Zaccaria/IGR Armada Game PCB Repair

Zaccaria/IGR Armada Game PCB Repair

Background

One incomplete instance of a game PCB titled "Armada" had previously been found and was thought to be a bootleg of Zaccaria Sea Battle. Some of the ROMs were missing, including the program ROMs, so it was unknown what the game really was. A fellow collector, Bela, encountered another complete IGR Armada game PCB that I was able to acquire to both archive a complete set of ROMs that were submitted to MAME and repair to running state.

13/07/2023 - ROM Dump

The Zaccaria/IGR Armada ROM set was dumped and submitted to MAME.

Zaccaria/IGR Armada ROM Set (25KB).

14/07/2023 - Armada game PCB PA20230714 repair

The physical condition of the PCB set was very good and complete. The main PCB was very similar to the Sea Battle main PCB and the sound PCB very similar to the Scorpion sound PCB, also IGR. There were twin pots and twin amplifiers, hinting that it could be a stereo sound game.

A side-by-side comparison illustrated the near identical hardware between Armada and it's predecessor Sea Battle. The Sea Battle matrix input & DIP switches had been removed from Armada to instead use conventional inputs via the 8255 on the top sound PCB following the same design as the Scramble-derived Scorpion platform.

Initial investigation with a multimeter found the power pinout matching the Konami standard. On first power on there was no video or sound but there was speaker hiss. Checking the supply voltages at IC 24 (2708) confirmed all three power voltages were correct. The scope showed active video sync output at IC 82 (2621) pin 1 but no active colour outputs. Checking IC 86 (2650) found:

pin 38 - active - CLK
pin 23 - idle lo - ~R/W
pin 24 - active - OPREQ
pin 26 - active - DB7
pin 27 - active - DB6
pin 28 - active - DB5
pin 29 - active - DB4
pin 30 - active - DB3
pin 31 - active - DB2
pin 32 - active - DB1
pin 33 - active - DB0

The main CPU was active but there was no response to coin.

Attention switched to the lack of any video output. Checking IC 24 (2708) "blue" found both address & data bus idle. Suspecting a problem with the video timing circuit, checking IC 82 (2621) found:

pin 1 - active
pin 2 - idle lo
pin 3 - active with poor level overtones
pin 4 - active
pin 5 - active
pin 6 - active
pin 7 - idle lo - GND
pin 8 - idle lo
pin 9 - active
pin 10 - active
pin 11 - active
pin 12 - active
pin 13- active
pin 14 - idle hi - VCC

IC 82 pin 2 was the VRST line that should have been active but showed as a marginal idle low at ~0.7V. Replacing IC 82 (2621) didn't result any video display but there were now random sounds being output and IC 24 address & data were active. The 2650 CPU reset line was also now toggling with watchdog resets. A subsequent power cycle brought the game to life with video, coin-up and sound all active (including the Scorpion "Thank You" speech on coin-up).

On the LCD monitor the video display was illegible. Switching to the CRT test monitor had a better display with a bad colours and a fault with the sprite display. The game also appeared to play OK, however right and left were oddly reversed versus the Konami standard pinout.

A scope confirmed the blue colour drive was missing from the edge connector video pin. Working back to C7,C8,C9 using a video probe & scope found:

C9 - blue - blank video - -0.3V -> +1.3v & floated down with scope lead
C8 - green - active video - -1.3V -> +0.3V
C7 - red - active video - -1.3V -> +0.3V

The signal at C9 didn't look correct and there was damage on the top of capacitor C9. I didn't see any significant differences between the Sea Battle & Armada video drive circuits so I suspected the video output was inverted. Using inverted video displayed correct colours (blue ocean, green land, and light blue sky).

14/07/2023 - Armada graphics repair

Moving on the sprite fault using a video probe starting with IC 24 (2708) found:

pin 1 - active
pin 2 - active
pin 3 - active
pin 4 - active
pin 5 - intermittent poor 0.7V signal
pin 6 - clock
pin 7 - clock
pin 8 - clock
pin 9 - active
pin 10 - active
pin 11 - active
pin 12 - GND

There were no schematics found for this game. Searching for the origin of pin 5 across the whole PCB was unsuccessful after several attempts, and I suspected the track may be open. Trying other pins as a potential hint on where pin 5 may originate found:

IC 24 (2708) pin 8 -> IC 56 (LS161) pin 14
IC 24 (2708) pin 7 -> IC 56 (LS161) pin 13
IC 24 (2708) pin 6 -> IC 56 (LS161) pin 12
IC 24 (2708) pin 5 -> ?
IC 24 (2708) pin 4 -> IC 17 (LS161) pin 11
IC 24 (2708) pin 3 -> IC 16 (LS161) pin 14
IC 24 (2708) pin 2 -> IC 19 (LS175) pin 2 (Q)
IC 24 (2708) pin 1 -> IC 19 (LS175) pin 7 (Q)

I suspected IC24 pin 5 as going to IC 56 pin 11 and noted that IC 56 had been replaced. Since Armada was very similar to Sea Battle, a comparison found that IC 5D pin 5 routed to IC 6L pin 11, equivalent to IC 56 pin 11 on Armada and confirming an open track. Visually and meter tracing the track found the continuity lost between R1 VCC and the top of IC 9 that was one long ~6" track. There were a couple of small blemishes on it near where capacitor C9 was also damaged. Scraping the solder resist and bridging across it fixed both continuity and the graphics fault. The game ran & played OK on the bench.

19/08/2023 - Armada colour repair

To keep the colours matched all three capacitors C7,C8,C9 (10uF/16V) were replaced. Overall contrast looked a better with deeper blue ocean and lighter blue sky.

20/08/2023 - Armada in cabinet

The game PCB on its original mounting plate fit perfectly in the Super Cobra cabinet and the Hantarex MTC-900 was switched over to the inverted video input option by changing a small plug on the chassis. Armada is a horizontal game, but this was good enough to be able to see how it looked and played on an original monitor.

The picture quality was good, and a few screen shots were taken.

20/08/2023 - Armada game play video

prswan@gmail.com