WELCOME
ABOUT Pin4
ARCADE
COMPUTING
CUSTOM-ICs
FPGA Coding
HOMEBREW
REPAIR logs
LAB-TOOLS
LINKS
HAM radio site
This page is a database for issues with arcade boards, computers and general electronics and ways to repair them.
It assumes you have a certain expertise in this field, use this information at your own risk!
Use a proper work place with proper equiment and good ESD grounding. Otherwise the repair will cause further
damages - and this you do not really want, do you?
ARCADE repair
Phoenix (G.G.I.) arcade board
Symtoms:
The board does not start up and gives garbage on the screen and some undefined sound.
Rarely it did start up, but some characters on the screen where wrong, then it crashed after some dozens of seconds.
Analysis and repair:
- Disconnected the board connector so no voltage is applied on it.
- C29 (ceramic cap. 47nF) on TIM-8002 board had a broken leg, C3 (ceramic cap. 47nF) on TIM-8001 board was broken. Replaced both with new ones (no improvement).
- Measured the capacitive load between +5V/GND, I got out that the capacitors where quite good, so I did not replace them.
- Checked EPROM contents against some MAME ROM contents (all ok), removed them and put them in again (no improvement).
- Replaced 8085 CPU with NOS one (no improvement). Put back the old one.
- Set up a test program to check the RAM (writes it and reads it back, halts the CPU on fail - here are the sources for download, use at your own risk).
- Programmed an EEPROM and replaced ROM0001 on location IC45 on TIM-8002 (pinout against 2716 EEPROM required - used a simple adapter).
- Connected a logic analyzer to the 8085, checking the last RAM access before the HALT condition. The CPU did crash sometimes, but if the RAM test was able to run, it consistently indicated a read-back fail at the same adress. This address/data indicated a defective IC27 on the TIM-8002 board.
- De-soldered IC27, soldered a 18-pin socket, inserted a new IC (NEC uPD2114LC, SRAM 1k x 4bits, 450ns) I got from the usual sources.
- Board runs stable again, it seems the defective RAM also had a significant impact on the board supply which caused even crashes with the simple RAM test I set up.
COMPUTER repair
JOYSTICK / C128D "portable non-metal" version (probably also valid for C128DCR)
Symtoms:
Joystick on Port 2 (or Port B) did not allow to control to the left. Right, top, bottom and fire did work. Keyboard did work fine.
Analysis and repair:
- Measured pull-down current on the Joystick port. All shown about 1mA, pin 3 (responsible for left moevement) shown about 10mA.
- Removed the mains cable to have no high voltage on the device. Removed all other cables as well.
- Opened the C128D case (4 screws on the bottom, two screws on the mains connector - this allows to put the cover away completely). Disconnected and removed the floppy PCB (2 screws and a metal 'hook' at the middle of the PCB). Removed the floppy drive (4 screws to the bottom, 2 screws holding the front, pulled out the floppy lever on the front). Opened the 4 screws of the power supply somewhat to allow some space to remove the metal shielding of the C128 board (removed all screws, including the middle larger one connected to the video circuit shielding).
- The line from pin 3 (named JOYB2) is connected (over some EMI stuff) to PA2 (pin 4) of the IC U1 (CIA 6526). Measuring the voltage while moving the connected joystick (middle position: about 5V, left position about 0V) indicated that the connector and the PCB is ok and there was no short or so.
- Exchanged IC U1 with IC U4 (same type), it showed that the joystick was working again.
- On U4 the defect PA2 line has an impact on the user port only. As I don't use it I left it this way for now and ordered an used CIA 6526. Will be replaced when the new one arrives.
- Mount the metal shielding of the main board and re-mount the other parts in reverse order. The shield itself must be placed that it lies on the board by itself, all 'fingers' will touch the ICs below. Do not bend the shield somehow by force!
DISK / C128D "portable non-metal" version (probably also valid for C128DCR)
Symtoms:
The disk drive suddenly stopped reading disks. Happed in the middle of loading a disk. The head still moved and the disks where spinning (indicated by the usual sound), but it just looked that there was no disk inserted or the disks were not formatted.
Analysis and repair:
- Removed the mains cable to have no high voltage on the device. Removed all other cables as well.
- Opened the C128D case (4 screws on the bottom, two screws on the mains connector - this allows to put the cover away completely).
- Removed the metal shield of the floppy drive (two screws, one on each side).
- Noticed that the top head of the floppy was not really in plane. This was caused because the elastic metal holding the upper head was partly broken.
- When gently (and carefully!) pushing the head in place it was possible to read disks again.
- Ordered a replacement head at the usual sources. It did not arrive yet.
ELECTRONICS repair