90 HP won't turn over with plugs in.

[jerryjerry05]

Bad starter, bad battery, bad battery cable leads, bad compression, timing way off, bad lower unit bearings,.
'​A smokey starter means it's been used. It's gonna smell.

 

[Kenny323]

I unhooked everything and put power directly to the starter and nothing else. The starter must have a short in the armature that is only noticeable under a load. I am going to order a new starter and see what happens.

 

[tommarvin]

E-Bay has great prices for boat starters, but check every inch of wire first, please.,It' could be 55 bucks down the drain.,now you have to resell a used starter on e-bay..
If the starter works , the odds favor the starters not bad under load, and It's something else.
You said you took the starter apart and cleaned it,everything looked ok, put it back together again?
Did you check all wires and cables?

 

[Jiggz]

I unhooked everything and put power directly to the starter and nothing else. The starter must have a short in the armature that is only noticeable under a load. I am going to order a new starter and see what happens.

Well, there is tool called growler to test for shorts in an armature. However, it cost so much you will be better off just getting a new starter.

 

[jerryjerry05]

You can test the armature with a regular analog meter(not as good as Jiggz test) but will tell if it's shorted.

 

[Kenny323]

Tested the armature with a multimeter. The ohms zero out on the comutator bar. There is a short some where inside that is causing a drop in voltage . Not enough to stop it from working , just not enough power to crank the engine under compression.

 

[Jiggz]

To test an armature for shorts with a multimeter, is almost impossible as the resistance of each coil is usually less than 1 ohm. You can test resistance but you will need a micro-ohmmeter which measures resistance less than an ohm and up to a millionth of an ohm. Otherwise, all resistance in the armature will read zero if you just used a regular multimeter. Additionally, you will need to identify the "span" of the coils to be able to read each coil's resistance on the correct commutator bars. It is not an easy feat nor will it be accurate especially if you are just using a regular multimeter.

Now if you want to check for "grounded coils", you will need a megger or megohmmeter which measures the dielectric strength of the wiring insulation. Again, a simple multimeter will not detect insulation failure unless it is totally grounded already. A "megger" applies high voltage to intentionally "break" insulation that is already weakened. And yes, insulation testing is a form of destructive testing like pressure testing a pipe.

In short, coil short testing is resistance testing of each coil while ground testing is the dielectric strength of the insulation, i.e. there should be no direct resistance reading from each coil or commutator bar to ground or in this case the shaft or laminates.

 

[Kenny323]

It was zeroing out to the shaft , which let me know something was touching that shouldn't be. I traced down the suspect wire, which wasn't hard to find luckily. Then I covered it with epoxy. Next I put the armature in the lathe and touched up the commutator to make sure it was nice and flat. Made sure the brushes were nice and flat , then put it back together. Works like a brand new one now.

I will still be buying a new one just in case the epoxy doesn't hold. I am interested to see how long it lasts.

 

[Jiggz]

The epoxy should hold. Although, insulating varnish will probably do better. The problem I see when using epoxy is the bulkiness required to insulate a wire. Now since we are talking about rotating wires, the centrifugal force will eventually peel off "high mass" or bulky epoxy patch. This is the reason varnish is used. Even with thin application it actually have a high dielectric strength. And because it is thinly applied, it can hold itself under high centrifugal forces. Good troubleshooting for finding that grounded wire.