Eye on electronics

Motor, Mar 1998 by Dale, Mike

This month, Mike gives us an engineer's look into the world of ignition coils-including how they're built, what goes wrong with them and how to test them.

Scott took his 1989 Cutlass with the 3800 V6 into a chain tire store because it wasn't running right. The car was difficult to start in the morning and frequently stalled at traffic lights.

The chain store solution was to replace the entire ignition system, to the tune of $800. The shoebox full of "dead" parts Scott was given included three DIS coils, the ignition module and the crank sensor, plus all the spark plugs and plug wires.

The stuff in the box represented a financial disaster for Scott. Eight hundred bucks is a lot of money for anyone, let alone a student trying to work his way through college. Putting that much money into a 10-year-old car is another issue. On top of that, the chance that all of those parts went bad at the same time is extremely unlikely. The tire dealer's defense was that he had no way to test all the parts, and the only way to make sure the repair would stick was to "shotgun" the system by replacing everything in sight.

The worst part of the whole story is that Scott's car isn't running any better now than it did before. It still won't start easily in the morning and is still prone to stalling. By now he's been to an Olds dealership and was told to put up with it, as his car is probably not worth what it would cost to fix it.

It shouldn't have to be like this. The whole purpose of tech education is to be a rifle, not a shotgun. Maybe there isn't a magic solution to Scott's dilemma, but let's start with the ignition coils. This month, we'll look at how they're built, what can go wrong with them, what's not likely to go wrong and how you can test them.

There are really just four major components that go into an ignition coil. The primary coil consists of approximately 160 turns of rather heavygauge wire. Usually, this is wound into a tall, rectangular coil, two or three layers deep. In most coils, the wire used has a special, meltable coating. It's wound on a square spindle and held temporarily in its correct shape. A heavy pulse of current is then passed through the wire to heat it up, causing the meltable coating to liquefy just long enough to reflow. When the current pulse is terminated, the coil cools and the meltable coating solidifies, holding the shape of the coil permanently as it's ejected off the winder.

The secondary coil consists of approximately 10,000 turns of very fine-gauge copper wiresmaller in diameter than a human hair-that's coated with a very thin layer of polyurethane varnish. Being that small and flimsy, you can imagine that something has to hold it in place. That something can be either of two things-a plastic bobbin or a paper-layered cardboard spool.

Most modern design coils use a plastic bobbin separated into bays, or segments. When the bobbin is being wound, the machine starts at one end and fills the first bay. Once that's full, the machine stops, gently guides the wire through a crossover in the dividing flange, then starts to wind the second bay. Any number of bays between four and eight can be used.

Older style coils are wound one layer at a time on a cardboard spool. After each layer of the coil is wound, another layer of paper is placed on top of it. Then the next layer is wound on top of the first, and so on. Each coil design has its own advantages. Paper-wound coils are somewhat more efficient, while plastic bobbin coils have better dimensional repeatability.

The next important coil component is the lamstack, or core of the coil. It's a rule of coils in general and transformers in particular that they're much, much more efficient if there's iron in the core. These coils work on the principles of magnetism and inductance. By adding iron to the core, you add an easy path for the flow of magnetism. Magnetism doesn't flow only in the core, however. The flux forms a complete circle down through the middle of the core, out around the sides and back down into the middle. Most ignition coils have lamstacks that are rectangular in shape, with the coil located on an extra leg in the middle.

If you look closely at the core, you'll see that it's not a complete circle. Most coils of this type have an air gap, or slice, cut out of the lamstack. Even though this slice might be welded shut on one side, it still represents an air gap in the core. This very accurately set air gap is used to keep the magnetic core from saturating.

There is an important feature to the lamstack that may not be immediately obvious-it's laminated. Layers of steel are stamped out, stacked up and then riveted or joined together. The reason for this is eddy currents. Eddy currents in magnetics can be likened to the eddy currents of water that swirl around rocks in a river. The currents represent a loss of energy that can be prevented by making the magnetic flux flow in layers created by the sheets of iron.

Probably the least understood component of the coil is the epoxy filler. While it's true that the epoxy keeps water and road debris out of the coil, that's not its main purpose. Typical automotive coils will put out close to 40 kV under the right circumstances. It's the insulation system of the coil that's responsible for hanging on to that voltage, and this is accomplished mostly by the epoxy. The wire insulation, the bobbin itself and the housing also play a role.