Taito G-NET¶
A Sony ZN-2 motherboard with two Taito boards grafted on, and one unusual idea: the game does not live on the cartridge. The PCMCIA cart is only a carrier — on first boot the board reprogrammes itself with its contents. Understand that and you understand every fault on the platform.
Identification¶
| Field | Value |
|---|---|
| Manufacturer | Taito |
| Years | 1998-2005 |
| Base | Sony ZN-2 motherboard (COH-3002T BIOS, “G NET” sticker) |
| CPU | CXD8661R @ 100 MHz (R3000A) |
| GPU | CXD8654Q @ 53.693175 MHz |
| Sound | Sony SPU + MN1020012A (Panasonic sound CPU, QFP128) on the FC board |
| Taito boards | FC PCB (sound + storage flashROMs) · CD PCB (PCMCIA interface) · optional Save PCB |
| PCMCIA controller | Ricoh RF5C296 (TQFP144) |
| Game media | PCMCIA cart — but the game runs from flashROM |
| PCB format | JAMMA (through the ZN motherboard) |
The mechanism that explains everything¶
At power-up the BIOS decrypts its bootloader from the flashROM at U30, then looks for a PCMCIA cart. Three cases — and they are exactly the three behaviours you see on a board:
- No cart → the Taito G NET logo appears with “SYSTEM ERROR”.
- Cart matching the flash contents → the game boots immediately.
- A different cart → the board re-flashes U27, U29, U30, U55 and U56 with part of the cart's contents. It takes two to three minutes, then the board resets and boots the game.
A three-minute boot is not a fault
That is case 3. A board that seems frozen after a cart change is re-flashing itself. Cutting power then leaves the flash half-written — and that is a real fault.
What it changes for preservation
The game lives in the board's flash, not in the cart. A board separated from its cart therefore keeps its last game; and a cart on its own is not enough to run a game if the board refuses to re-flash.
Sound: two configurations¶
Most games use the Taito Zoom hardware on the FC board. Five do not: Otenami Haiken, Otenami Haiken Final, Zoku Otenamihaiken, Zooo and Space Invaders Anniversary.
→ “No sound” therefore does not diagnose the same way depending on the game fitted.
What a G-Card manual says ✅¶
The Usagi manual — the G-Card, and what it demands
The 兎－野性の闘牌－ (Usagi) manual (Taito, © 2001, ref. G2002193A, 82 pp.) is the only G-Card manual archived here, and it is explicit about what it is not: "本製品「兎－野性の闘牌－」はGカードであり、マザー基板に装着 しなければ動作しません" — the product is the card, and the mother board is not included.
ⓘ The box holds seven items, and no mother board:
| Item | Part number |
|---|---|
| The Usagi G-Card | S0010025A |
| The operator's manual | G2002193A |
| Mahjong harness ×1 | S0010014A |
| Earth harness ×1 | (same plate) |
| Clamps ×2 | (same plate) |
| Instruction card | S0010023A |
| Display sticker | S0010024A |
ⓘ And the manual asks you to keep the box: "出荷時、Gカードは梱包材で保護してあります。梱包箱 は、Gカードを保管する時及び故障時の搬送用に大切に保管して下さい" — the packaging is for storage and for shipping it back for repair.
⚠️ Five ways to kill a G-Card, as the maker lists them
These warnings recur three times in the document — installation, settings, maintenance. They are not decorative:
- ⚠️ Static electricity destroys it. "Gカードは精密部品であるため、静電気によって破損 することがあります" — it is the first risk named, ahead of impact.
- ⚠️ Insertion and removal, DIP switches included, are done with the power off. "コネクター およびGカード等をゲーム（マザー）基板から抜き差しするときおよびDIPスイッチの切り換えを行なう ときは、必ず電源スイッチをOFFにしてから行なってください" — and on insertion, "注意して ゆっくり確実に挿入してください。無理に差し込むと故障の原因となります".
- ⚠️ Never into anything but a mother board. "Gカードはマザー基板以外の基板及びパソコン等に 絶対に挿入しないでください" — not a PC, not another board.
- ⚠️ The supply adjustment can destroy both boards. "スイッチングレギュレータの調整において 出力電圧を基準値より高くすると、ゲーム（マザー）基板およびGカードを破壊してしまいます". Setting the voltage above the reference value does not degrade: it breaks.
- ⚠️ The slot is kept dry and clean — "Gカード差し込み口を濡らしたり汚したりしないでください".
⚠️ And a clause that closes the door on improvised repair: "メンテナンスにおいて、タイトー純正 部品以外を用いた作業を行なったときは不正改造とみなし、修理等は行ないません" — work using non-genuine parts is treated as unauthorised modification, and repair is refused.
The error codes, as printed ✅
This is the most useful part of the document: the platform says what is wrong.
| Display | What it means | What the manual prescribes |
|---|---|---|
| COIN ERROR | ⓘ the coin takes 0.5 second or more to pass the microswitch inside the coin rejector | consult the cabinet's manual |
| SERVICE SW ERROR | the cabinet's service switch is stuck or broken, so it reads as held down · or its harness is shorted | consult the cabinet's manual |
| TILT | the tilt is active | consult the cabinet's manual |
| システムエラー / SYSTEM ERROR | ⚠️ the program can no longer be read from the G-Card | check the G-Card is fully inserted |
| LINK ERROR | the two boards are not correctly connected | check the wiring per the link kit's instructions |
| 通信に必要な以下の設定が違います | ⚠️ the two boards' settings do not match | change the settings named on screen so they agree |
✅ The first three point at the cabinet, not the board — that is triage information: faced with a COIN ERROR or a SERVICE SW ERROR, Taito tells you to go and look at the cabinet, not the platform.
And for the most disorienting fault — a machine that resets over and over:
| Suspected cause | Prescribed remedy |
|---|---|
| incorrect supply voltage | adjust the switch-mode supply |
| poor contact at the edge connector | ⚠️ wipe the connector pads with a clean soft cloth, lightly |
| open or shorted wire in a harness | replace the harness |
ⓘ Cleaning the edge connector echoes what this wiki publishes elsewhere: Atari prescribes a soft pencil eraser and forbids file and burnisher on gold contacts; Taito prescribes a soft cloth. Two houses, two gestures, ⚠️ the same exclusion of anything abrasive. → Traces, vias & reflow
Mahjong harness pinout, as printed
ⓘ Rare: few manuals publish the wiring of a mahjong panel. The table gives, for each signal, the four connector designators and the wire colour.
| 3P / 4P | K | J | L / R | Colour | Signal |
|---|---|---|---|---|---|
| 3P (1) | K (4) | J (5) | L (1) | white/orange | OUT 1 |
| 3P (2) | K (3) | J (3) | L (2) | white/red | OUT 2 |
| 4P (1) | K (7) | J (9) | L (3) | white/blue | OUT 3 |
| 4P (2) | K (11) | J (7) | L (4) | white/black | OUT 4 |
| 4P (5) | K (5) | J (4) | L (5) | white/yellow | IN 1 |
| 4P (6) | K (6) | J (8) | L (6) | white/green | IN 2 |
| 4P (7) | K (9) | J (11) | R (1) | white/grey | IN 3 |
| 4P (8) | K (10) | J (6) | R (2) | white | IN 4 |
| 4P (9) | K (12) | J (1) | R (3) | white/pink | IN 5 |
| 4P (10) | K (2) | J (2) | R (4) | white/brown | IN 6 |
ⓘ The document also publishes the panel matrix — which keys sit at which OUT row and IN column. ⚠️ It is not copied out here: its key labels are in tight Japanese, and a mis-transcribed matrix is worse than no matrix. It is in the archived document, page 80.
Notable games¶
Ray Crisis · Super Puzzle Bobble · Shikigami no Shiro · Psyvariar (and Revision) · Chaos Heat · Kollon · Zooo · Flip Maze · Go By RC · Otenami Haiken.
Common faults¶
| Symptom | Probable cause | Next step |
|---|---|---|
| “SYSTEM ERROR” under the G NET logo | no cart detected | Check the cart and the PCMCIA connector (RF5C296) |
| 2-3 minute boot | re-flash in progress — normal | Do not cut power: this is case 3 above |
| Board frozen after power was cut during a re-flash | half-written flash (U27/U29/U30/U55/U56) | Restart the re-flash with a valid cart |
| Cart not read | PCMCIA contacts / RF5C296 | Clean the contacts with isopropyl alcohol |
| No sound | Taito Zoom or the FC board's MN1020012A | ⚠️ five games do not use Zoom — check which one is fitted |
| No picture | ZN-2 motherboard (CXD8661R / CXD8654Q) | PCB diagnosis |
Repair (technician)¶
- Three stacked boards: a fault can sit on the ZN-2, on the FC (sound, flash) or on the CD (PCMCIA). Separate the layers before probing.
- The flash is the real storage: it wears and it corrupts. A board looping at boot with no cart points at U30.
- Shared ZN-2 base: for the motherboard part, everything that holds for a Sony/Capcom ZN applies.
→ Methods: PCB diagnosis · Recap
Procedure — a dead board, step by step ✅¶
What this adds
The general method lives in PCB diagnosis. This is its application to this platform: the same order, but with the values actually expected here, so you do not have to go back and forth.
Tools: multimeter, oscilloscope (or frequency counter), current-limited bench supply, desoldering braid. → Tooling
If the board is still in a cabinet
A CRT holds its charge with the machine unplugged. → CRT safety
1. Inspection, power off. Corrosion, cut tracks, oxidised EPROM legs, leaking SMD capacitors, dubious earlier repairs.
2. The connector, before anything else. A large share of "dead boards" are a contact problem, not a fault. Clean the contacts with isopropyl alcohol and reseat — not with an abrasive eraser on a gold edge, which takes the plating off and brings the problem back worse. It costs two minutes and settles the question.
3. Power, measured on the board. Aim for +5.0 V measured as close as possible to where power reaches the board — not at the PSU output: the drop is in the loom, and that is exactly what makes a board boot on the bench and not in the cabinet. Where you probe depends on this board's connector — JAMMA edge, JVS loom or a proprietary harness — plenty of makers kept one long after JAMMA arrived — see Identification above. → Wiring Check −5 V and +12 V too if the sound stage uses them. Suspected short: bring it up current-limited and find the hot spot.
4. /RESET. It must release after power-up. Held low = the CPU never starts. A 1–2 Hz loop is not the reset circuit, it is the watchdog: the CPU is crashing at boot, so look at the bus, not at the reset.
5. Clocks. There is no measurement table for this platform — MAME's sources yield nothing usable for it. Any frequency quoted under Architecture on this page serves as a marker; otherwise compare against a working board rather than against a number: an invented value condemns healthy parts. → PCB diagnosis
6. Bus. Activity on the main CPU's address lines tells you whether it is running or frozen. Frozen with good clocks and good reset → work RAM, bus buffers, ROMs.
7. It boots but the picture or the sound is wrong — that is another job: see Common faults on this page, then signal tracing.
Modernisation & mods¶
- Conversions: the platform converts through the cart — that is the whole point of the re-flash mechanism. → G-NET conversions
- ⚠️ Do not host ROMs — method only.
Emulation & FPGA¶
MAME emulates the platform (sony/taitogn.cpp), re-flash included: the description
of the mechanism above comes from that work.
Photos / assets¶
(To take: the three-board stack, the PCMCIA connector and its RF5C296, the U30 flashROM, a G-NET cart.)
Archived original documents ✅¶
Manufacturer document archived locally, found on Internet Archive on 13 August 2026. ⚠️ The item declares no licence — kept for preservation, like the rest of the collection, and its provenance is named.
- PDF — taito-gnet-usagi-manuel-jp.pdf — 兎 -野性の闘牌- 取扱説明書 (Usagi — Yasei no Toupai), 82 pp. The cover carries the TAITO G CARD logo, publisher Warashi Inc. and 株式会社 童, reference G2002193A. It is the manual of a G-NET game, and therefore the first document on this page describing the hardware in service. Via Internet Archive
Sources & attribution¶
Text (synthesised here; sources listed for attribution):
- MAME —
src/mame/sony/taitogn.cpp(BSD-3-Clause / GPL): the three-board make-up (ZN-2 + FC + CD, optional Save PCB), the COH-3002T BIOS, the complete boot sequence — bootloader decrypted from the U30 flashROM, “SYSTEM ERROR” with no cart, U27/U29/U30/U55/U56 re-flashed in 2-3 minutes when the cart differs —, the MN1020012A and RF5C296 with their packages, and the list of the five games without Taito Zoom hardware — https://github.com/mamedev/mame/blob/master/src/mame/sony/taitogn.cpp - MAME —
src/mame/sony/zn.cpp(ZN-2 motherboard clocks: CXD8661R at 100 MHz, CXD8654Q at 53.693175 MHz) — https://github.com/mamedev/mame/blob/master/src/mame/sony/zn.cpp
Review & corrections
How this space is used
Spotted a wrong value, an outdated procedure, a chip reference that does not match your board? Say so here, with what you observed (model, board revision, serial number, measurement). Every report is cross-checked against a source before anything changes — an unverifiable correction is published as “reported by …, not cross-checked” rather than silently applied.
Reading is open to everyone; posting requires signing in with Discord. Reports from the wiki's declared reviewers are handled first; anyone else's are read too, but go through a human before anything is changed.