Hopefully you guys can help me out here. Sorry for the long post.
Last week my brother picked up a '86 Dyna (supposedly). It has a RevTech 88" motor and a S&S 5 spd. trans. It had a major oil leak out the inner and outer primary. So I pulled the primary cleaned it all up and replaced all the seals and gaskets. After putting it all back together, as soon as I turned the key on and hit the start button smoke began rolling out from the battery area. It turns out the cable from the starter to the solenoid was very hot and smoking like it was shorting out, it actually started sizzling the battery. I disconnected it and it was fine and everything cooled back down. As soon as I reconnect it the starter to the solenoid it started baking again. I pulled the starter and took it apart to see if anything looked out of place and didn't see anything. I reinstalled it last night and tried it again and it worked fine. I did some other work and went to start it later and it did the same thing, began cooking the cable from the starter to the solenoid and the battery again. I'm stumped! Typically on a car the lead that goes to the starter from the solenoid is a positive lead that's isolated from the starter case (which is grounded). I know usually to test a starter you can hook a jumper from the - to the housing and the + to the lug and that should crank the starter. But I hooked an Ohm meter to this starter case and the lug, and the lug is grounded to the case, so that doesn't make sence to me. I bought a new starter and the lug is also grounded to the case.
The way it's wired now, + from battery to solenoid, solenoid to starter. The ignition wire on solenoid should connect the batt. with the starter cranking the starter. But if the starter lug is grounded it would instantly fry everything once the ign. is pushed.
What am I missing? Is the starter cable really supposed to be grounded? Could the solenoid be bad?
I found a wiring diagram on this site and I have everything wired right.
It uses the old Shovelhead starter with the solenoid that uses a jack shaft.

Thanks for any input!

 

Went back out this morning to work on it some more, and found out whatever the problem is fried my battery. Off to get a new battery.

 

The starter motor is probably not shorted.
Connecting an ohmmeter to it measuring the resistance from the positive 12v connector and the frame of the starter motor will show little or no resistance. All that is in there are four carbon brushes (they have very low resistance) and copper wire.
Doing a bench test, if the starter runs when you hook a battery up to it alone (neg. to frame of starter motor and pos. to lug on starter motor), it's okay. Your problem lies elsewhere than the starter if it tests good.
Was the cable from the solenoid to the starter motor hot as well? If it was then I'd suspect something in the starter motor but it probably wouldn't bench test good either. It may be that something is wrong with the solenoid. I'd suspect it if the cable from the battery to the solenoid got hot but not the one from the solenoid to the starter motor.
Bummer about the battery frying.

 

The bench test did show the starter was good.
Both leads going to the solenoid did get hot, but the lead going from the starter to the solenoid was very, very hot. I couldn't touch it. The only thing I could think of is there was no jumper wire grounding the starter body to the chassis and I did add a thin cork gasket between the starter body and the primary case. I don't really think this could cause it but who knows, maybe. I added a jumper wire and will try it again tomorrow when I buy him a new battery. Oh yeah, also there was no relay for the starter when I tore into it, but I did add one.
It just seems like the + lug on the starter is directly grounded to the chassis.
I'll try a new solenoid this week and see what happens.

Thanks for the responce.

 

If the cable from the solenoid to the starter motor is also hot then I would NOT suspect the solenoid. Both cables have high current flow to get hot so the obvious question is, "Is the cable from the solenoid to the starter shorting to ground anywhere due to bad insulation or positioned on the starter motor lug so as to touch ground?"
If it ain't that...low battery voltage will cause high current flow. Was the battery fully charged? One would think the starter motor would be sluggish or not turn at all with low voltage.
The longer of the two cable studs on the solenoid is supposed to have the battery cable on it and the short one has the cable to the starter motor. Is that the way it is wired? I have not wired it the other way around so I don't know if it would cause a problem but the manual does specify that the battery cable goes on the longer of the two studs.

 

I installed a new batt. today and no luck. Now the solenoid got very hot as soon as I hooked up the cables, even with the key off.
The longer lug is connected to the batt. +, so that's right. I also installed a 4 gauge jumper cable from starter case to frame. I checked the + cable from the battery and it's in good shape and not grounding. Funny thing is though, as soon as I touched the - cable to the batt., the starter very shortly engaged, and the key was off. Now that's got me thinking it's the solenoid.
I can't believe after many years as an auto mechanic at a Dodge dealership, who's built several off-road race buggies, countless motocross bikes, and a couple Harley's that it's coming to this, I'm about to take it to the local shop.

 

Now it is sounding like it is something to do with the solenoid, as you suspect.
In your first post you said the ignition wire was hooked up to the solenoid and later mentioned the ignition button.
I figured you meant the start button and not ignition. Right?
Have you checked the solenoid (with all wires disconnected) to see if the battery cable connection (long stud) is grounded? That one should not be grounded.
The starter motor cable connection will show ground because of the pull-in and hold-in windings...as I recall...it's been a while. But it could be shorted to ground. Now I remember that because of those windings, the battery cable should not be connected to it. That's why the long stud is used for the battery cable.
Do you have a "known good" solenoid to try on the bike?

 

Yeah, when I said ignition wire I meant start wire that comes from the start button.
What your saying about the windings makes alot of sence. With the only thing between the + lug (magnets) and the windings is the brushes it should show grounded with a meter.
I don't have a spare solenoid, but I think I'm just going to replace the starter and solenoid. That will eliminate both posibilities. I went out and tried it again earlier and found the starter is now dragging real bad, barely even turns the motor over with a fresh battery. So obviously that is one problem. Now I'm thinking the whole time it was the starter and just pulling so much draw that it was toasting the cable.

 

I hope throwing a new starter and solenoid at it will fix the problem. It's kinda expensive way to go.
You said you had the starter apart and it looked good inside. Are you going to pull it apart again to see if there is something you missed the first time?
I'm curious as heck what is wrong with the bike so please post what you find and if the new parts fixed it.
Are you sure all the cables are good?

 

anybody figure this one out yet? I got this same fricking problem, with a 79 shovelhead and I really can't afford to take it to a shop...

and I can't afford to fry another starter, battery, cables and solenoid either...