New Life Games LLC
NLG Users Repair Logs and Other Ramblings. Request your very own topic. Just ask any site staff. => Channelmaniac's Arcadecomponents' Old School Repair Logs => Topic started by: channelmaniac on May 07, 2014, 06:26:18 PM
-
Model: Rally Bike
Symptom: Garbage video - dead
The screen was black with a small section of garbage characters on it. Board would make the coin up sounds when the coin mechs were triggered. Looked at the board carefully and found a corroded pin on a capacitor. This corrosion went down into the PC board and ate an internal trace. This disconnected power from 2 ROMs. Removed the cap, cleaned up as much of the damage as possible, and ran 2 jumper wires to fix the game.
-
Model: Renegade
Symptom: Dead
Board was dead - garbage on screen. Replaced the 68A09 CPU and the game would run but glitch and lock up. Replaced the 6502A CPU and the game would play but had no music. Voice worked fine. Replaced a bad op-amp in the audio amplifier section and played a game to test it.
-
Model: Hishou Zame (Japanese Sky Shark)
Symptom: Player shadow, enemy plane shadow, and explosion graphics are corrupt
Replaced a bad ROM at location B02-04 and tested.
-
Model: F3 main board
Symptom: Makes a loud noise
Replaced bad fan and tested
-
Model: F3 - Bubble Memories
Symptoms: Locks up and resets. Audio is not correct.
Cartridge was a converted cartridge. Replaced 4 slow EPROMs and 2 bad EPROMs. Replaced 1 incorrect PAL chip and tested.
-
Model: F3 - Bubble Symphony
Symptom: Convert from Japan region to USA
Erased and reprogrammed the rightmost EPROM to change the region and tested.
-
Model: Bubble Bobble
Symptom: Constantly resets
Replaced bad EPROM at A78-08. Added jumpers at JPA and JPB and tested.
-
Model: F3 - Bust-a-Move Again / Puzzle Bobble 2
Symptom: Garbled graphics
Background and text graphics were good. The motion graphics were corrupt. ROM E10-01 had a hole in the top and E10-02 was bubbled up. Replaced them with 27C800 EPROMs programmed with the code from the MAME dumps. Reprogrammed E10-09 to convert the game from Japan region (Puzzle Bobble 2) to the US region (Bust-a-Move Again) and tested.
-
Model: F3
Tip: Convert games to different regions
The F3 system uses keyed cartridge connectors for different regions. To change regions on the game requires either the modification of the connectors, removing the keys, or reprogramming of the EPROMs.
To reprogram the EPROMs, simply change the last byte of the last Program EPROM on the right: 01 = Japan, 02 = USA, and 03 = Europe
-
Model: F3
Tip: Making a multi-region cart
CAUTION: Be careful wiring the switch! If you miswire it you could create a dead short between +5v and ground which could damage the cartridge and/or game board. DO THIS AT YOUR OWN RISK.
Some carts look MUCH better in Japan region than they do in the US region: Puzzle Bobble 2 is an example of one. To change the region one must change the last byte of the right most EPROM so to make a dual region cart, simply use an EPROM that is 2x the size and copy/edit the code.
The Puzzle Bobble 2 cart uses a 27C020-12 EPROM. Read the right most EPROM into your programmer. Copy the code region from 00000 to 3FFFF and paste it into region 40000 to 7FFFF. Next edit the byte at 7FFFF and change it to 02 for the US region. Program it into a 27C040-12 EPROM and verify it. Next, mount a SPDT switch to the top case of the cartridge. Wire one outside contact to +5v and the other to ground. Take the EPROM, bend pin 31 out straight from the body of the chip and trim the skinny part of the leg off. Solder a wire from the uppermost address line which, in the case of the 27C040, is pin 31 to the center contact of the switch.
To run one region, flip the switch then power up the unit. To change regions, turn the game off then move the switch to the other position and power it back up.
DO NOT flip the switch while power is on. It could crash the game - but won't cause damage. You WILL need to reset the backup data when changing regions, just like you would if you switched carts on the F3 system.
-
Model: Bubble Bobble
Symptom: Stuck in reset
Replaced bad EPROM at A78-8 and bad Z80 CPU at location 36. Installed jumpers at JPA and JPB to change coin up functions. Board now works but has missing audio. Replaced physically damaged caps in the sound section and tested.
-
Model: F3
Symptom: Missing Video RAM chip
Board was missing a surface mount IC and had attempted repair damage. Rebuilt 4 pulled pads, soldered in a replacement video RAM IC at location IC8, and soldered one jumper wire. Replaced dead cooling fan and tested.
-
Model: F3
Symptom: Up control doesn't work
Game does not register an up motion on the joystick. Checked the pin on the JAMMA edge and found it was not tied high, but was grounding properly when the stick was pushed up. Replaced a bad 390 ohm resistor pack at location RB17 and tested.
-
Model: Bubble Bobble
Symptom: Sound intermittant and graphic corruption
Text characters were wrong color and some were replaced by colored blocks. The Bubble Bobble title screen was corrupt. Replaced bad EPROM at A78-15 and 3 bad capacitors and tested.
-
Model: Battle Shark
Symptom: Corrupt graphics
Fixed bent pin on IC29. Reflowed IC29 to fix smeared solder issue. Replaced a cracked RAM at IC120 and tested.
-
Model: Twin Cobra
Symptom: Stuck on self test screen
Board would not boot past self test screen. Custom protection chip "GXC-03" was bad. Swapped out the USA ROMs and the matching "GXC-04" protection chip from a parts board and tested game.
-
Model: Light Bringer
Symptom: Flickering graphics with horizontal lines on the screen
Repaired broken ground pin on the PAL chip at IC62. Tested.
-
Model: Renegade
Symptom: Dead. Garbage on screen
Replaced defective ribbon cables between the 2 boards. Tested.
-
Model: Bubble Bobble
Symptom: Dead. Stuck in reset.
Replaced bad PAL at location IC43. Tested.
-
Model: Bubble Bobble
Symptom: Dead. Stuck in reset.
Replaced bad ROM at IC50 and IC51. Replaced bad PAL. Replaced bad 74LS157 at location IC40 and IC13. Replaced bad 74LS245 at location IC54. Tested. Board booted with an I/O Error. Replaced cracked PC030CM hybrid module. Tested.
-
Model: Bubble Bobble
Symptom: Dead. Stuck in reset.
Replaced bad CPU socket. Replaced bad 74LS245 ICs at IC28, IC29, IC54, and IC18. Replaced bad 74HC245 at location IC59. Installed jumpers at JPA, JPB, and JP6. Tested.
-
Model: Bubble Bobble
Symptom: Plays blind
Board had nothing on screen except some random splashes of color. Replaced a bad 7LS374 at IC58 on the video board to return the background and text. There were no characters and no blue colors. Replaced bad 74LS163 at IC25 to fix character problem. Replaced bad PC040DA hybrid at IC85 to fix the lack of blue output. Tested.
-
Model: Bubble Bobble
Symptom: Distorted music
Sound effects were normal, but music was distorted. Recapped audio section and replaced 1 bad TL074 Op Amp. Tested.
-
Model: Cadash
Symptom: Dead
Replaced missing TC0070 hybrid. Board reported bad Work RAM. Replaced IC16 and tested.
-
Model: Bubble Bobble
Symptom: Dead. Stuck in Reset
Replaced bad ROM at IC52. CPU 1 would continuously reset when trying to read the Work RAM. Replaced bad 74LS245 at IC52 and the board booted to an IO ERROR message. Replaced cracked PC030CM hybrid module and installed PC style jumpers at JPA and JPB. Tested. Player 1 would not move left. Cleaned edge connector and retested successfuly.
-
Model: Final Blow
Symptom: No sync and corrupt graphics.
Cleaned board and edge connector. Resoldered the TC0070 SIP and fixed 2 bent pins on the SMD custom ICs. Tested.
-
Model: Rastan
Symptom: Static in audio. No sampled sounds
Recapped audio section, replaced missing volume pot, and replaced a bad TL074 op amp at location IC106. Tested.
-
Model: Bubble Bobble
Symptom: Will not boot
Board displayed garbage on the screen. The main CPU was not booting properly. It would pass RAM tests with the Fluke but would not boot. The data lines and address lines on the shared RAM chip were missing signals when under CPU control and the display would intermittently lose sync.
Replaced the 60 pin ribbon cable to fix the sync issue. Replaced IC56/IC57, 74LS244 bus buffers to fix the address line issues and IC54/IC55/IC18 to fix the shared ML Data Bus issues. Set the game for English mode, replaced 2 board standoffs, and played a few games to test.
-
Model:Sky Shark
Symptom: Buttons stuck
Replaced 2 cracked 4.7K resistor packs in the input circuit. Tested.
-
Model: Arkanoid
Symptom: Blue screen and no sync.
Cleaned edge connector and tested.
-
Model: Bubble Bobble
Symptom: Dead. Garbage on screen.
Shared memory bus was stuck on 5 of 8 bits. Replaced bad 8K x 8 SRAM IC and 2 bad 74LS245 ICs. Bus was OK but CPU would not start. The program EPROMs had the correct checksum but the main Z80B CPU was bad. Replaced CPU and tested.
-
Model: Bubble Bobble
Symptom: Dead
DIP Switch 3 was turned on leaving the game in Pause mode and causing it to be stuck in watchdog. Flipped the switch and the board would boot but get into a crash loop with repeating game sounds and garbage on the screen. Replaced a bad 6264 SRAM IC and tested.
-
Model: Jungle King
Symptom: No sound, corrupt graphics, and crashes. Will not read DIP switches.
Replaced 2 bad AY-3-8910 sound ICs at IC50 and IC51 and EPROMs at KN37 and KN38 to fix the sound and DIP switch issues. Replaced EPROMs at KN26 and KN27 to fix graphic corruption and crashes. Tested.
-
Model: Bubble Bobble
Symptom: Dead
Reseated loose ribbon cable between boards. Board booted but had no sync. Cleaned edge connector and tested.
-
Model: Bubble Bobble
Symptom: Stuck in reboot loop.
Replaced bad A78-08 ROM and tested. Board had no player 2 buttons and incorrect colors on the playfields. Cleaned edge connector and replace a bad A78-15 ROM and retested.
-
Model: Bubble Bobble
Symptom: Low audio output
Replaced 2 x 100uf, 1 x 220uf, and 3 x 10uf electrolytic capacitors in the audio circuit. Replaced a physically damaged filter cap on a corner of the CPU board. Tested
-
Model: Bubble Bobble
Symptom: Dead
Cleaned the CPU board, replaced corroded sockets on IC12, IC17, IC37, IC49, IC51, and IC52 and cleaned the pins on the ICs before reinserting them. Game would now play blind. Signal 0A0 was missing, causing all objects, backgrounds, and text to be missing. Replaced IC20 (74LS163) on the video board as pin 14 was stuck. Board had video but character would not move to the left. Resoldered broken solder joints on several coils on the input circuits and tested the game.
-
Model: Bubble Bobble
Symptom: Garbled screen
Game had sync to monitor but graphics weren't synchronized to the screen and were all scrambled.
Screen would intermittently clear up. Traced it backwards through the circuit to the SY0 signal coming off of pin 11 of IC75 on the video board. Replaced the chip and had the same results. Put a multimeter on pin 11 of IC75 and found it to be shorted to ground through 58 ohms. Used an ESR meter to find the lowest resistance point to ground on the circuit and found a bad 74LS74 with shorted inputs at IC61. Replaced the IC and tested.
-
Model: Bubble Bobble
Symptom: Dead
Replaced the 74LS245 latches on the shared RAM bus and tested
-
Model: Bubble Bobble
Symptom: Random video glitching
Replaced bad 74LS374 IC at IC77 and tested.
-
Model: Bubble Bobble
Symptom: Dead
Board had no clock signals. Replaced a bad 74S74 at IC88 to restore the clock. Board booted and played but had flickering graphic corruption. Replaced a bad 74LS374 at IC94 and 74LS373 at IC95 to fix. Tested.
-
Model: Violence Fight
Symptom: Screen RAM Error
Replaced bad 62256 RAMs at IC10 and IC20. Game would boot and play but had corrupt player and title graphics. Replaced a corroded socket at ROM C16-04, patched a bad ground trace, and cleaned the ROM legs. Tested.
-
Model: Championship V'Ball
Symptom: No sound
Replaced physically damaged volume pot and one loose electrolytic capacitor. Board booted but was out of sync. Cleaned the edge connector with a fiberglass pencil and successfully tested the game.
-
Model: Bubble Bobble
Symptom: Dead
Cleaned the dirty edge connectors with a fiberglass pencil. Cleaned the .156 power headers. Cleaned dirty legs on 3 EPROMs and the 3 PAL chips on the CPU board. The logic probe revealed that IC58 and IC18 had stuck pins between them. The ICs and the 6116 RAMs tested good out of circuit. When flipping the DIPs the board suddenly came to life, but would quickly crash. Probing the outputs of the 74LS253 chips would cause the board to come back to life. Replaced all 4 74LS253 ICs on the input circuits and tested successfully.
Note: The 74LS253 chips were connected to the pins on IC58 and IC18 and were the cause of the stuck pins.
-
Model: Bubble Bobble
Symptom: White screen
Replaced 2 missing standoffs, installed new jumper at A and missing jumper at B, and cleaned the edge connector. Removed excess solder blobs from the power connector. Tested.
-
Model: Elevator Action
Symptom: Lines on screen, corrupt graphics, and locks up
Cleaned corroded ROM legs on the CPU, ROM, and sound boards. Replaced bad socket on IC54 on the CPU board. Replaced bad video RAM at IC106 and IC108 on the video board to fix the graphics issues. Board played, but the player and enemy sprites only showed the rear half of the character, doubled. Replaced a bad 74LS86 at IC86 on the video board to fix the sprite issue as pin 11 (output) was dead. Replaced a broken poly/foil cap on the sound board and tested.
-
Model: Elevator Action
Symptom: Red screen with some garbled characters instead of text.
Replaced 2 bad 40193 ICs on the ROM board. Replaced 2 broken poly foil caps and tested.
-
Model: Elevator Action
Symptom: Stripped. Parts board. Please rebuild.
A parts board for a local museum was revived - it was severely stripped because it was dead.
Installed the following:
19x 2732 EPROMs
1x 24 pin socket
1x 93419 Bipolar RAM
6x 2114 RAMs
4x AY-3-8910 sound chips
5x .1uf poly film "chicklet" caps
This board didn't have the special encrypted CPU board so I programmed the ROMs with a mix of elevator and elevatorb MAME ROM sets. The elevatorb MAME set was for the unencrypted version.
Replaced a shorted chicklet cap on the CPU board to revive the game. Cleaned the power connector and the edge connector and tested.
-
Model: Elevator Action
Symptom: Stripped. Parts board. Please rebuild.
Another parts board for a local museum was revived - it was severely stripped because it was dead.
Installed the following:
19x 2732 EPROMs
1x 93419 Bipolar RAM
4x 2114 RAMs
6x 6116 SRAMs
4x AY-3-8910 sound chips
2x Z80A CPUs
This board didn't have the special encrypted CPU board so I programmed the ROMs with a mix of elevator and elevatorb MAME ROM sets. The elevatorb MAME set was for the unencrypted version.
Cleaned the power connector and the edge connector and tested. Dead. Swapping boards showed the CPU board was the culprit. Reprogrammed ROM at IC69 and tested successfully.
-
Model: Bubble Bobble
Symptom: Resets
Replaced 3 badly corroded ICs, cleaned the edge connector and flipped all the DIPs to see it reset. That excluded the matching 74LS253 chips as a cause. Noticed the graphics stuttering before the reset line was activated by the watchdog. Checked the microcontroller and found its reset line was pulsing during the stuttering. Checked backwards through the circuit to the LS273 at IC55 and found that pin 11 (Clock) was floating. This was causing the LS273 to output spurious data. Replaced bad 74LS139 at IC39 to fix the reset issue. Replaced 2 corroded caps and a bad TL074 Op Amp at IC21 to fix the missing music. Tested.
-
Model: Bubble Bobble
Symptom: No sound, 1/2 screen
Replaced bad 74LS163 in the vertical clock divider circuit to fix missing vertical sync and restore the screen. Replaced a bad TL074 in the audio section to restore sounds. Tested.
Note: The YM3014B audio DACs will show stuck outputs if the audio feedback loop through an external Op Amp is bad.
-
Model: Bubble Bobble
Symptom: Dead
Replaced a bad 74LS273 at IC55 and LS245 at IC56 on the main CPU and 4 other LS245s on the CPU busses. Board would boot and crash. Resoldered broken connections on the hybrid and diagnosed the MPU as bad. Customer had a spare on another board and still wanted this boardset fixed.
On the video board, replaced a bad RAM at IC74 to fix vertical lines issue. Replaced a bad 74LS257 at IC55 to partially restore the crosshatch test screen. Replaced bad 74ALS169 chips at IC65 and IC48 to finish fixing the screen display. Board would play, but had graphical corruption. Chased this one around a bit to discover the IDT6116-45 SRAMs have slightly different timing than the Sony CXK-5814-45 SRAMs. Replaced IC74 with a Motorola 2018-45 SRAM to fix the corruption.
Game would play but would lose sound. Chased this one around for a long while to discover it was simply a bad ribbon cable. Tested.
-
Model: Double Dragon
Symptom: No video. Missing cable
Replaced missing ribbon cable. Cleaned very dirty edge connector to restore video and sound. Replaced 2 physically damaged 100uf capacitors and played the game through to the end to test.
-
Model: Empire City 1932
Symptom: Dead. Missing ribbon cables.
Replaced 2 missing ribbing cables and tested. No sprites. Replaced a Fujitsu 74LS283 with dead outputs at V12 to fix the sprite issue. Backgrounds were off in color. Replaced bad Fujitsu 74LS174 at A7 that had one dead output. Replaced 2 missing PCB standoffs and tested game.
-
Model: Elevator Action
Symptom: Dead
Removed the encryption CPU board. Installed the 5 ROMs from the elevatorb.zip file to convert the game to unencrypted ROMs. Replaced a bad 6116 SRAM chip. Repaired several solder smeared traces and one broken trace. Tested.
-
Model: Elevator Action
Symptom: Dead
Replaced IC3, 74LS32 to fix a sync issue. Replaced IC94, a 74LS32 to fix an issue with the CPU stuck in /WAIT and getting reset. Board booted but had graphic and text corruption. Replaced EPROM at IC64, 74LS32 at IC42, and another LS32 to fix most of the graphic corruption. Player sprites were divided into every other line and were placed horizontally in 2 different places on the screen. Replaced bad 74LS32 at IC6 on the video board and tested.
-
Model: Elevator Action
Symptom: Graphic corruption
The corruption made the screen look like an Atari 2600 - stretched out and blocky vertically. Horizontally, everything was OK.
Installed 10 ROMs to deconvert the PCB from a multi-kit to fix missing sounds. When the kit was installed, only the music played and all other sounds (shooting, dying, sliding down rope, etc...) were missing. Replaced bad 74LS173 at IC5 to fix the graphic corruption issue.
-
Model: Rastan
Symptom: P1 start not working
Cleaned edge connector and resoldered a broken solder joint on a capacitor in the audio section. Tested.
-
Model: Bubble Bobble
Symptom: Dead. Garbage on screen
Board had attempted repair damage and damage where someone removed the hybrid module.
Replaced bad PAL at IC 43. Replaced 6 corroded sockets on the video board. Patched damaged trace under the ROM socket on the main CPU. Board booted to I/O error. Installed missing hybrid and patched multiple damaged traces. Replaced 3 sockets in the audio section, patched another bad trace, and recapped the audio section to fix intermittent audio issues. Tested.
-
Model: Frontline
Symptom: Dead. Garbage on screen.
Repaired 3 smeared solder shorts on the back of the CPU board. CPU was stuck on /RESET. Replaced a broken 10uf tantalum and 10K ohm resistor in the reset circuit on the top board and tested.
-
Model: Rastan
Symptom: Dead
Replaced 1 bad Work RAM and patched 2 gouged traces. Tested to find graphic corruption and no FM music or sounds. Patched a dead trace between a 74LS244 buffering the ROMs to the PC0090 custom to fix the graphics corruption. Recapped the PCB and replaced a bad TL074 op-amp to fix the FM music and sounds. Tested.
-
Model: Frontline
Symptom: Dead. White screen
This happens when the Z80 cannot boot at all. The protection board socketed ICs were corroded - Z80, Protection CPU, and EPROM. Removed the protection PCB, cleaned the Z80 pins, installed the Z80 onto the bottom PCB where the Protection PCB plugged in, and converted the game to Elevator Action. Tested.
-
Model: Cadash
Symptom: Dead. Nothing on screen.
Board had a broken/missing TC0070RGB hybrid module and solder shorts where it should've been installed.
Cleaned up the broken pins and solder shorts from the PCB. Inspected and found one broken trace for the hybrid module and 2 lifted pins on the corner of the custom next to the module. Straightened one pin and reflowed the lifted pins on the custom. Installed a replacement hybrid and patched the broken trace. Tested.
-
Model: Operation Thunderbolt
Symptom: Dead
Created a JAMMA adapter for the game and it was dead.
Cleaned the Program ROM legs and tested. Replaced 9 capacitors in the audio section and 2 broken filter caps and retested.
-
Model: Rally Bike
Symptom: Dead
Found a dead internal board trace on pin 28 (+5v) of a 8k x 8 main CPU RAM. Jumpered the trace and tested. Board had scrambled colors and no audio and would make popping sounds when tapped. Resoldered multiple cracked solder joints and one factory unsoldered capacitor leg in the audio circuit and multiple bad factory solder joints in the video section to fix the issues. Tested.
-
Model: Elevator Action
Symptom: Missing sprites
Board had attempted repairs done to it.
Sprites were hacked up into 1 bit wide vertical lines and were all over the screen.
This one was a tough one to figure out. Ran through the previous work and found one EPROM with a leg folded under. The problem was with the sprites, but not in the sprite section, but rather right off the SYNC buss on the Video Board. IC39 pin 11 was stuck low. This affected a lot of signals throughout the video board: /LNCL2, /LNCL1, /LNLD2, /LNLD1, /LNSL1, /LNSL2, /PHA348, HBLM, 128HM, 64HM, 32HM, and 16HM.
Replaced IC39 and tested.
-
Model: Space Invaders Part II
Symptom: Dead. Nothing on screen.
Pulled ROM H and the board displayed horizontal lines properly. Replaced 4 bad 7408 chips buffering the CPU address lines. Now board had activity on the address lines but a test ROM in socket H was not being accessed. Replaced bad 7442 IC in the address decoding circuit. ROM was now being read, but data was not making it to the CPU. Replaced 2 8216 data latches and board booted the test ROM. Replaced one bad RAM chip and tested.
-
Model: Legend of Kage
Symptom: Dead
Board was missing 6 ROMs.
Programmed and installed 3 missing Program ROMs and 3 missing Graphics ROMs. Board booted to an IO ERROR message. Followed the traces for the hybrid and discovered it was installed backwards. Removed the hybrid, cleaned the flux from the PCB, and reinstalled the hybrid correctly. Tested.
-
Model: China Gate
Symptom: Graphic corruption
Socketed the four graphics Mask ROMs. Replaced one bad ROM and tested. Board had intermittent graphics issues that would come and go when the board was flexed. Jumpered a cracked power trace to another graphics Mask ROM. Tested.
-
Model: Majestic 12 (Super Space Invaders '91)
Symptom: Dead. Nothing on screen.
Replaced 2 corroded Program ROM sockets. Cleaned ROM legs and patched 2 corrosion damaged traces between the ROM sockets. Tested.
Game version was not in MAME. Dumped ROMs and submitted them.
-
Model: Elevator Action
Symptom: No sync
Patched a damaged trace to fix the sync. Game would try to boot and would crash. Replaced a 74LS08, 74LS139, and 74LS86 on the video PCB. Replaced 11 24-pin, 2 18-pin, and 1 40-pin corroded sockets on the CPU board. Game would boot and run attract mode successfully, but sprites were corrupted. Replaced 2 bad 93422 Bipolar RAMs and tested. Board was missing background audio and had bad thumping sounds in the speaker. Replaced 5 220uf and 1 22uf capacitors and a missing 50K volume pot. Tested.
-
Model: Cadash
Symptom: Corrupt graphics
Patched 3 gouged traces and tested.
-
Model: Double Dragon II
Symptom: Crashes when sprites are supposed to be on screen
CPU 2 (Z80) was dead. Power to the CPU was good. /RESET and Clock signals were present. IRQ and NMI were good. /BUSRQ was low, meaning the CPU was offline waiting for something to release control of the Address and Data busses to it. Traced the /BUSRQ back to a 74LS04 and the input to the gate was floating. Patched dead trace and tested successfully.
-
Model: Crime City
Symptom: No sound.
Sound CPU was stuck in watchdog. Replaced improper OpAmps, fixed solder shorts on the OpAmp sockets, replaced several cracked capacitors, and replaced bad YM2610 sound chip. Tested. P1 Up and P1 Button 2 were stuck on. Replaced TC0220IOC custom I/O controller IC and tested.