Delco Starters -- Anyone familiar with them?

[Averagesupernova]

So I am working on a 24 volt starter from an old International track type dozer that has been in working condition. To start out with when I hit the starter button nothing happened. I suspected a poor battery terminal connection but that turned out to not be the case.
-
What happens next is this: I hit the starter button, the starter takes off, but I did not have the throttle open (fuel is cut off) so I release the button but the starter keeps turning. Ugh! Not about to pull a battery cable off on a 24 volt system drawing this kind of current. After about 20 seconds it quits. I notice one of the battery terminals had dripped some molten lead off so I assume the connection is broken and I safely disconnect the batteries. I had wondered if low system voltage had caused the solenoid to stick so I go about repairing the battery cables and hook it back up also having replaced the starter push button switch and installed a battery disconnect. I try it again and sure enough the starter continues to run after I release the button. I quickly pull the disconnect and naturally the starter comes to a stop. But yet again when I reconnect the disconnect the starter does not turn and when I hit the starter button the same thing happens again. The starter turns and releasing the button does not stop the starter. It seems to be latching itself on.
-
Fast forward to having pulled the starter off the machine and disassembled it in the shop. If find the inside of the starter looks excellent. The brushes and commutator are in fine shape and everything is very clean. However, inside the starter solenoid there are some heavy contacts that are quite worn. For those who know, I don't need to explain how this works but for those that don't I will explain it. Solenoid current causes a magnetic field that pulls on a plunger which pulls a rocker arm which slides the gear ahead on the starter shaft that engages the ring gear on the flywheel. The other end of the solenoid pushes a conductive disk against two heavy contacts. One contact connects to the battery and the other connects to the starter motor itself. When the solenoid plunger moves it completes the circuit causing the starter motor to turn.
-
So I just so happen to have a pair or contacts on hand and the disk to replace the worn ones with. I had assumed that the old contacts were sticking as they were badly burned. Guess what? No change in operation. I tested it with a battery and pair of jumper cables. The starter latches itself on.
-
I had a little experience in the past with starter solenoids but kind of rusty on how they are wired. Here is the deal as far as I can remember: There are actually 2 coils. One is a holding coil and the other is a pull in coil. The pull in coil draws a fair amount of current and the holding coil considerably less. The holding coil is wired how you would expect it. One end of the coil is grounded and the other connects to the starter push button. If I recall correctly the pull in coil is wired one end to the starter push button and the other end to the starter motor terminal. In other words, the pull in coil is bypassed when the plunger pushes the conductive disk against the contacts that complete the circuit to the starting motor. Very clever.
-
So my mind is spinning now as to how this can work. Once the starter is engaged wouldn't there be feedback through both coils that would cause the thing to latch on? What is it that prevents this? Unless I am not remembering this correctly it would seem that how this starter is acting is exactly how I would expect it to yet I know there are millions of starters out there designed like this that work flawlessly. At this point my assumption is that there is a short between the two coils. Why the thing does not latch in normal operation I assume is a combination of cancelling fields and creative positioning of the coils.
-
@jim hardy this sort of thing seems to be right up your alley. Maybe not on an International TD20 dozer but I would guess you have had your share of starters apart.
-
Any ideas? Anyone?

 

[jim hardy]

I'm trying to remember the old Ford starters.
What you describe sounds actually like mechanical binding.
See this video

Sounds like something holds the gear extended keeping the big contacts made.
Are both springs nice and free ? Does gear slide back when you let go? Friction someplace would keep gear extended as it runs out along the helix.

Still trying to remember last starter i had apart with that kind of 'engage' solenoid. Seems it was a Ford that had an external solenoid too. I had opposite problem - gear engaged but it never got power to armature.

As you suggest, if it's wired as you remember current could flow from pull-in coil up to pushbutton and back down through hold coil.

will sleep on it.

 

[jim hardy]

hmmmm another interesting Delco video

 

[Bystander]

hmmmm another interesting Delco video

" ... protects the starter from costly 'stuck contact' situations caused by low system voltage." Are you certain you've got 24 volts? Reason I ask is because Sears sold me a brand new Diehard that died soft a couple years ago and nearly set my car on fire (the melting) and almost stranded me south of Pueblo.

 

[Averagesupernova]

It definitely seems to be an electrical latch situation. The mechanism moves freely. The new contacts are not sticking due to low voltage. One thing I forgot to mention is that upon disassembly of the solenoid I notice a smell that I assumed was burnt contacts. But now I remember where else I have noticed that smell. A coil that I replaced in a 7.5 HP motor contactor years ago. I suspect the varnish insulation on the windings has burned.

 

[jack action]

This guy had similar symptoms and the solution was not electrical:

FINAL SOLUTION: MY BUDDY "JR" (COMMANDER OF VFW POST 1388) FIGURED IT OUT. THE SPACER IN THE NOSE END OF THE STARTER WAS GONE, CAUSING THE BENDIX GEAR TO EXTEND TOO FAR INTO THE FLYWHEEL. THIS CAUSED IT TO HANG UP IN THE ENGAGED POSITION AND AS LONG AS IT'S ENGAGED, IT WILL CRANK UNTIL THE BATTERY DIES!

 

[Bystander]

new Diehard that died soft

Tests/ed "good," any load reversed cell polarity; i.e., nominal twelve cranked at eight, and you might want to check loaded output voltage for twenty rather than twenty-four.

 

[jim hardy]

So you say it doesn't retract even when tested on the bench ?
I'd find a way to pull the solenoid into engaged position by hand instead of electrically and see if it disengages.
I think it's moving too far or there's insufficient spring tension .

 

[Averagesupernova]

When the holding coil connection is removed then only the pull in coil is connected. So what happens is the pull in coil pulls the plunger as it should. When the disk makes contact and completes the circuit so the motor starts then the pull in coil is bypassed. With nothing to hold the solenoid in place the solenoid releases the plunger which of course places the pull in coil back in circuit. The solenoid is again activated and the cycle starts over. It will cycle like this until I remove power from the terminal on the solenoid that would normally go to the start button on the dash. This proves to me that the mechanism is not physically sticking.

 

[dlgoff]

I'm thinking the spring constant of contact closure spring may have decreased from heat/aging etc. causing the copper contact rings to arc (due to lack of contact speed/closure force) and fuse.

 

[Averagesupernova]

Bought a solenoid and put it on and the problem is gone. Just the solenoid and electrical contacts. Not the plunger or linkage to the gear.

 

[jim hardy]

That burnt varnish smell probably tells the tale.here's my hypothesis, sorry i didnt get it worked out earlier. I had to draw it out and I DREAD PAINT!...

(Pardon my typo Ttop means top...)
Draw in my hypothetical short and work it in your head. I think i got it right..,.Can you separate the coil wires and check for shorts?
If they're shorted , put an amp of DC through one coil and measure millivolts from other coil to both ends (of the one with current ) to estimate how far into the windings the short is.

Congratulations on a fine job of troubleshooting , My Good Man !

My neighbor Harry has three such starters on his "fix me" pile. I'll get a look inside one.

old jim

 

[Averagesupernova]

@jim hardy
-
I agree with your drawing and description of how this works. Fields cancel, etc. I have had a hard time accepting that these can work due to the difference in current draw between the 2 coils. However, when they are wired in series fields opposing there is nothing to say that the fields will not balance. The pull in coil will have larger wire but that is not to say its field is not the same as the holding coil when they are in series. A clever setup indeed.
-
As popular as these starters are I would think that somewhere along the line this problem would have cropped up on google. Nothing jumped out at me.
-
I did not keep the old solenoid. My dad likes to hang onto too much old crap the way it is so I let it go. He will be happy shoving trees tomorrow though.

 

[jim hardy]

The pull in coil will have larger wire but that is not to say its field is not the same as the holding coil when they are in series. A clever setup indeed.

yes. it's amp-turns that count. . I would make them have same number of turns but hold coil of much smaller wire. If Harry has a burnt up one i'll see if it's accessible.
It is clever.
Pull in coil can have huge current when it's connected to battery through start pushbutton, but hold coil will be just 24V/Rcoil.
When start PB gets released the two coils fall into series so current through them must become the same. They're backfed from the disk but if #turns on them is same they'll have equal and opposite MMF's so flux drops to zero.
That postulated short allows unequal amp turns when we need them equal.

(sorry - just rambling , i always quadruple check myself.)

He will be happy shoving trees tomorrow though.

My 'Dozer Envy' is becoming acute .

old jim

 

[Averagesupernova]

My 'Dozer Envy' is becoming acute .

old jim

It gets old in a hurry. Although I would rather have spent the time operating it than pulling and replacing that starter. We used a strap over the hood as a hoist with one guy on the strap and the other on the starter itself.
-
It's kinda funny. We are all kids at heart. Most kids watch something that they think looks fun and beg to do it. Without fail, when they are allowed to it soon turns into work and they get tired of it. As an adult with 'big toys' it is no different. When the novelty of being able to push large things out of the way with ease wears off, it is a noisy, dirty, boring job.

 

[Averagesupernova]

This one was a tough pull. We used a cable attached about 10 feet up. No way to dig at roots so we just tugged. Normally when we push trees out a shallow pit is dug on the side you are pushing from. You cut roots and you can lift on those roots and push at the same time which gives a lot of traction. Under no circumstance do you want the wad of dirt attached to the roots to get under the blade. When the tree goes over, the trunk can get under the blade and the dozer is lifted and gets hung up.

 

[jim hardy]

I had an AHA! moment in the shower...

To start out with when I hit the starter button nothing happened. I suspected a poor battery terminal connection but that turned out to not be the case.

So why didn't the starter turn ?

I had assumed that the old contacts were sticking as they were badly burned. Guess what? No change in operation. I tested it with a battery and pair of jumper cables. The starter latches itself on.

Hmmm contacts were burnt ?

Another hypothesis
Those contacts get burnt spots from the heavy current. I'm accustomed to them being a flat disk that rotates slightly with each actuation so as to 'wipe' the contact clean. When it gets rough spots it'll stop turning and black slag builds up until it no longer makes contact.. I've filed many of them back smooth.

So that time when you pushed 'start' and nothing happened..

might this explain all ?

The coils are separate in that drawing but in the videos they appear tp be concentric,
of course they both surround the same solenoid... So a short between them is plausible..

old jim

 

[Averagesupernova]

The very first symptom was absolutely no action of any kind when I hit the button. If it were a contact issue I would have heard the solenoid move and I did not. I suspect the original problem was the solenoid winding.

 

[jim hardy]

The very first symptom was absolutely no action of any kind when I hit the button. If it were a contact issue I would have heard the solenoid move and I did not. I suspect the original problem was the solenoid winding.

Okay, scratch that idea.

We understand it better now and that's what counts.

Thanks for letting me play in your sandbox,

old jim