interesting starter behavior -- opinions please

[tmctguer]

after wrastlin' with the top nut of the starter on my '89 911, i finally loosened it and was able to remove my starter.

in an earlier post, i detailed that i could hear the solenoid actuate, but only intermittently would the starter engage and turn the motor. most of the time, i would turn the key and get nothing.

suspecting an undercharged battery, i took it to my local parts store to have it tested. it tested good. that put me on the starter path.

once i freed up the starter, and removed it from the flywheel, i left all electrical connections on, reconnected my battery, and had my assistance (a.k.a., girlfriend) turn the key while i watched the starter. each time she turned the key, the starter turned, but sparks flew out from the mechanism. i was sure that my connections weren't grounding to the chassis, then watched sparks fly a few more times when the key was turned.

i surmised that there is a short somewhere in the starter motor, and under load (i.e., trying to turn the flywheel), it shorts out and won't turn (or would intermittently turn). however, when not trying to turn the flywheel, it shorts but still is able to turn itself.

what do you think of my theory?

clearly, the starter should be throwing sparks, but is it more likely to just be "dead" rather than turning & sparking?

i ordered a new one and i'll report next week if my problem goes away when it is installed.

[Jim Sims]

Sparks equal failing brushes?

[tmctguer]

i forgot to mention that the car has 3,400 miles on it. it is an extremely low mileage car.

[jester911]

An 89 with 3,400 miles? Wow, must have been in a museum.
Post some pics of that baby.

[DCN 911]

In addition to the pictures, I would love to hear about the background history of the vehicle and how it ended up in your life.

Dean
88 911 Coupe

[Tim Polzin]

Unless you have a ground strap attached to the starter, could you be seeing sparks from where the starter body is casually grounding?

Tim

[HotPorsche]

Could be maginal ground on starter. While not under load it runs OK and when cranking the engine the ground breaks down and the starter does not get sufficent current. If you can get a DC meter on the starter connections you should see 0V in the ground side while cranking. If you see some voltage there your ground is bad.

[Quicksilver]

To keep from seeing sparks I would expect that you would need a very, very good ground. If you are just holding the starter up against the tranny or something, the torque of the starter spinning up is going to make it impossible to hold still.

A starter under load is an unbelievably overloaded device. It will pull close to 1000 amps as the engine is spooling up. It is so overloaded that if you crank the engine for 15 seconds you can feel the battery cable heating up. A lot of things need to work correctly to make it crank the motor.

I suspect that the contactor inside the solenoid has a problem. If you run jumper cables directly to the starter's input and casing I would bet it spins fine every time. (bypassing the solenoid)
With the low miles you describe, I can't imagine that the brushes are worn out. They may have a problem with how well they contact the commutator. I would let the starter spin freely for a minute so the brushes get happy with the commutator again.

Getting back to what I suspect is the real problem... The solenoid has 2 bolts that stick out of its end that are the major power leads. One connects to the battery and the other connects to the starter's input lead. Inside the solenoid these bolts have a square flat head. On the end of the solenoid's plunger is a copper disk that hits both of these bolt heads to make the circuit. The disk spins freely to keep the copper disk's surface from eroding too much from the arc and failing.
The problem is the heads of the contactor bolts do erode quickly and they have no way to renew themselves. If you can open the solenoid you can fix this by simply reseating the bolts after turning them 180 degrees. The disk in the solenoid only contacts 1/2 of the bolt's head so rotating them gives you a fresh contact surface. On my 87 I wasn't able to get the end cap off but I can't recall why. (I might have been lazy. I just don't remember.)

I would bet you real money that the problem is the solenoid. If you are getting a new starter get the hi-torque starter.

It is a much better unit and weighs less.


Wayne

[tmctguer]

after reading the replies, i realized that i had disconnected the ground strap when lowering the transmission to gain access to the starter. i DID see that the ground strap and the connection to the chassis were in pristine condition. not surprising for a car with only 3,400 miles on it. (what is surprising is that the starter failed !!)

perhaps the sparks were from the housing grounding to the bell housing, and not the internal starter mechanism??

i wonder if i should re-connect the ground strap, re-connect the starter, and be extra careful not to have the connections close to the the chassis or transmission housing? perhaps that would stop the sparking.

i was intrigued that the starter motor turned, which it would only intermittently do while installed in the car.

[tmctguer]

i received my starter on friday from our host, and replaced it this morning. the car fired right up, so all is well.

thanks for everyone's replies.

[RoninLB]

Originally posted by tmctguer
so all is well.

.

thanks for the up-date