Circus (Gottlieb, 1980): Faults & System 80 Replacement Boards

Is your Gottlieb Circus refusing to start, resetting mid-game, stuck in attract mode, or playing with sound so faint you can barely hear it? Those four complaints cover most of what comes through the door on a Gottlieb System 80 wide body. The original boards are now well past forty, and on Circus they sit behind a NiCd battery that has usually been leaking for decades. Modern replacement boards, battery-free and Plug & Play, put the game back in service.

About the Circus (Gottlieb, 1980)

See it: full photo gallery for Circus on IPDB · owner gallery on Pinside

Gottlieb released Circus in June 1980 as model 654 and built roughly 1,700 of them. Ed Krynski designed the game, Gordon Morison did the artwork, and it was sold on the promise "THE GREATEST PINBALL ON EARTH!".

What makes it worth knowing about is the cabinet. Circus is one of only a handful of true wide-body cabinets Gottlieb built for System 80, and that width is filled: five flippers, two pop bumpers, four clown drop targets and a spinner. In the middle sits a roto target, a rotating drum of scoring values — a mechanism far more typical of Gottlieb's earlier System 1 games than of its solid-state wide bodies. Nothing else on System 80 plays quite like it, which is also why Pinside's archive calls it difficult to operate and repair, and consequently a rare find. Note too that several unrelated machines are called Circus; this article is about the Gottlieb 1980 System 80 wide body only.

How it plays

The numbered 1-4 top lanes light the matching clown drop targets, and knocking those down builds your multiplier. Sequence A-B-C to light special at the roto — you must hit it with the star showing in the middle. Clear the top lanes and all the drops and the roto lights for an extra ball. Turning off both side lanes lights the spinner. Advance bonus to 20,000 and land in the kickout hole and it pays 20,000 times your multiplier.

What one is worth

Treat these as ranges, and note that the sources disagree because they measure different things. Pinside, working from asking prices, estimates $1,390-$1,610 with a median ask near $1,500. PinballPrice, which tracks completed eBay sales, shows yearly averages of $1,448 for 2017 and $1,995 for 2016, with individual auctions closing as low as $516. Both agree on scarcity: the Museum of the Game census ranks Circus 2 out of 100 for how often it turns up.

Circus specifications

• Manufacturer: D. Gottlieb & Co.
• Released: June 1980
• Model number: 654
• Production run: approximately 1,700 units
• Game design: Ed Krynski
• Artwork: Gordon Morison
• Electronic system: Gottlieb System 80
• Cabinet: wide body, 4 players, manual plunger
• Display: alphanumeric
• Playfield: 5 flippers, 2 pop bumpers, 4 drop targets, spinner, roto target
• No: multiball, ramps, magnets, captive balls
• IPDB: machine #515

Common faults on Gottlieb System 80

The battery comes first

Every System 80 conversation starts here. The CPU board carries a NiCd battery to hold settings and scores, and after forty years it has almost always vented. The residue does not stay put: it wicks along traces, climbs into connector pins, and keeps working long after the battery is gone. Cleaning is therefore not enough — residue travels back out of an old housing into fresh pins, so replace clips and housings together. If the battery is still fitted, remove it first, working machine or not.

What is actually on the CPU board

Knowing the chip map saves a lot of guessing. The processor is a 6502 at U1. Three 6532 RIOT chips — combined RAM, I/O and timer devices — sit at U4, U5 and U6 and handle switches, lamps and displays. Battery-backed memory is a 5101 static RAM at Z5. Game ROMs live at U2 and U3, and here is a real generational difference: on System 80 the EPROM is soldered down, where 80A and 80B have it socketed. Around them, Z1 (4528) and Z4 (4081) generate reset while Z2 (7474) and Z3 (7404) make the clock, with Z7-Z10 doing chip-select logic. A board with no clock or no reset will not boot however healthy the CPU is.

This is why a test fixture pays for itself. A System 80 tester steps through the memory devices and reports the failure by flash or digit code: no flash means the CPU, one flash the 5101 at Z5, and two, three or four flashes the RIOTs at U4, U6 and U5 respectively. That turns an afternoon of probing into a five-minute answer.

Grounds: one fault that imitates ten others

System 80 distributes ground by daisy-chaining it through connectors, from power supply to CPU to driver board. As those connectors age their resistance climbs and the ground reference drifts above 0 V. Logic that expects a clean zero starts misreading its own signals, and the symptoms look unrelated to each other: random resets, coils locking on, misbehaving displays, erratic scoring.

The accepted fix, the ground mod, bypasses the connectors with heavy wire, typically 16 AWG: backbox lock ground straight to the power-supply frame, the negative lead of C1 on the power supply to that frame, the same from C1 on the CPU board, the driver board ground plane to the frame, and optionally doubled +5 V and ground lines in the CPU-to-driver connector. Adding one wire at the connector alone is a partial job, because it leaves the daisy chain through the CPU board's own J1 untouched.

Displays, driver board, power supply

System 80 display wiring uses IDC feed-through pins, sometimes with two wires sharing one contact. When they loosen you get the classic complaint: wiggle the wiring and a segment drops out or digits flicker and show garbage. Reseating with an IDC tool works better than swapping in crimp pins, which suit this connector style poorly.

On the driver board the coil transistors are the wear items — failures like Q53 and Q54 cracking or melting at power-up are typical, and Q53 drives the knocker. To separate board from machine, ground a transistor tab with the game on: if the coil fires, the fault is upstream on the board rather than in the coil or wiring. The power supply also drifts with age, and a soft 5 V rail produces the same intermittent nonsense as a bad ground.

Circus-specific problems owners report

Faint or absent sound — the fault Circus is known for

This is the most reported Circus problem and the most satisfying to fix. Owners describe sound either completely dead or so quiet it is barely audible, while the volume control itself tests fine. Suspect capacitor C8 on the sound board: it leaks on its underside, where you cannot see it without lifting the board, and corrodes the connections underneath. Replacing C8 restored normal volume in documented cases.

Two things to check alongside it. The A6J1 connector uses punch-pin contacts with a high failure rate, so converting those to crimp is a sensible upgrade while you are in there. And the LM380N amplifier is the part everyone suspects first — it is often innocent, so test before replacing. If you would rather not chase leaking electrolytics through a forty-five-year-old board at all, a replacement sound board sidesteps the circuit entirely.

Battery acid on the MPU and connectors

Circus turns up with this often enough to plan for it. Owners buying one describe bad acid damage on the MPU with corrosion on most pins, and conclude the connectors are simply shot. Experienced responders point out that treating only the white connectors is a false economy, because alkaline transfers back from the old housings — board and connectors need doing together. Connector clips, housings and alkaline repair kits are available from the specialist suppliers.

The roto target

The roto target is the heart of the game and the part most likely to have quietly stopped working. The failure is rarely dramatic: the leaf-spring blade switches that read the rotating disc bend back over the decades, sometimes sitting an eighth of an inch clear of the hub, and the drum then spins past without registering. Because those blades are riveted at the base you cannot bend them straight at the root; the practical approach is a blade forming tool to curve the middle of the blade and bring the fan back towards the disc, forward enough to clear the disc on a spin.

The hub and similar parts are Bakelite or phenolic paper, so they tolerate water or cleaner — just dry them promptly. A perfect restoration means drilling out rivets, flattening bends and re-riveting, which owners agree is not worth it outside a museum-quality rebuild. If the roto does not turn at all rather than not scoring, move upstream to the drive coil and its driver transistor, using the ground-the-tab test above to place the fault on the board or in the mechanism.

Stuck in attract mode

A Circus that sits in attract mode and answers nothing — not a coin, not the test button, not the tilt — is telling you the processor runs but cannot see its inputs. Suspect the switch matrix, the CPU connectors and the grounds before deciding the CPU has failed.

Pop bumpers

With two pop bumpers working a wide playfield, these take a beating. The characteristic System 80 failure is the two rods linking the ring to the solenoid breaking away from the ring itself. Do not try to save it: experienced restorers are blunt that JB Weld and welding both fail within seconds of the coil firing, and that a new pop bumper ring is the only real answer. Bumpers that are weak rather than broken may be a driver-side fault instead.

Documentation

At around 1,700 built, Circus is scarce enough that owners regularly go hunting for manual information on the forums. Reprints are available from the usual parts suppliers, and it is worth having one: a five-flipper wide body with a roto target is not a machine to work on from general System 80 knowledge alone.

Replacement boards compatible with the Circus

All of these fit System 80, which is what Circus uses. We test boards before they ship and support them afterwards, and we would rather confirm your configuration than have you order the wrong thing: contact us with the references from your boards.

• GottFA80_Plus (Light) — all-in-one CPU, driver and power-supply board, without sound.
• GottFA80_Plus (Full) — the same board with sound integrated, worth considering on Circus given the C8 problem.
• Gosof — replacement sound board, with or without speech.
• Godri80 — replacement driver board.
• Lisy80 — CPU board with web-based diagnostics.
• GoPOP80 — pop bumper driver board, replacing the original MA-922 / A-19741.

Installation is Plug & Play, with no soldering, and every board runs without a battery. Tutorials and technical support are free.

A note on LISY. LISY is a free, independent DIY project by its own developer, documented at lisy.dev — credit goes there. It is not our project; the Lisy80 above is a board we stock, test and support.

One variant point. Because the EPROM on a System 80 board is soldered at U2/U3 rather than socketed as on 80A and 80B, a ROM-level repair on an original Circus board means desoldering — which tips the economics towards a replacement board sooner than on later games.

Step-by-step diagnosis

1. Look before you measure. Power off, check the CPU board for battery residue around the chip positions and connector pins, and remove the NiCd battery if still fitted.
2. Check the 5 V arrives. Measure at the power supply and again at the CPU board; a rail that sags under load behaves exactly like a failed chip.
3. Do the grounds. Ground mod in 16 AWG: backbox lock, power-supply C1 negative, CPU C1 negative and driver ground plane, all to the power-supply frame.
4. Re-pin, do not just clean. Replace corroded Molex clips and housings, especially on the CPU-to-driver link.
5. Read the fault code. With a test fixture: no flash the CPU, one flash the 5101 at Z5, two, three or four flashes the RIOTs at U4, U6, U5.
6. Displays. Never plug or unplug one powered. Reseat the IDC pins with the proper tool and test against a known-good display.
7. Sound. If faint, lift the board and inspect C8 from underneath before suspecting the LM380N. Check A6J1 too.
8. Mechanisms. Set the roto blade switches close to the disc, check the pop bumper rings, and fire each coil from the test menu.
9. Final test. Start-up, credits, all five flippers, coils, displays and sound. Still off? Get in touch — support is free.

FAQ — Circus (Gottlieb, 1980)

Why is the sound on my Circus so faint?
Usually capacitor C8 on the sound board. It leaks on its underside, out of sight, and corrodes the connections beneath; replacing it restored full volume in documented cases. Check the A6J1 punch-pin connector too, and test the LM380N amplifier before replacing it, because it is often innocent.

My Circus will not start at all. Where do I begin?
Look for battery corrosion on the CPU board and measure that 5 V actually reaches it. If the board is clean, powered and still dead, a System 80 test fixture reads the fault out: no flash means the 6502 CPU, one flash the 5101 RAM at Z5, and two, three or four flashes the RIOTs at U4, U6 and U5.

My Circus is stuck in attract mode and ignores the coin and test buttons.
The processor is running but cannot see its inputs, which is usually the switch matrix, the CPU connectors or the grounds rather than a dead board. Re-pin the connectors and do the ground mod before condemning anything.

The roto target on my Circus does not score.
Nine times out of ten it is the leaf-spring blade switches that read the drum. They bend away from the hub over the years, sometimes by an eighth of an inch, and the drum spins past without making contact. They are riveted at the base, so use a blade forming tool to curve the middle of the blade back towards the disc.

The battery has leaked on my Circus CPU board. Can it be saved?
Sometimes, but budget for more than a clean-up. Owners report corrosion on most pins, board and housings both, and residue travels back out of an old housing into fresh pins, so replace clips and housings together. If the traces under the chip positions have gone, a battery-free board is cheaper.

Why does my Circus reset at random?
System 80 daisy-chains its ground through connectors. As they age their resistance rises, ground drifts above 0 V, and the logic misreads its own signals: random resets and coils locking on. The ground mod hard-wires the boards to the power-supply frame instead.

Which replacement board should I choose for a Circus?
It depends how much of the original you still trust. GottFA80_Plus replaces CPU, driver and power supply in one board, Light without sound or Full with sound built in. If a single board has failed, the Godri80 driver, Gosof sound board or Lisy80 CPU cover it individually. Send us your board references and we will confirm first.

Does a battery-free board keep the settings and high scores?
Yes. These boards use non-volatile memory, so settings and scores survive a power-off with no battery in the machine and no corrosion risk.

See also

• Star Race (Gottlieb, 1980) — the other 1980 System 80 wide body
• Panthera (Gottlieb, 1980)
• Counterforce (Gottlieb, 1980)
• James Bond (Gottlieb, 1980)
• Time Line (Gottlieb, 1980)
• The Amazing Spider-Man (Gottlieb, 1980)

Sources

• IPDB — Circus (#515)
• Pinside — Circus specifications
• PinballPrice — Circus sales
• Museum of the Game — census
• Pinside — Circus no sound (C8, A6J1, LM380N)
• Pinside — Circus acid damage and connectors
• Pinside — Circus rulesheet
• Pinside — roto target blade switches
• Pinside — System 80 pop bumper repair
• Pinside — System 80 ground mod
• Pinside — driver board (Q53/Q54)
• Pinside — display connector problems
• Pinitech — System 80 tester (chip map, flash codes)
• Great Plains Electronics — QuickScan80 manual
• Home Pinball Repair — System 80 CPU
• Kineticist — Circus
• lisy.dev — the independent LISY project

Own a Circus? Send us a photo and we'll add it here with credit — get in touch.