Pinball Coil Stuck On: How to Stop the Damage and Find the Fault

You switch the machine on and something under the playfield starts to buzz and stays buzzing. A pop bumper fires and never lets go. Within a minute there is a smell like hot varnish.

A stuck coil is one of the few faults on a 1970s or 1980s machine that gets worse every second you leave it. These coils are built to be on for hundredths of a second at a time. Held on, a coil cooks its own windings, and the transformer feeding it does not enjoy the experience either.

The short version: switch off first and ask questions afterwards. If the coil locks on the instant you switch on, before the game has booted, it is almost certainly a shorted driver transistor. If it locks on during a game, suspect a stuck playfield switch or the CPU. If it is a flipper, stop looking at the driver board — flippers on these machines are not driven by it.

Contents

• Switch off, then pull the right fuse
• Why a coil locks on
• First question: is it a flipper?
• When it locks on tells you most of it
• Testing the driver transistor
• When it is a switch, not the board
• What to check on your system
• When replacing the board is sane
• FAQ
• Sources

Switch off, then pull the right fuse

This is not a fault to live with while you read up: power the machine down at the switch. When you come back, you want the game running but with no current reaching that coil, so pull the fuse feeding the solenoid supply. Your manual names it; on Williams System 3 to 7 it is F2, with F3 feeding the lamp matrix. The game will boot, the displays will light and the switches will read, but nothing will fire.

Be careful which fuse you pull. Playfield lamps here are strobed by the CPU rather than held on, so if the CPU stops strobing, lamps that are still fused sit at full voltage and burn out in seconds. And if the coil is visibly scorched, leave the machine off and come back to it cold.

Why a coil locks on

A coil has two lugs. One is permanently connected to the solenoid supply whenever the machine is on. The other runs to the driver board, where a transistor connects it to ground when the game wants that coil to fire. The coil is always half-live; nothing happens because the ground side is open. Firing means closing that path for a few milliseconds.

So a stuck coil means something is holding it closed. Three things can:

• The driver transistor has failed short. Collector and emitter are joined — a wire, effectively. Nothing the CPU says will change it.
• The CPU is genuinely commanding the coil on, because a switch it watches looks stuck closed, or because the board locked up mid-cycle.
• The pre-driver has failed, the smaller transistor between the CPU's output and the power transistor.

For the wider picture of which board does what, see our guide to pinball machine electronics.

First question: is it a flipper?

Answer this before anything else, because people lose evenings testing transistors that were never in the circuit. Flipper coils of this era are not switched by the driver board. The cabinet button closes a normally-open switch and that completes the path to the coil. Early Bally and Stern games add a flipper enable relay on the solenoid/regulator board, driven by the Q15 circuit, but the switching is still the button.

A flipper coil has two windings in series: a high-power winding of a few ohms, drawing something like 13 amps at 43 volts, and a hold winding of around 265 ohms, taking about 160 milliamps. The EOS switch — end of stroke — is normally closed, and while closed it shorts out the hold winding so all the current goes through the high-power one. When the flipper reaches the top of its travel the EOS opens, the hold winding comes into circuit, resistance rises and current drops to a safe trickle.

So if the EOS contacts weld closed, or the flipper never physically reaches the end of its stroke, the high-power winding stays in circuit at full current and the coil burns. Same if the cabinet button sticks. Check the EOS gap and contacts, the button switch, and whether the mechanism binds — not the driver board.

When it locks on tells you most of it

With the fuse back in and your hand near the switch, watch when it happens. This one observation splits the problem in half.

What you see	Most likely cause
Coil energises the instant you switch on, before the game boots	Shorted driver transistor — the CPU has not had time to command anything
Coil locks on part-way through a game	Stuck playfield switch, or a CPU that crashed mid-cycle
Coil fires repeatedly and never resets	The switch that should report the job done is stuck open
Several coils affected at once	Shared circuit: pre-driver stage, a supply rail, or grounds
A flipper only	EOS switch, cabinet button switch, or a binding mechanism

Several coils misbehaving together is worth treating as a grounding problem before you replace anything, particularly on Gottlieb — our System 80 ground mods guide explains why.

Testing the driver transistor

With the machine off and unplugged, any multimeter with a diode function will do this. Most of these drivers are Darlingtons in a TO-220 package — the flat black rectangle with a metal tab. Hold it with the tab facing away and the legs down: left leg is the base, middle is the collector, right is the emitter, and the tab connects to the collector.

A healthy NPN Darlington reads like this in diode mode:

• Red probe on the base, black on collector then emitter: roughly 0.3 to 0.9 V each time.
• Black probe on the base, red on collector then emitter: open both times.
• Between collector and emitter, either way round: open.

A dead short between collector and emitter is your locked-on coil.

The trap, and it catches everyone once: in-circuit readings lie. Surrounding resistors and diodes give current somewhere else to go, and Darlingtons contain resistors and a diode inside the package that produce readings which look wrong on a good part. There is a well-known forum thread in which every transistor on a Bally driver board read bad. They were not. Desolder one leg and retest before you believe a bad reading.

Then check the coil itself before you call the job done. A coil that has been held on has been overheated, and an overheated coil can develop a partial short between windings that lowers its resistance and destroys each new transistor you fit. Measure it against the manual, or against its twin on the other side of the playfield, and put the manual's fuse rating back — not a bigger one.

When it is a switch, not the board

If the coil behaves until a game starts and then locks on, the board may be doing exactly what it was told. The game fires a coil in response to a switch and stops when another switch reports the job done. A drop target bank reset coil that will not let go is often a bank switch stuck closed, telling the game the targets are still down.

Find the switch for that coil in the manual, go into switch test if your board offers one, and confirm it reads open when it should. Bent leaf switches, contacts closed by forty years of dust, and a switch a previous owner adjusted too tight are all more common than board faults — and all free to fix.

A diagnostic board earns its keep here: a Lisy80 on a System 80 lets you watch switch states and fire individual solenoids from a web interface. It is built on the LISY project, an independent developer's free, non-commercial DIY project; we supply it ready to use and support it, and we did not create it. And if a coil is dead rather than stuck, check its own in-line fuse and see our guide to testing a pinball coil with a multimeter.

What to check on your system

Gottlieb System 80, 80A and 80B

Solenoids and playfield lamps share one board: the A3 driver board, which carries nine solenoid driver circuits — labelled QS1 to QS9 — alongside its lamp drivers. The original design uses MPS-U45 pre-driver transistors, now genuinely hard to source, and 2N3055 power transistors with an unpleasant habit: when they fail they can fail spectacularly, and burning a hole through the board is a documented outcome rather than an exaggeration.

So inspect the board, not just the components — a new transistor soldered into a scorched pad will not last. Note too that some high-current System 80 coils are driven by a transistor under the playfield rather than on the board. Machines such as Black Hole and Haunted House have their own fault notes.

Bally and Stern, 1977 to 1985

The board you want is the solenoid driver / voltage regulator in the A3 position: AS-2518-16 on early games, AS-2518-22 later. Driving is done by TIP120 or TIP102 power transistors, and the pre-driver stage is a CA3081 transistor array — one IC containing several transistors.

That detail matters: if the power transistor tests good out of circuit and the coil still locks on, the CA3081 is next, and because it is an array a single failed IC can affect more than one solenoid. The same board carries the LM323K 5 V regulator, a well-known age-related failure. If the CPU is misbehaving too, our guide to Bally MPU LED flashes settles that in a minute.

Williams System 3 to 7

Each solenoid circuit is two stages: a 2N4401 pre-driver feeding a TIP120 power transistor. Many rebuilders fit TIP102s instead, which handle more current. Test the power transistor first, then the 2N4401 behind it — a shorted pre-driver turns the power transistor hard on just as effectively. F2 is your solenoid fuse, and if the CPU LED is also odd, our System 7 diagnostic codes guide will decode it.

When replacing the board is sane

Plenty of stuck coils are a two-euro transistor and twenty minutes. If that is your fault, do it yourself and enjoy it — we would rather say so than sell you something you do not need. The calculation changes when the board has been damaged rather than simply failed: scorched traces, lifted pads, a hole through the laminate, or transistors failing one after another because the board is worn out. Our honest comparison of repairing versus replacing a board sets out the arithmetic.

If you go that way, the Godri80 replaces the System 80 driver board, the BallyDri covers the Bally and Stern power and driver position, and the WillFA7 handles MPU and driver for Williams System 3 to 7. A Gottlieb pop bumper circuit has its own small board, replaced on its own by the GoPOP80. We test boards before they ship and support them afterwards. Keep the original either way — a damaged driver board is still a good winter project once the machine is playing again.

FAQ

Why is my pinball coil stuck on?

Something is holding the coil's ground path closed. Most often a driver transistor has failed short, grounding the coil whatever the CPU wants. Otherwise: a stuck playfield switch, a failed pre-driver, or — on a flipper — a welded EOS or cabinet button switch.

Will a stuck coil damage my pinball machine?

Yes, quickly. These coils are rated for milliseconds at a time; held on, they overheat within minutes, which can destroy the coil, blow the fuse and stress the transformer. Switch off as soon as you notice, and pull the solenoid fuse before testing.

How do I tell a shorted transistor from a stuck switch?

Watch when the coil energises. If it locks on the instant you power up, before the game has booted, the CPU cannot be commanding it and a shorted driver transistor is likely. If it only starts once a game is running, look for a stuck switch.

Can I test a driver transistor without removing it from the board?

You can, but do not trust a bad reading. Surrounding components give the meter alternative paths, and Darlingtons contain internal resistors and a diode that produce odd readings on good parts. Desolder one leg and retest before replacing anything.

Why does my flipper stay energised and get hot?

Flipper coils here are switched by the cabinet button and the EOS switch, not the driver board. If the EOS contacts weld closed, or the flipper cannot reach the end of its stroke, the high-power winding stays in circuit and the coil burns.

My driver transistor keeps failing. What am I missing?

Usually the coil. An overheated coil can develop a partial short between windings, lowering its resistance and pulling enough current to destroy each new transistor. Measure it against the manual or its twin on the playfield, and check the fuse rating.

Sources

• Home Pinball Repair: Pinball solenoid stays locked on
• Home Pinball Repair: Understanding pinball transistors and diodes
• Home Pinball Repair: How serial flippers work in early SS and EM machines
• pinrepair.com: Williams System 3–7 repair guide
• PinWiki: Gottlieb System 80
• PinWiki: Williams System 3–7
• Gottlieb System 80 driver board circuit designations
• Action Pinball: Bally AS-2518-16 solenoid driver
• Pinside: Gottlieb System 80 coil lock mystery
• Pinside: Bally driver board, all transistors read bad

Stuck coil and not sure what your meter is telling you? Send us the machine, the coil and the readings. Our technical support is free, before and after any purchase — including when the answer is a transistor and a switch adjustment rather than a board.