Gottlieb System 80 Won't Boot: Display Shows 0, or Nothing at All

The short version: before you suspect the CPU board, check that the two jumper wires in the interconnect harness between driver board and CPU board are fitted, measure +5 V at the silk-screened test point on the power supply, and find out whether your game should have a reset board on TC1. A System 80 gives you no flashing LED and no error code when it fails to boot, so sequence is everything — and the first few checks cost nothing.

System 80 ran 1980-1982, 80A 1982-1985, 80B 1985-1989. If you own a Black Hole, a Mars God of War, a Haunted House, a Devil's Dare or a Rock, this is your platform.

Contents

• Why there is nothing to read at power-up
• What the symptom narrows down
• Step 1: the two interconnect jumper wires
• Step 2: +5 volts, measured properly
• Step 3: corrosion, connectors and grounds
• Step 4: the reset board and the pull-up
• Step 5: clock, processor and chips
• The trap: fine in test mode, dead in a game
• Reading the memory test error codes
• When a board is genuinely the answer
• FAQ
• Sources

Why there is nothing to read at power-up

A Bally MPU of the same era blinks an LED and the count narrows the fault (what each Bally MPU LED flash means); Williams put numeric codes on the displays (Williams System 7 diagnostic codes). Gottlieb fitted neither, and the clearest proof is what the community built instead: the widely used System 80 test EPROM by Marco Albus asks you to wire your own LED to an address line to get flash codes at all. So work the five steps below in order. Three are free.

What the symptom narrows down

Watch the machine for the first ten seconds after switch-on. What the displays do, and what the playfield does alongside them, narrows things a long way.

What you see	Where to look first
Nothing at all: no displays, no illumination, no sound	Mains, main fuse, transformer
Illumination on, displays dark	+5 V to the CPU board, then the display supplies
A single 0 in the credit window, no attract mode	The CPU is not finishing its boot: steps 1 to 5
Zeros flash up, then lamps flash erratically, tilt relay clicking in time	A RIOT at U4, U5 or U6, or a corrupt ROM
Coils or lamps lock on at power-up	Switch off now. Missing reset board, or grounding
Boots some days, not others, or worsens as it warms	Connectors and grounds, not chips
Green or grey crust near the battery	Step 3, and do not power it up
Test mode works, but a game will not start	The test mode trap below

No sign of life at all? Our guide to a pinball machine that will not turn on is the faster route in.

Step 1: the two interconnect jumper wires

The cheapest check on the list, and it catches people on any machine that has recently had boards out.

Thirty-three discrete signals pass from CPU board to driver board across the A1-J4 to A3-J1 connection. Two of them are not signals: they carry logic ground and +5 V to the driver board, as jumper wires in the harness. PinWiki is blunt — "be sure to install the two jumper wires in the interconnect harness between the driver board and MPU." Missing or badly crimped, they do not fail tidily: lamps, coils, relays and sound can lock on or stay dead, and the game may not boot at all. Check both ends on continuity rather than trusting the look of a pin.

Step 2: +5 volts, measured properly

The CPU board tolerates very little either side of 5 volts. The power supply carries a +5 VDC test point silk-screened onto it, and the reading you want is 5.0 to 5.1 VDC, measured across the electrolytic capacitor next to the MPU power connector. Measure with the game running: a rail that reads fine unloaded can sag under load.

If your System 80 or early 80A still has the big orange 6800 µF filter capacitor on the cabinet floor, assume it is finished. At forty-plus years they are all dead or dying, and a tired one pollutes the supply line and makes every other measurement unreliable — which is why replacing it is step one of the standard John's Jukes procedure. Check fuses on continuity too: a fuse can look perfect and be open.

Step 3: corrosion, connectors and grounds

System 80 CPU boards carry an on-board battery to keep audits and high scores in the 5101 CMOS RAM at Z5. Four decades on, many have leaked, and what leaks creeps along traces and into connectors.

• Do not power up a visibly corroded board. Corrosion bridges tracks and turns a repairable board into scrap. Our guides to spotting and stopping battery corrosion and getting the battery off the board for good cover the rest.
• Clean edge connector contacts with a pink rubber eraser. The attached warning matters: "do not use sandpaper or fiberglass erasers on the contacts — they will only thin the material."
• Take care re-pinning A1J2 and A1J3. Several wire colours there are very similar — 344 and 677 on A1J2, 433 and 766 on A1J3. Label before you cut.

Grounding belongs in the same breath. Unlike Bally and Williams, Gottlieb gave its boards no common metal ground plane: ground is daisy-chained board to board through the harness, and logic ground reaches the CPU board through essentially a single path. When that path gains a few ohms the machine does things no schematic predicts, intermittent booting among them — our System 80 ground mods guide explains the flaw and the fix, and the voltage-drop test in burnt pinball connectors finds a bad pin faster than swapping boards. If your machine boots differently on different days, stop reading about chips.

Step 4: the reset board and the pull-up

System 80 shipped without a watchdog circuit, which most competing systems had. Gottlieb addressed this later with a small reset board that monitors the display digit strobes: if the CPU locks up and the strobes stop, it resets the CPU, which is what stops a frozen processor leaving coils and lamps locked on.

It plugs into TC1 via a 40-pin DIP connector, with two consequences for a no-boot. On a machine that came with one, the game may not boot at all with that board removed — worth knowing if you have been swapping parts. And the board itself fails, usually through dried electrolytic capacitors, producing intermittent resets rather than a clean failure.

Check for the documented pull-up too. Alongside the reset board, Gottlieb added a 3.0 kΩ resistor across pins 7 and 11 of the reset board connection at TC1, to fix a real level mismatch: read/write from the 6502 at U1 pin 34 is only guaranteed at 2.4 V or higher for a logic 1, while the 4069 at Z36 needs 4 V or higher on its input. The resistor holds read/write high when the 6502 is not pulling it down; in practice you will also see 3.3 kΩ, the nearest common value. A missing pull-up gives exactly the works-sometimes behaviour that sends people chasing chips for a fortnight.

Step 5: clock, processor and chips

Only now is it reasonable to suspect the silicon. The map of the board:

Position	Device	Job
U1	6502	The processor
U2, U3	Masked ROMs	The operating system
U4	6532 RIOT	Switch matrix
U5	6532 RIOT	Displays
U6	6532 RIOT	Solenoids, lamps, sound
Z5	5101	256 x 4-bit battery-backed CMOS RAM
Z3	7404	Clock buffer
Z7, Z8, Z9, Z10, Z12	Address buffers	Address bus
Z4, Z36	Selection circuitry	Chip select and control logic

Three measurements before anything leaves a socket. Pin 40 of the 6502 at U1 should sit at 5 volts, and there should be a clock signal at pin 37. If the clock is absent, suspect Z3 — with one caveat that trips up careful people: Z3 must be a genuine 7404. An LS04, HCT04 or S04 is pin-compatible and looks like a sensible modern substitute, and will not do the job.

Beyond that, a jammed address or data line anywhere stops the processor dead, so a failed address buffer at Z7 to Z12 presents identically to a failed CPU. This is where a diagnostic tool beats a parts cannon: our guide to using the Lisy80 as a troubleshooting tool covers testing lamps, coils and switches one at a time, and pinball machine electronics explained maps which board does what.

The trap: fine in test mode, dead in a game

This one is counter-intuitive and has cost people weeks. A machine that passes its tests but will not start a game feels like software or a switch. It is often an address line.

A documented Counterforce case ran through replacement ROMs, a replacement CPU, new capacitors, ground mods, a reset board and a swapped Z36 before the owner found the real fault: a broken trace from U3 pin 21, the BAB12 address line, to Z10. Why did diagnostics pass? A plug-in diagnostic board does not use the BAB12 and BAB13 signals during its tests. The board meant to find the fault was bypassing it.

So when diagnostics and symptom disagree, the diagnostics may not be exercising the broken path. A continuity check from each ROM address pin to its buffer is tedious, and it is what finds this. If your game boots and then falls over mid-play instead, see our full diagnostic of a crashing System 80B.

Reading the memory test error codes

Once the machine boots far enough to enter diagnostics, it will tell you things. Open the coin door, press the red momentary switch, then press credit/start once to skip bookkeeping and reach the tests: 16 lamp, 17 coil, 18 switch, 19 display, 20 memory. On an 80B the top alphanumeric display shows TEST MODE. Test 20 reports:

• 7641-1 or 7641-2 in the player 1 window (System 80 and 80A): a ROM checksum error, the digit telling you whether PROM 1 or PROM 2 is at fault.
• 5101 (System 80A): a CMOS RAM error, or a problem accessing it.
• ERROR- Z5 (System 80B): most likely a failed 5101, but it can also mean the processor cannot reach the 5101 because of a chip select, address line or data line problem.

That last distinction matters: "ERROR- Z5" is not automatically a dead RAM chip, and replacing the chip without checking the path to it is how people end up with two 5101s and the same fault. If the game boots but displays stay dark, that is usually a supply problem — is it the display or the power supply? tells them apart. If it boots but makes no noise, System 80 no sound picks up from there.

When a board is genuinely the answer

We would rather you fixed a jumper wire than bought anything, and on this platform that is often how it ends: steps 1, 3 and 4 are free or nearly free. A board becomes the honest answer in narrower cases — battery damage that has eaten through traces and pads rather than merely coated them, a board repaired badly enough that the next fault is unpredictable, or a hunt for a broken internal trace that has already cost more hours than the board is worth.

Two routes we work with. The Lisy80 is built on the LISY project — an independent developer's free, non-commercial DIY project, which we support but did not create — and its value here is diagnostic as much as replacement, since it gives you the per-device testing Gottlieb never fitted. The GottFA80_Plus replaces CPU, driver and power supply with one board on System 80 and 80A, removing several of the fragile ground and connector paths above simply by removing the cables. Our comparison of the all-in-one options sets out where each fits, and repair or replace a pinball circuit board covers the decision. We test boards before they ship and support them afterwards, and our technical help is free whether or not you buy anything.

FAQ

Why won't my Gottlieb System 80 boot?

In this order: check the two jumper wires carrying logic ground and +5 V in the interconnect harness between driver board and CPU board; measure +5 V at the power supply test point and confirm 5.0 to 5.1 VDC; inspect the CPU board for battery corrosion and clean the edge connectors; check whether your game should have a reset board on TC1. Only then suspect the chips. The first three cost nothing.

My System 80 displays just show a 0. What does that mean?

A single zero in the credit window with no attract mode means the CPU is powering up but not completing its boot, so it never reaches the game program. It is a no-boot, not a display fault. Start with the interconnect jumpers, the +5 V rail and battery corrosion.

Does a Gottlieb System 80 have a diagnostic LED?

No. Gottlieb fitted no boot LED and no flash code to the System 80 CPU board. The community test EPROM asks you to wire up your own LED on an address line precisely because there is nothing on the board to read.

My System 80 works in test mode but won't start a game. Is the ROM bad?

Not necessarily, and this is a known trap. A documented Counterforce case turned out to be a broken trace from U3 pin 21 — the BAB12 address line — to Z10. Diagnostics passed because a plug-in diagnostic board does not use the BAB12 and BAB13 signals, so it bypassed the exact fault it was meant to find. Check continuity from the ROM address pins to their buffers before buying ROMs.

Can I substitute an LS04 or HCT04 for the 7404 at Z3?

No. Z3 is the clock buffer and must be a genuine 7404. An LS04, HCT04 or S04 is pin-compatible but will not work reliably there, and the symptom is a missing clock at pin 37 of the 6502 — which looks exactly like a dead processor.

What does "ERROR- Z5" mean on a System 80B?

A memory error from test 20. Most likely a failed 5101 CMOS RAM at Z5, but it can equally mean the processor cannot reach the 5101 because of a chip select, address line or data line fault. Check the path to the chip before replacing the chip.

Sources

• PinWiki: Gottlieb System 80 — chip designations, +5 V test point, interconnect jumpers, self-test sequence, error codes, TC1 pull-up
• Marco Albus, Gottlieb System 80 universal test EPROM V3.3 — Z3 7404 clock buffer, 6502 pins 37 and 40, address buffers, RIOT assignments
• Pinside: Counterforce won't boot, display shows 0 — the BAB12 broken trace at U3 pin 21
• Pinside: System 80 (The Games) boot problems, fine in test mode — RIOT symptoms and the flashing-lamps pattern
• Pinside: Gottlieb reset boards — the watchdog, the TC1 connection and the pull-up
• Pinside: Gottlieb System 80A no boot, displays show 0
• John's Jukes: Gottlieb ground cures — the orange filter capacitor and ground procedure
• Action Pinball: System 80 EPROM modifications — PROM sockets and board part numbers
• The LISY project — independent, free, non-commercial DIY replacement CPU boards

Stuck on a System 80 that will not wake up? Tell us the game, exactly what the displays do in the first ten seconds, and what you read at the +5 V test point. Our technical support is free, before and after any purchase — and on this platform the answer is very often a wire, not a board.