How much starter shaft end play should there be? We have a rebuild starter where the bendix hits the flywheel and there is zero in and out play on its shaft. The starter came with a distorted spring, after replacing it with a new old stock one the bendix still hits, but it is a little better.
I opened the starter and found two thrust washers. Is this correct? I understand the starter has been fitted with an oil seal and one of the washers was perhaps the old bushing end machined off to work as a thrust surface. If the thinner of the two washers is removed and the shaft is pulled out fully, the brushes remain on the edge of the amateur, but still making full contact. With the thinner washer removed, the bendix no longer hits.
The starter shaft should have virtually no end play. When the starter is actuated, the bendix gear moves backwards towards the starter ring gear. If the bendix gear is not engaging, you have other issues. Does the bendix drive gear have the small pin and spring installed on the underside of the collar?
Apologies for not being clear. The Bendix engages and turns the motor over fine (all the proper parts are in place). The problem is the Bendix gear is hitting the flywheel when it's retracted. A new spring helped, but it still hits.
Should there be two thrust washers?
It's a little hard to picture the problem, but as I see it the problem is this:
The starter itself, exclusive of the Bendix assembly, seems fine. there is little or no end play in the starter shaft (which is correct).
When the Bendix is installed and the unit is mounted in the engine, it works correctly except that when the Bendix retracts (which means the gear is as far from the starter motor as it can get), the gear still contacts the flywheel's gear.
If I understand it, moving the thinner spacer from the front of the starter to the back, cures the problem of the gear hitting the flywheel, and everything seems to work correctly.
The potential problem with moving the starter shaft in or out would be that it is important for the brushes to have full contact with the commutator, and not ride off the end, because there's a tremendous amount of current flowing through that junction, and full contact is important. But I think you said you have tried moving the thin spacer, and the brushes do in fact still have full contact with the commutator.
All that being correct, I'd say the thin spacer belongs at the brush end of the armature.
But if moving the armature by relocating the spacer makes the brushes run with even a little of their length hanging off the commutator, then don't do it, and look for some other fix for the problem.
Might there be two gaskets or too thick a gasket where the starter mounts to the hogs head?
What does the rebuilder have to say about the problem?
Thanks, Peter. You summed up my problem well. With the thinner bushing moved, the brushes ride exactly flush with the edge of the commutator with zero overlap.
Ted, I acquired the starter through a friend. The rebuilder is unknown. There is only one thin gasket on the hogs head. Thanks for the idea, though.
That whole armature looks odd. Were the later ones made with a shaft like that? All the ones I have see so far just have a smooth stepped shaft not that area that looks like a forging in the center with a lip to retain the bendix. After market or home made?
We solved our problem with advice and guidance from Ron P.
The previous rebuilder upgraded the brush-cap bushing with a ball bearing, but neglected to install a spacer. They then removed the armature end-play by adding an extra shim at the rear bushing (Bendix end). This shifted the armature too far forward which caused the Bendix to interfere with the flywheel.
The solution was to install two machined heavy brass washers under the ball bearing and to shim the armature rearward. We now have the correct end-play of .003" - .005" and the brushes are making full contact with the commutator.