The Starter Story: Rebuilt Bosch, Gear Reduction or Permanent Magnet?

I have owned rear-engine Porsches and VWs since 1974. I can't remember a single one that I've owned that did not have an occasional no-crank issue at some point in its life, either when cold or hot. Replacing the starter with a rebuilt Bosch unit would remedy the problem for a while, but sure enough—within a year or two—the same problem would return. For years I mulled this over, scheming possible solutions. This is the type of thing that drives engineers crazy.

One common theory was that the ignition switch was located too far from the starter, and the resulting voltage drop (because of the long wire), coupled with the high current demand of the starter solenoid, was so great that there simply was not enough power at the starter to make it go. On that premise, many people (myself included) tried to remedy the problem by adding a load-reduction relay or Ford-type starter solenoid back near the starter. The Bosch corporation made such a kit (their "WR-1") specifically for rear-engine VWs and Porsches. With this setup, the ignition switch and (long!) length of wire would have to supply only enough power (very little) to trip the relay. Great idea! The only problem was... it didn't work consistently. I continued to puzzle over it for another ten years or so until I finally got so frustrated that I started studying the problem. I discovered that the issue was in the design of the starter itself.

The original Bosch starter design is two cylindrical parts bolted together: a solenoid and an electric motor. The assembly of the two is the "starter." The smaller of the two cylinders is the solenoid, and the larger is the electric motor. The solenoid is an electromagnetic device. Inside this small cylinder is a smaller (and also cylindrical) solid iron core with a couple of miles (literally!) of copper wire.

The wire is wound around in such a way that it surrounds the iron core but does not touch it. When current flows through the wire, a magnetic field is created which causes the iron core to get sucked up inside the windings, thereby creating a pulling force. One end of the iron core is mechanically linked to a seesaw device that operates a "Bendix" drive, which has a small gear on it. That small gear mates with the big gear on the flywheel, which is bolted to the crankshaft of the engine. By pulling on one end of the seesaw, the iron core within the solenoid engages the two gears together. At the other end of the iron core is a big, fat switch. It is through this switch that the electric motor gets power via the big fat wire that is bolted to the starter and goes straight to the battery. So now, with the two gears mated, and power supplied to the electric motor, the engine cranks over. Seems pretty simple and bulletproof, right?

I started studying the starter circuit and noticed that the solenoid has two separate sets of electrical windings. One set is called the "pull" set, and the other is the "hold" set. The pull set is much stronger than the hold set, which is made possible by more loops of wire. What is interesting is that the pull set gets its ground path through the windings of the electric motor. The hold windings get a separate ground path directly to the chassis ground. So, when you turn the key in the ignition switch, electricity has to flow through the mile of wire between the ignition switch to the solenoid, then through another few miles of wire that are the pull windings of the solenoid, and then through another million miles of wire that are the electric motor windings. When everything works right, the solenoid pull windings yank on the iron core, which engages the Bendix drive. The other end of the iron core energizes the motor windings. That gets the motor spinning, but in doing so kills the ground path of the pull windings, so the hold windings keep everything going until the ignition key is released. The problem with this design is all the light-years worth of wiring.

My theory is that the resistance of all that wire increases with age. Thus, when these vehicles were new, the system would work fine for years and years. The reason replacing the starter with a rebuilt one does not solve the problem is that these rebuilt starters have old wiring in them. When Bosch rebuilds the starter, the motor section just gets new brushes, bushings, and maybe an occasional armature and solenoid. And since these vehicles have been out of p for so long, brand new Bosch starters of this design are no longer available.

Our first solution was to replace the starter with a brand new one of a completely different gear reduction design, often referred to as “high torque." The gear reduction concept itself is not new—Chrysler products have been using the design since the '50s. If you're a true gearhead, you probably always wondered why a 1966 Dodge Dart has such a wheezy sound when it was cranked over to start. Well, it's the gear reduction starter. By using a gear reduction principle, a smaller but faster spinning (4.44:1 in this case) electric motor can produce more torque with less current (the GoWesty starters produce between 40-75% more torque than the original Bosch starter).

Fast forward to 2024 and all current vehicles have switched to a permanent magnet-type starter. This design fundamentally changed the way the starter works by eliminating the many miles of copper wire used to generate the needed magnetic field. It is a much simpler, stronger, cheaper, and more reliable design.

We finally found a company that could make one for our beloved Buses and Vanagons, albeit only for the manual transmission application—with no plans in sight for an automatic transmission starter.

If you have an automatic transmission in your Bus or Vanagon, we still have you covered with one of our gear reduction starters. And you can bet we will continue to push the manufacturer to make one for the automatic application as well.

Stay tuned!