Farmall A starter gear recoil spring removal

Farmall Super A: The small spiral recoil spring for the starter's pinion gear is too weak to push the gear back after starting. This is ok when the engine starts, but when it doesn't, the gear jams and the starter must be removed. Unhandy. Cleaning and lubing the spiral shaft that drives the gear to the flywheel doesn't work. The recoil spring is simply too weak. I'd like to replace the spring (I don't see a need to replace the entire final drive for want of a spring) but can't get the spring off. The sheet metal collar at the nose end of the final drive has no apparent way to be removed. (The thin spiral recoil spring goes up into this.) Anyone know how to gain access to the spring/remove the collar? I have not yet tried removing the bolt that attaches the Bendix torsion spring to the final drive.

It is usually just a part of the problem to work on the spring. A solution that is proving to be better is to use a clutch type starter drive. This also has a modest spring, but is far less prone to jamb. CaseIH and others have them, they are a direct replacement. Jim

James - That light spring is not intended to return the starter pinion to "home". Rather it is to keep the pinion at home and prevent it from rattling while the engine is running.

If the starter is periodically jamming in the ring gear it means either the starter pinion or the ring gear are worn. Sometimes installing a new pinion will solve the problem. More often however it also requires a new ring gear to correct.

Parts are relatively cheap - less than $100 will get both a new bendix and ring gear. The bad news is you gotta split the tractor to replace the ring gear.

Thanks BobM and Janicholson for both of your answers. I had found nothing in a search yesterday, but today got your mssgs and also found the following links. Bob, I can understand the purpose of the light spring as being to keep the pinion gear in the aft position while the engine is running. However in the special case where the engine does NOT start and the non-starting engine does not fire to kick the Bendix pinion back down its spiral to a resting position, there is no other mechanism in place for the Bendix pinion gear to be disengaged from the flywheel gear. Rocking the tractor in gear (to take any load off the pinion gear to allow it to slide) used to work, but doesn’t now. Off the tractor, on the bench, with the armature level, if I gently spin the pinion gear forward up the spiral to where it touches the collar, then release it, it stays put and doesn’t move. If I move it back from that position about an inch, the spring is strong enough to continue to push the pinion back to where it belongs.

I agree that any dings on the pinion and/or flywheel will complicate proper operation, and those parts in any 60 year old tractor are likely not to be perfect. I realize I may have to replace one or both, or, short of replacing, dress out rough teeth on one or both. But I think the links below contain some useful suggestions. From those posts it appears that there are three types of starters available: 1) The original Bendix starter final drive, where the initial rotation of the starter armature, acting against the spiral spline of the pinion gear, drives the gear forward and into mesh with the flywheel gear. The original Bendix starter also has a heavy, flat spiral spring that transmits torque from the starter armature to the starter final drive and acts as a crude shock absorber. 2) A (poorly named) “clutch drive”, where the starter pinion is moved forward on the starter final drive shaft to engage with the flywheel by a solenoid. 3) A “clutch drive” that uses some rubber pucks or flexible couplings to absorb torsional shock between the armature and the final drive.

There seems to be nothing terribly wrong with the first design; they’ve worked for 50+ years. But in the first the pinion is rotating and the flywheel is stationary at the instant of engagement, and that suggests wear will occur. In the second design the starter armature does not begin to rotate (is not supplied electricity) until the pinion engages the flywheel ring gear. Any metal-to-metal contact results from the speed and strength of the solenoid’s action, and that would be less than a Bendix type. The third type uses the same non-rotating engagement as #2, but reduces starter windup torque shock. So while #1 seems adequate, #2 is a better design, and #3 is best. One poster says the “clutch type” starter is available only from Case IH, but doesn’t detail I this is a type 2 or type 3.

At this point I would still like to simply replace the spring. But if that doesn't work, and if costs are similar, it seems #3 would be the best replacement choice. Any comments?

I now see that as a first time poster I can't include links so those will have to wait, but they were in nnalert and ytforums

Several points to make:
The big coiled flat wound spring should be the return spring for the drive. This spring is pretty heavy, and connected with two bolts into the drive collars. The little spring is not a return spring.
No driving force is put through the spring, the drive force is only in the spiral grooves. The little spring is a noise reducer, keeping a little tension on the pinion gear when retracted.

Farmalls in the Letter series, and X00- X50 number series. never used a starter with a solenoid engagement. The X00-X50 series used a starter relay, but it did not pull or push linkage to make the drive move on its shaft.

The clutch type starter drive is way different, it uses a ratchet style one way clutch in the moving part of the drive. Ebay item # 150502997216 is an example of what they look like. They solve many issues particularly jamming. they also have a little spring, but it works.

It operates similar to the type 1, but the internal one way clutch allows the tractor to overrun the drive.

The worn drive on yours is binding at the flywheel and has gotten all the value out of it.(opinion) don't mess with it it is toast.
Jim.

Thanks for your reply Janicholson and clarification on solenoid type starters never having been used on this tractor. IH/Case said the same thing. It should be obvious by now that if I just wanted to get this tractor running I would have bought a new part by now. I had found the part you noted from Ebay (which BTW IH/Case doesn’t even list). But I want to understand how the starter works too, which helps one figure out why it doesn’t work. If I was going to buy a new part I’d have to take off the old one anyway so I removed the nose casting of the starter and removed the fixing bolt on the end of the large flat spring nearest the pinion gear. Hint: Don’t allow the (other) brush end casting to come off to avoid having to figure out how to hold 4 brushes out while you put the armature in (use wires for that). See photos.

Once that bolt is removed the “tube “ type shaft containing the spiral grooves, the pinion gear, the light spring, and the thin metal cap that retains the spring are free to rotate around the (inner) armature shaft which is continuous from the bushing at the brush end to the bushing at the nose end of the starter. That shows that the torque from the armature to the pinion gear is indeed transmitted only through the large flat spring. That’s why other posters have said that diesels break these springs more often than gas engines—the starter has to transmit more torque. When the starter is engaged, the spin-up of the armature shaft working against the inertia of the pinion gear drives the gear up the spiral until it engages the flywheel teeth. When the engine starts it spins the pinion back down the tube shaft. When it starts. When it doesn’t then something needs to move that pinion back down the spiral for another try. The large torque spring may have some windup tension in it that might help retract the pinion, but in my case that spring is in good condition and the pinion isn’t disengaging. But I do realize that rough flywheel and pinion gears may be preventing retract too.

The “tube” shaft should have, but would not, slide off the end of the armature shaft. Wear over time on mine had created a tiny ridge at the end of the larger diameter portion of the shaft near its left end. I filed that ridge off and the tube shaft assembly slid off.

Before I give up and buy the part you mentioned I’m going to do three things as an experiment. 1) Replace the small spring with one that is a little, but not a lot stronger than the current one. 2) Clean and polish the armature shaft and inside of the tube shaft. 3) Dress the gear tooth noses.

I had previously not known how to remove the metal end cap retaining the small spring. Once I got the tube assembly off and cleaned up it is clear that the cap is retained by four punch dents deforming the end of the tube shaft. I’ll de-dent those, get the metal cap off, and find another spring.

The pinion gear tooth noses look too beat up to allow easy engagement so I’ll dress those with a Dremel so they are consistent in length and have clean angled noses like new ones.

Thanks again for your help. I do appreciate the responses.

With real respect and admiration for your efforts to understand the starter operation, I stand behind what I said before. The torque on that spring would make toast of it if it were transmitting power. The sleeve that has the external spiral grooves in it is driven by the motor shaft. It is driven so that the pinion's inertia causes it to spin out on the grooves and engage the ring gear. If the sleeve was free to rotate, nothing would move into mesh. The gear stops moving into mesh when the gear finds the end of its travel (grooves and end cap) the flat wound spring is torqued to a slightly smaller diameter and stretched longer in the process.
The retraction happens because of that big spring.

Your starter drive is worn in a way that causes it to jamb in the ring gear teeth. The loosening of the starter frees it to snap back (yours might be worn enough that it is not even doing that)

I am certain that if it were in gear and not jammed, it would spin the engine. The fact that it does not is defacto proof of the jamming.

There is no looking inside the housing, so direct evidence is tough to come by. I have had this situation happen on 6 tractors. It is always progressive and ends in removal to free the drive.
(even putting a high CCA 12v battery on a jammed starter will not make it spin.

Cleaning it up and putting a bit heavier spring on the end might make it work a few times. It has for me, but I have found the time to mess with it when the issue is a better design makes reliability and monetary sense.

If you bench test the starter with it in a vice, chain wrench, or clamped down firmly, with a good battery and cables, it will show you what I mean with a violent snapping out of the pinion against the end of travel. Do hold it down with restraints, it will spin the case opposite of armature rotation with massive force.

Jim

Jim,

Thanks for your reply. I have figured it out, though I still differ with a couple of your details, and that is based on my observations this morning. If you’ve had this happen on six tractors, you have a lot more experience than me. Probably 10 times as much because I find I never manage to do repairs just once. So I take your reply very seriously. You made a great point that the starter had jammed, as opposed to meshing correctly and then not recoiling correctly. I hadn’t thought of that simple conclusion. If the pinion was meshing correctly, the starter would have turned the engine over, and it wouldn’t in that situation position. The pinion HAD been jammed.

I was going to say you were wrong here: “The gear stops moving into mesh when the gear finds the end of its travel (grooves and end cap) the flat wound spring is torqued to a slightly smaller diameter and stretched longer in the process.”

After looking at the thing again and waving my hands in circles to get the rotations right I agree that the big spring does act to retract the pinion. The pinion rotates on its spiral about one full turn, which would be the max spring wind-up. If the engine fires, the flywheel will overrun the pinion and shoot it back down the spiral. If the engine does fire the spring will do the same thing. I agree the pinion locks to the final drive shaft only when it reaches either end of its travel at the end of the spiral.

But after reading your last post first thing this morning I went back to the shop, went over all my assumptions, thought about your post, slid on the starter final drive, and looked for all possible ways that it would be forced to turn by the armature. There aren’t any, except the spring. Without the forward spring attach bolt the hollow final drive shaft and its pinion spin round and round freely. The aft face of the (removed portion of the) final drive is flat (at the other end of the final drive in my first pic), and the forward face of the bushing with the aft bolt still in it in pic 2 is flat. No cogs or such. The forward spring attach bolt, the one I removed to take the pictures, COULD have gone all the way through the final drive tube to nest in a pocket in the armature shaft, locking the final drive to the inner shaft. It doesn’t. There is no pocket in the armature shaft. The final drive tube has a 5/16” thick wall, the spring is 3/8”, and the lock-tab washer under the bolt is 1/16”, equaling ¾”. The fixing bolt is 11/16”, 1/16” less than needed to make any contact at all. If you said “It is this tab, or cog, or projection, or bolt, or roll pin that locks the forward part of the final drive to the armature”, and then I found it on mine, I’d have to agree. But you haven’t said that yet and I can’t find anything that would function like that.

I said I figured it out, but didn’t say why. It became clear when I removed the aft spring fixing bolt, which I hadn’t done til now. The aft spring fixing bolt DOES extend down into a pocket in the armature shaft, locking the after part of the big spring through the light colored bushing there to the armature shaft. Which is exactly where the armature shaft had fatigue-cracked and broken in two. I had noticed some axial play in the shaft at that aft bushing before and thought that having a bushing connect two halves of an armature shaft was a crummy way to engineer that part. Hard to keep the shaft straight over its length. But I never guessed the armature shaft could be broken and the starter could still work. Duh. Which kind of explains why the starter was always a bit, uh…..slow and loud. Though it still did work, except when it jammed. To be clear, though the armature shaft had broken, the two halves had never rotated relative to each other. The bolt in the shaft pocket (halves) still held them in register to each other, and the fracture face showed no rotational wear.

I can’t explain the difference between my observations here and your experience. Are we talking about two different types of starters? Doesn’t seem likely. But I could have bought a new final drive thinking it would solve my problem and found out later it was no solution. If you want more pics I can post them. Fortunately I have a crank to start the A, which is why I have this tractor and not something new where we would be talking about computers, which are less rational and predictable than broken shafts.

I had mentioned another post somewhere talking about how the big spring DOES transmit torque to the pinion. He may be wrong too, but here’s the link and that specific post:

http://ytforums.ytmag.com/viewtopic.php?t=553535
Fri Nov 06, 2009 1:36 pm Just an FYI. Those "replacement" drives are actually trying to solve the torque spring problem. The Bendix drive has that hefty flat wire coil spring that connects the motor to the drive pinion, and it tends to break. All the gas over diesels guys can tell you about it in detail. Leave it in diesel and try to crank it and it may be your last attempt at tractor driving for the day. Nothing a $100 and a couple of hours can't fix, if you can get the replacement spring. The heavy engagement of a 12V on a gasser may do the same thing over time. The spring takes a beating and eventually cracks. PS: The later Bendix style drives got rid of the coupling spring.

Anyway I don’t expect to change your mind; experience is a strong teacher. I do appreciate your posts. But I think I’m in the market for a new starter.

Jim

OI just did a bunch of google research on the drive, then went and looked at an example I have in my garage. (it is not from a Farmall, but uses the same design.
The spiral sleeve is definitely attached to the armature shaft. The pinion is attached to the spring, but the spring only retracts the drive, the end of the grooves limit the travel of the pinion into the ring gear. If the spiral splines on the tube were free to rotate with the pinion, there would be no reason to assume the pinion would move out to engage at all.
I think yours is broken where the sleeve attaches to the drive collar closest to the armature, where the bolt that is long enough to drive the collar is located. It could have broken off at that point due to the forces that were strong enough to snap the shaft at the bolt hole.
It is a great discussion. Jim

Jim,

Could you take a picture? Is there any detail I could photo that would clarify whether mine is the same, or differs in some way from yours?

The one I have is neither the same size, nor of the same Farmall type. So the comparison would not be valuable. Inertia (Bendix style) drives have been used on everything from aircraft to boat motors. A simple way to solve the controversy is to look at (and hold in your hand) the drive at NAPA, or other store. If the sleeve with spiral grooves is attached to the hub with the long bolt, it is as I am convinced it is. If it is free to turn with the pinion I just can't understand what makes the pinion spin out on a spiral that can spin with it. I think almost every part of your drive is grenaded. The investigation is always worth the effort. Jim

Jim,

Last pics should be enough. Layout is roughly as it would be assembled. Other than differences you can see, things to note:
The bare shaft is the forward (now fractured off) end of armature shaft with the fractured end on the right end. Pocket for aft spring retaining bolt is rotated down to highlight key and the edge of the pcket (where it fractured) is just barely visible as a tiny white spot in the second pic. This shaft inserts into the right collar about as deep as the darkened end of the shaft. The pocket and fractured end of the aft (right) armature shaft is just barely visible inside the bolt hole of the right collar. The detached (left) portion of the armature shaft is smooth over its length, with the left end diameter reduced for the nose bushing. The inside of the assembly bearing the pinion is also smooth. The faces of the two collars are smooth. The right spring retaining bolt has an extension that goes into the armature shaft pocket to fix the right collar longitudinally on the armature shaft, though torque is transmitted through the shaft key and keyway. The left bolt is short and does not contact the armature shaft. The only way for the armature to transmit torque to the pinion is via the spring. The engine turns counterclockwise viewed from the back, so the starter must rotate clockwise, with forward being to the left in the photos.

How does this differ from yours?

Jim

A the university I looked back into books I had when I was in mechanics school. Answer! We are learning:
The drive spring is the only force driver involved in the link between the pinion and the armature shaft. It does get smaller in diameter as a result of taking shock loading of the pinion teeth in the ring gear. You are correct.

The weak spring assures that the pinion stays out of engagement when the cranking stops (due to either the engine overspeeding the pinion pushing it back on the spiral sleeve, or the starter motor shuts off with no starting of the engine. The sleeve is designed with either groove limits, or a stop (like yours)that force the pinion to turn with the sleeve when it reaches this limit. THe forces are contained to the sleeve.

There are several designs of this type of system, and applied to many different engines.
I Learned also the new style has a shock absorber internal to it, and does not use the big spring.
It is more reliable because of the reduces number of parts, and the more solid positioning of the drive pinion to the ring teeth.

If you purchase a used starter with good armature, and a new style drive, it should be magnificent.
A 1970 motor service -Automotive Encyclopedia- provided the context in which it was made clear (along with your good photos.)
Thanks for the enriching discussion.
Jim

Glad to see you have found differences between the two starters we each were describing. There must be many variations on the theme. Thanks for the replies.

Jim