Midway T-Unit (Mortal Kombat II, NBA Jam…)¶
The Midway T-Unit hardware (early 90s): a TMS34010 (graphics processor) running Mortal Kombat II, NBA Jam, NBA Jam T.E. and a T-unit version of Mortal Kombat. Digitised actors, big sprites, and ADPCM sound (Williams board) or DCS depending on the game. A descendant of the Y-Unit (Smash TV, Total Carnage, original MK1).
Identification¶
| Field | Value |
|---|---|
| Manufacturer | Midway / Williams (WMS) |
| Years | ~1992-1994 |
| Type | ROM-board platform; TMS34010 CPU/video |
| Connector type | Midway harness (not JAMMA) |
| Settings | DIP + CMOS (coinage can come from either) |
Architecture¶
| Block | Chip | Clock | Role |
|---|---|---|---|
| CPU / video | TMS34010 (GSP, Graphics System Processor) | 50 MHz | CPU and graphics engine (integrated blitter, framebuffer) |
| Video | framebuffer driven by the TMS34010 | pixel ~4 MHz | big digitised sprites, wide palette |
| Sound (MK1 T-unit) | Williams ADPCM board: M6809 + YM2151 + OKI6295 | FM + samples | |
| Sound (MK2, NBA Jam) | DCS (Digital Compression System, ADSP-2105) | high-quality compressed ADPCM audio | |
| Settings/scores | CMOS/NVRAM + a DIP bank | bookkeeping, high scores |
The TMS34010 is a graphics processor that serves both as CPU and display engine (bit-addressable, internal blitter) — hence the big digitised-actor sprites typical of MK/NBA Jam. The sound evolves: the Williams ADPCM board (6809/YM2151/OKI) on Mortal Kombat T-unit, then DCS (ADSP-2105 compression) on MK II and NBA Jam — a real audio quality jump.
Notable games¶
Mortal Kombat (T-unit version), Mortal Kombat II (1993), NBA Jam (1993), NBA Jam Tournament Edition (1994). The earlier Y-Unit (same TMS34010 family) carried Smash TV, Total Carnage, Mortal Kombat (1992) and Terminator 2.
Settings (DIP + CMOS)¶
Unlike the Williams 6809 (100% CMOS), the T-Unit has a real DIP bank — but hybrid: a "Coinage Source" switch chooses whether the coinage comes from the DIP or the CMOS. Gameplay settings (Attract Sound, "Low Blows", "Comic Book Offer"…) are on the DIP; bookkeeping and scores in CMOS. → Transcribed table: DIP settings — Mortal Kombat.
Common faults & preservation¶
| Symptom | Likely cause | Lead |
|---|---|---|
| Lost settings / scores | Flat / leaking CMOS battery | Replace the battery; clean the corrosion |
| No sound (MK2/NBA Jam) | DCS stage (ADSP-2105) | DCS board diagnosis |
| No sound (MK1 T-unit) | Williams ADPCM board (6809/YM2151/OKI) | Sound-stage diagnosis |
| Corrupted sprites / page flipping | Video RAM / TMS34010 | RAM test, contacts |
| No image | Power, TMS34010, harness | Base diagnosis |
CMOS battery leak = corrosion
As on the Williams 6809 and many boards of the era: relocate/remove the CMOS battery to avoid trace corrosion.
- Emulation: MAME (
williams/midtunit.cpp) emulates the T-Unit. No FPGA/MiSTer core (the TMS34010 is hard to reproduce in logic).
Conversions¶
- ROM swaps between games on the same base (T-Unit) are possible in the collector scene; mind the sound variant (Williams ADPCM vs DCS) which differs per game.
- No suicide encryption; the CMOS only stores settings/scores.
- ⚠️ Do not host ROMs — method only.
EPROM repair¶
TMS34010 (GSP) program + ADSP/DCS sound. Program and graphics on a ROM board (mask ROM + boot EPROM). Converting = swap the game's ROM board (Mortal Kombat, NBA Jam). See MAME midway/midtunit.cpp.
Modernisation & mods¶
The common ground for every JAMMA board
Video, power, control panel and upscalers depend on the cabinet rather than the board: everything is centralised in Modernising a cabinet. To run the board outside a cabinet (at home, on a display): supergun.
- Battery 🟢: a save battery is fitted (settings/counters/RTC). ⚠️ On this platform, a dead battery does NOT kill the game — it is a leaking battery that corrodes traces. Check and replace it before it leaks → save batteries.
- Storage 🟡: the game lives in EPROM/mask ROM on the board. Changing or restoring a game means reprogramming the EPROMs (and, for a mask ROM, using a pinout adapter) → EPROM method.
- Media preservation: replacing the original storage, multicarts, ODEs and modern options by family → PCB preservation hub.
- Scan rate 📺: a TMS34010-based board: the screen configuration is programmed by the game in the graphics controller and does not appear in the driver — not measured here. In practice these boards stay at standard ~15 kHz. (Derived from the architecture documented above.) → tri-sync and upscalers.
- FPGA / emulation: see the Emulation & FPGA section below.
Repair (technician)¶
- TMS34010 (GSP): CPU + graphics engine — corrupt/absent image = TMS34010 / framebuffer RAM / blitter → diagnose the GSP + video RAM.
- Sound: Williams ADPCM (M6809 + YM2151 + OKI6295) (MK1) or DCS (ADSP-2105) (NBA Jam/MK2) — a daughterboard to diagnose separately.
- ROM board: contacts. Capacitors: recap. EPROM sockets: turned-pin socket.
→ Methods: PCB fault-finding · Recap · EPROM · Traces & vias
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.
The factory schematics — 33 sheets made navigable ✅¶
What this wiki had, and what it was missing ✅
Since 13 August this wiki has held pages 3-1 to 3-4 of the Mortal Kombat II — Operations Manual (16-40029-101, October 1993). ⓘ A second harvested document, titled "Midway T-Unit Schematics", is the rest of that same chapter: pages 3-5 to 3-38.
✅ The join is proven, not inferred: the last page of the first (3-38) and the last sheet of the second are the same assembly drawing — CPU (T-UNIT) ASSEMBLY A-17266-40029 and SOUND BOARD A-17319-40029, with the same cable references (H-17297-2, H-17442-2, ribbon 5795-10937-18, spacer 03-8338-1).
ⓘ The two documents are therefore two halves of one chapter, catalogued separately and under two titles that look nothing alike.
⚠️ What these sheets carry in their header
Every sheet carries this, in plain sight, above the title block:
"INFORMATION CONTAINED HEREIN IS CONFIDENTIAL PROPERTY AND MAY NOT BE USED OR DISCLOSED WITHOUT EXPRESS WRITTEN CONSENT FROM THE COMPANY."
ⓘ This wiki reproduces it rather than smoothing it away. The document is public on Internet Archive; the reservation is that of Williams Electronics and Games Inc. / Midway Manufacturing Company, 3401 N. California Ave., Chicago IL, and it dates from 1992-1993.
The three drawing sets, and what they cover¶
| Drawing set | Ref. | Sheets | Project / date |
|---|---|---|---|
| CPU BOARD | 16-9470, rev. 5 | 22 | 40023 · first usage A-14816 09DEC92, current project 19AUG93 |
| Image ROM board | 13664-01, rev. 1 | 5 | drawn 26 July and 20 August 1993 |
| DCS Video Sound Board | 5766-13825-01 | 5 | 40029 · eng. John Lowes, drawn 1 September 1993 |
⓵ Plus an assembly drawing (3-38) showing where it all mounts in the cabinet.
The contents of the 22 CPU board sheets ✅¶
⚠️ The sheets carry no contents list: you have to leaf through them to know which one to look at. Here is each one's title, read off its title block.
| # | Title block | # | Title block |
|---|---|---|---|
| 1 | 34010 PROCESSOR | 12 | IMAGE ROM BANK 2/BANK 3 |
| 2 | CLOCK GENERATION/PLAYER MONITOR | 13 | I/O DECODING |
| 3 | ADDRESS GENERATION/BUFFERS | 14 | MISC. I/O AND DIP SWITCH |
| 4 | SYSTEM MEMORY | 15 | PLAYER 1/PLAYER 2 INPUTS |
| 5 | IMAGE/VIDEO MEMORY DATA BUFFERS | 16 | EXTRA INPUTS/COIN/START |
| 6 | SOUND INTERFACE | 17 | ⓘ PLAYER 3/PLAYER 4 INPUTS |
| 7 | ⓘ FPGA | 18 | ⓘ A/D |
| 8 | DMA | 19 | RAMDAC |
| 9 | VIDEO RAM BANK 0/BANK 2 | 20 | OUTPUTS |
| 10 | VIDEO RAM BANK 1/BANK 3 | 21 | POWER |
| 11 | IMAGE ROM BANK 0/BANK 1 | 22 | CONNECTORS |
ⓘ Two sheets say something the usual description of the T-Unit does not: an FPGA sheet (7) and an A/D sheet (18) — an analogue-to-digital converter. ⓘ And sheet 17 wires a player 3 and a player 4, while the same manual's JAMMA Chart assigns only two players: the inputs exist on the board, the JAMMA harness does not use them.
ⓘ Sheets 9-10 (video RAM) and 11-12 (image ROM) are organised by banks 0 to 3 — useful when a symptom hits only part of the graphics.
Two manuals, one drawing, two references ⚠️¶
This wiki also archives the MORTAL KOMBAT KIT — Using the T-Unit CPU (16-43725-101-B, February 1993, 74 pp.). ⓘ It contains the same CPU board drawing: the same 22 sheets, the same titles, the same first usage A-14816 09DEC92, the same engineer REM.
⚠️ But not the same drawing reference
| Mortal Kombat Kit (Feb 1993) | Mortal Kombat II (Oct 1993) | |
|---|---|---|
| CPU drawing ref. | 16-9423 | 16-9470 |
| Revision | — (none) | 5 |
| "Current project" | 40023, 09DEC92 | 40023, 19AUG93 |
| Change notes | none | three (below) |
⚠️ So one board schematic circulates under two drawing numbers, eight months apart. ⓘ Asking for "schematic 16-9423" and for "schematic 16-9470" gets you the same drawing, two revisions apart — and the reverse is the trap: believing you hold two different boards.
The three factory changes, as printed ✅¶
The October 1993 printing carries notes the February one does not:
| Note carried on the sheet | Sheets affected |
|---|---|
| ⚠️ "19 AUG. 1993 REM — CHANGED UG7 74LS32 TO 74F32." | 4 (System Memory), 13 (I/O Decoding), 19 (RAMDAC), 20 (Outputs) |
| "25 MAY 1993 REM — ADDED C1001, C1002, C1003" | 21 (Power) |
| "25 MAY 1993 REM — ADDED PWR, GND, IROMSEL~, & IBANKSEL TO P14 & P15." | 22 (Connectors) |
⚠️ UG7: one package, four sheets
UG7 is a single chip, but the note about it is repeated on four sheets. ⓘ Looking for it on one sheet is a way to miss it.
⚠️ And the change is not cosmetic: 74LS32 → 74F32 swaps a quad OR gate for its fast version. ⓘ Fitting a 74LS32 back on a late board undoes a factory change — this wiki does not say what symptom that produces, and neither does the document. ⓘ The same lesson as the Wells-Gardner yokes: a part that fits is not thereby the right one → Wells-Gardner.
What the kit demands of the host cabinet ✅¶
The Mortal Kombat Kit manual is a conversion manual: it says what the salvaged cabinet must provide. ⓘ That is usable for any T-Unit board dropped into a second-hand cabinet.
| Item | What the manual writes |
|---|---|
| Power supply | +5 V dc at 5 A and +12 V at 2 A — "These operating voltages are necessary for your kit. Your power supply must be FCC approved." |
| Monitor | ⚠️ "This kit is not intended for use with X-Y monitors." It needs a raster set with a horizontally mounted CRT, separate red, green and blue inputs, and composite negative sync |
| Salvaged parts | "Always repair circuit board malfunctions and cabinet damage before conversion is attempted." |
| Connectors | "If they do not slip on easily, do not force them. A reversed connector may damage your kit and void the warranty." |
The transformer: seven taps, and the lowest brushes a known floor ✅
Mortal Kombat II's Transformer Chart gives the voltage selection table of transformer 5610-13326-00, by jumpering pins:
| Voltage | Jumper pins | Voltage | Jumper pins | |
|---|---|---|---|---|
| 103.5 V | 5-6, 7-8, 10-9 | 207 V | 5-9, 10-8 | |
| 115 V (as shipped) | 5-6, 7-11, 10-12 | 218.5 V | 5-9, 10-11 | |
| 132 V | 5-6, 7-14, 10-15 | 230 V | 5-12, 10-11 | |
| 247 V | 5-12, 10-14 |
ⓘ Seven taps, and the lowest is 103.5 V — just under the 105 VAC below which Bally/Midway writes that no game will work → mains chain.
⚠️ And moving to 230 V is not just the transformer. The drawing's note: "FOR 230VAC OPERATION: 1. FUSE IS 2.0A SLOW BLOW. 2. VARISTOR IS 250V." — against 3 A slow-blow and a 130 V varistor in the 115 V build. ⓘ Changing the transformer tap without changing the fuse and the varistor leaves two parts at the wrong rating.
ⓘ The drawing also names a 120 V isolated output for the monitor and the power supply, and 5.9 V for the coin door lights.
Photos / assets¶
(To add: T-Unit ROM board, TMS34010, DCS/ADPCM sound board, CMOS battery.)
Archived original documents ✅¶
Manufacturer documents archived locally, found on Internet Archive on 13 August 2026 and identified from their title page. ⚠️ The items declare no licence — kept for preservation, provenance named.
-
PDF — midway-mortal-kombat-2-operations-manual.pdf — MORTAL KOMBAT II — Operations Manual, Midway, October 1993, ref. 16-40029-101, 7 pp. The cover advertises Operation & Adjustments · Testing & Problem Diagnosis · Parts Information · Wiring Diagrams & Schematics**. Via Internet Archive
-
PDF — midway-mortal-kombat-2-schematics-section-3.pdf — MORTAL KOMBAT II — Operations Manual, section 3, pages 3-5 to 3-38, Midway, October 1993, ref. 16-40029-101, 33 sheets. ⚠️ The archive titles it "Midway T-Unit Schematics" and names no game: it is the continuation of the document above, of which this wiki held only pages 3-1 to 3-4. The join is proven by assembly drawing A-17266-40029, identical on both sides. It holds three drawing sets: CPU BOARD 16-9470 rev. 5 (22 sheets), image ROM board 13664-01 rev. 1 (5), DCS Video Sound Board 5766-13825-01 (5). ⚠️ Every sheet carries Williams Electronics and Games Inc.'s "INFORMATION CONTAINED HEREIN IS CONFIDENTIAL PROPERTY" legend. Harvested from Internet Archive's arcademanuals collection.
- PDF — midway-mortal-kombat-kit-t-unit-cpu-operations-manual.pdf — MORTAL KOMBAT KIT — Using the T-Unit CPU — Operations Manual, MIDWAY, February 1993, ref. 16-43725-101-B, 74 pp. ⓘ A conversion manual: what the host cabinet must provide, the procedure from a Street Fighter II cabinet or any generic video cabinet, JAMMA and non-JAMMA, the menu system, diagnostic tests, bookkeeping, parts lists, and the same CPU board drawing under reference 16-9423. ⚠️ The Internet Archive title reads 16-43275-101-B and "Feb 1983"; the cover reads 16-43725-101-B and February 1993. Harvested from Internet Archive's arcademanuals collection.
Sources & attribution¶
- MAME —
williams/midtunit.cpp(TMS34010 @ 50 MHz, Williams ADPCM vs DCS ADSP-2105 sound, MK/MK2/NBA Jam games,mkDIP coinage/CMOS) — https://github.com/mamedev/mame/blob/master/src/mame/williams/midtunit.cpp - Wikipedia — Midway T-Unit / Mortal Kombat II / NBA Jam — CC-BY-SA — https://en.wikipedia.org/wiki/Midway_T-Unit
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.