 kicks) shorted to FF00

For background information I mostly used my SuperGun for testing and I have Tekken 2 (ver B) and Soul Edge (ver. 2) which do not have issues so I ruled out problems with the SuperGun and harness. Also tested with just Jamma connector attached and it had no difference.

I got the original Tekken for cheap that was claimed to be working perfectly but it has had some repairs done (video and audio section) to it prior and the kick harness had not been tested properly since both players had issues.

Player one had a cut trace on the kick 1 that I repaired.

Player two couldn't use either of their kick since the 4 player controls are shorted and the kick buttons are borrowed from them. Since the kick is always pressed it also would mean that almost with every character the punches that did work would cause throw moves.

In the test menu you would normally see all four player controls and each one having 0000 when nothing is pressed, but the 4th player had FF00. I now spent several hours trying to trace where the short could have come from but haven't really made a breakthrough. There are two IR2C24AN IC that passthrough most of the 4th player controls. Both of them had leg number 10 (output 6) shorted to the ground and this could be verified by also testing out ground from edge connector and the where the output 6 was connected to in the edge connector.

On the first IR2C24AN it was connected A6 on the extended edge connector which according to Dunk Mania manual is 4th player Right which is unused in Tekken and I don't have it wired to anything.
On the second IR2C24AN it was connected solder side Jamma connector J which is coin counter 2.

By first trying to lift the shorted output pin from the first IR2C24AN I was able to clear the error to FC00 (not 100% sure of the exact code, but in the ballpark), this however did not unstuck the kick buttons.
By removing the first IR2C24AN completely the problem cleared enough that 4th player now has C000 when nothing pressed and now all the buttons for both players seem to work. Since only 4th player button 1 and 2 are required to work for the kicks this is technically now fixed enough, but I would still like to get to the root cause if possible.

Decided to remove the second IR2C24AN as well. Now the 4th player shows 0000 in test and everything seems to work as intended so I guess Tekken really didn't need them for gameplay.

Tekken manual:
https://www.arcade-museum.com/manuals-videogames/T/Tekken Operators Manual.pdf

Dunk Mania manual
https://www.gamesdatabase.org/Media/SYSTEM/Arcade//Manual/formated/Dunk_Mania_-_1995_-_Namco.pdf

Schematics of IR2C24N

One of the IR2C24N chips remove. The second one I removed was the one directly below in the picture.

I technically fixed the game to work, but I'm still pretty beginner at reading schematics and there are still a lot I didn't understand so if someone else has more experience with Namco board repairs would love to hear the answers.

- Why even without the chip the buttons still work? My reasoning that since the Namco 11 is a motherboard for multiple different games it must be use to some of them but Tekken doesn't need it? I did find that Point Blank games were also for the Namco 11 and they have an extra PCB layer with connectors for the gun controller. Perhaps the Gun borrows the 4th player inputs? I do not have these games so can't really try and trace them.

- Anyone know of good schematics of the Namco 11 board or had written up on them? I tried to trace where the control inputs where going but could only find a few points and the end trace was to the IR2C24 output for the 4-player controls and few others like the coin counters.

- It is probably unlikely that both IR2C24AN would have gone bad and both have the exact output pin shorting out. So I'm guessing the problem must have been somewhere further down the line but I couldn't find where the would connect other than ground and power, but nothing else is shorted out. There was a nasty looking repair near the audio section were a diode had burned up so might be related what caused these issues. Can take a picture if that helps.

 

[froztbite]

Game PCB: Fever SOS / Dangun Feveron

Symptom: Settings were not saving and on some full power off and on boot ups a very specific settings were loaded

Recently bought a Fever SOS board with Dangun Feveron ROMs installed. The game was sold without issues and but noticed that on some boots the default settings were loaded but sometimes the game would load a very specific settings that I had not set. These settings included inverted screen, hardest difficulty, only one life and no continues. Going into the settings with the test button and soft resetting the game would bring the default settings for the game but I couldn't make any changes that would save. For example I would want to set the autofire for C-button on but on every attempt it stayed off after leaving the settings menu.

Picture of the weird settings that were being loaded

While trying to replicate and test the issue on MAME I verified that I did do the correct procedure that the settings are saved by holding the test button while in the settings to get the game to reboot without powering down.
Reading on MAME documentation and other repair logs from games without dipswitch settings I determined that the settings were being write on a EEPROM chip 93C46. I went up and ordered some 93C46B-I/P for a replacement from mouser.

Picture with the suspected bad chip surrounded by red square

The chip came out pretty neatly

Decided to install a socket just in case with the new chip if this needs to be replaced again in the future.

Repair: Replaced EEPROM chip 93C46 used for storing settings. The chip had gone bad and would not save settings and sometimes either load the default ones or load up weird settings that I had not set up. After the new chip was installed the settings were saved correctly.

 

[froztbite]

Game PCB: Tekken (1)

Symptom: Blue color was missing

I got a second Tekken in a bundle with Tekken 3 with knowing that it had the color issue. Figuring it would be an easy fix.

Probing around the game with a multimeter when it was running the blue on JAMMA edge had close to zero voltage. So turned down the power, switched to continuity mode in the multimeter and started probing if there was a cut trace somewhere. To my surprise it was right next to the JAMMA edge so only a short wire was needed to fix the connection.

Seems like these boards are really susceptible for cut traces near the JAMMA edge .

Picture of the final repair done

Then a few pictures were comparing the working game before the repair was done. On the left is a working version and on the right the blue color is missing. Some of the before repair pictures look a bit more grayish when they were actually more of an yellow tint.

Title

Demo

Test

 

[froztbite]

Game PCB: Battlantis


Symptom: Game was constantly resetting on boot up. Some garbled image could be briefly seen before reset.


Bought this game knowing it wasn't booting up but the board was really clean looking and the price was right so thought that I could do with a challenge.

Started with the easiest part of checking all the socketed ROMs and all read fine and matched with MAME.
Decided to check voltages from JAMMA edge and some IC on the board and all were seemingly normal.

Found out that there are two pads marked JP near the JAMMA edge which when connected prevents watchdog circuit from resetting the board on error. For some reason I had tough time soldering there so I traced the pads to the long resistor array looking IC next to it and connected the circuit there. The board was no longer resetting but the game didn't go further from the garbled image. Didn't really spot any difference with IC signals around the board.

With no real leads took out the logic probe to check what was high (5v), what was low (0v), what was pulsating and if I'm lucky find pins that are none of the above which might mean that something wrong there.

Decided to hone in around the main CPU (63C09E) and some 74-series next to it. Found several pins that showed no activity.
- The main CPU (63C09E) had pins 17 (Address bus 9) and 32 (Read / Write output) silent. The pin 17 was going to the ROM at 8E described as program code.
- Data selector/multiplexer (LS157) had multiple pins not having any signal. The 'c' and 'd' inputs and output were silent
- Octal Bus Transceivers (LS245) had one pin either 19 (O/E) or 18 (Channel 1, b-side) that was not giving any activity.

Dreading that I would have to de-solder a 40-pin main CPU I decided to try out the the smaller logic IC first. Examining the LS157 closer it seemed that the pins that showed no activity might not actually be used or connect to anything. There have been instances where the pins connect under the IC and then continue either to the solder side or the trace comes out between some other leg. After I got the LS157 there was no mistaking it the legs that were not giving any signal were not connected to anything else on the board so this was checked off as a possible culprit.

Time to move on the second smallest LS245. With this I found continuity to another IC on the leg so I figured that it should probably be either high or low. After de-soldering, installing a socket and swapping between the original and an another the leg that previously gave no signal for the logic probe suddenly there was activity even with the old LS245. So either I fixed a cold solder joint when installing the socket or I just didn't get a proper reading the first time. Anyway this didn't help with the reset problem either.

So it was down to the main CPU (63C09E). Now that I had a warmed up on removing the smaller IC I was ready for the big one. The de-soldering process was pretty painless but one of the solder pads on component side came loose. Luckily it didn't tear so I was able to solder it back on before installing a socket on top of it. When soldering the socket I started with the leg that had the pad tearing and checked continuity before proceeding with rest of the legs.

With the new CPU in place in the socket it was time to test it out. On power up I was greeted with a familiar garbled image and was already feeling a bit deflated about another defeat, but a moment later the image changed and after ROM check giving all OKs the game booted up. It lives again!

On a side note I actually did the initial troubleshooting and bough the replacement CPU a year ago but decided I wasn't ready for de-soldering 40-pins without breaking something . Now with a bit more experience and better tools I had more confidence to attempt the repair. Year ago I was using a manual solder sucker but now with HAKKO FR-301 even the bigger de-soldering jobs are not so time consuming.

Picture of the whole board:

Close-up of the repaired area

Snippet of the 63C09 datasheet