Front Diff Gear Pattern - Again.

[Gerrha]

FWIW, I used to do this for a living, on a production axle assembly line. I'd send that pattern without a second thought.

A thought on printing for setup: use an old toothbrush and paint 4-5 teeth, both sides, 120° apart in three places. Don't put any more paint on the brush once you dip it initially. You can go over a tooth again if need be, but too much compund can mask fit problems. You want it thin enough that it doesn't take up room, but thick enough that it can be seen.

Two (general) things:
1» A pattern will never get any farther out; you can always move it in but never out. Gear teeth wear in; once they wear, you have to move deeper to get contact.
2» You should be able to turn the gearset by hand; if you have to use a drill, it's too tight. This is from experience building very large gearsets. Our hand crank was mounted directly on the pinion (we adapted it from a large Cincinnati mill we blew up).

I think I will repeat the whole thing using the toothbrush method.

When I was using a drill, the flange was easy to turn. My thought was that it takes three twists of my hand to get one full rotation and the drill was just an easy way to get several turns in a short time.

Let me see if I have this correct about the gear teeth. It is OK to be a little deep initially because as the teeth wear the contact area will naturally slowly creep out toward the edge of the teeth. You definitely do not want to be on the shallow side of things initially because the contact area will not get any deeper. Am I seeing this about right?

 

[Gerrha]

Just out curiosity, where do you apply the final preload on the carrier bearings? Ring side or Carrier side? It could add or decrease your backlash reason I asked but that also depends on the procedure. I apply the final preload once I get all my readings where I wanted, I don't finish off the pinion preload if I'm using crush sleeve but on a solid spacer I just get my preload and final torque all at once.

I am following the FSM, so the final carrier bearing preload is from the pinion side, which I think would slightly increase the backlash if anything. I have also been re-checking it once the carrier bearings are done.

I plan on using a solid spacer, but as of yet, I have not been able to quite grasp how that is going to work out. The crush sleeve makes sense, but the solid spacer seems like it could be a very long trial & error process. At least I have been using setup bearings on the pinion, so they do not require pulling and pressing every trial.

 

[Malleus]

I think I will repeat the whole thing using the toothbrush method.

When I was using a drill, the flange was easy to turn. My thought was that it takes three twists of my hand to get one full rotation and the drill was just an easy way to get several turns in a short time.

Let me see if I have this correct about the gear teeth. It is OK to be a little deep initially because as the teeth wear the contact area will naturally slowly creep out toward the edge of the teeth. You definitely do not want to be on the shallow side of things initially because the contact area will not get any deeper. Am I seeing this about right?

Opposite. When the teeth wear, you have to move them closer together to get a pattern again. If you have to be off the nominal pattern, toe to heel, it's better to be above, but then again, only slightly. We're talking about thousandths of an inch here.

 

[Malleus]

I am following the FSM, so the final carrier bearing preload is from the pinion side, which I think would slightly increase the backlash if anything. I have also been re-checking it once the carrier bearings are done.

I plan on using a solid spacer, but as of yet, I have not been able to quite grasp how that is going to work out. The crush sleeve makes sense, but the solid spacer seems like it could be a very long trial & error process. At least I have been using setup bearings on the pinion, so they do not require pulling and pressing every trial.

When we built axles on the production line, we used solid shims – a stack of them, not a solid tube spacer. You can get a set of precision ground shims and easily and quickly get to within less than 0.005–inch during the build.

We could do this because we knew what the depth from the bottom of the bore to a reference datum was, since it was measured at the first build station. Production axles use this method because it's the most economical way to get the pinion offset established correctly. A solid spacer would be best, but then you'd have to grind it to length, after you measured the setup, and that costs money, which is why it's not done in production.

There's no magic to where the datum is, but you have to be able to relate it to the gearset somehow. Since we also cut our own gears, we designed them so that there was a reference datum on the back side of the pinion, and that was based on the theoretical line of action of the gearset, when it was 100% correct. We also specified a reference datum on the carrier housing, based on the bore depth. This allowed us to measure the position of the pinion by measuring from the back pinion face to the datum on the carrier. The preload stack was calculated from that.

We never measured backlash; no one building production axles does that. The gear designers include a backlash measurement for rebuilders, because they assume shops won't have the necessary measuring tools, or the datum numbers (which are on the drawings and proprietary). Backlash is a backhanded way to get to the pinion offset dimension verification. It's a guess at best, just like measuring spinning preload with a torque wrench.

If you want to use a solid spacer, you need to measure everything first, build it so that you can verify your stack dimensions, and then tear it down, and rebuild it with the spacer. This is essentially how it's done in production, except they skip the intermediate build step; they don't need it because the build stations are validated and they trust them to be correct.

 

[Malleus]

My advice to you is to use the method in the service manual and use a crush sleeve. Unless you've built a lot of gearsets, and I mean a lot, it's too much trouble to do it any other way. Otramm's video on building a differential is one of the best I've seen. It's very clear and he follows the service manual procedure.

Oh, and ignore anyone who even hints there's magic involved. There isn't.

 

[Gerrha]

My advice to you is to use the method in the service manual and use a crush sleeve. Unless you've built a lot of gearsets, and I mean a lot, it's too much trouble to do it any other way. Otramm's video on building a differential is one of the best I've seen. It's very clear and he follows the service manual procedure.

Oh, and ignore anyone who even hints there's magic involved. There isn't.

A lot to digest there. Starting from the bottom working up...

The thing about the crush sleeve that started me going down the road with the solid spacer is how to hold the companion flange while torquing the pinion nut to 250 ft-lbs or thereabouts. Most of the references I have checked use a homebuilt contraption of one type or another that looks like welding is involved and I do not weld.

I had come up with a tentative plan for the solid spacer installation. The old crush sleeve was 1.865" tall. Assume that I start with a solid spacer + shims that is slightly taller, say 1.87" - 1.88". When I tighten the pinion nut, it should just lock up at some point, and the preload should be very low. Then, remove a few thousandths in shims and repeat. At some point, the preload will reach the FSM spec and it should be ok - I think?

 

[Gerrha]

A lot to digest there. Starting from the bottom working up...

The thing about the crush sleeve that started me going down the road with the solid spacer is how to hold the companion flange while torquing the pinion nut to 250 ft-lbs or thereabouts. Most of the references I have checked use a homebuilt contraption of one type or another that looks like welding is involved and I do not weld.

I had come up with a tentative plan for the solid spacer installation. The old crush sleeve was 1.865" tall. Assume that I start with a solid spacer + shims that is slightly taller, say 1.87" - 1.88". When I tighten the pinion nut, it should just lock up at some point, and the preload should be very low. Then, remove a few thousandths in shims and repeat. At some point, the preload will reach the FSM spec and it should be ok - I think?

I have watched Otramm's videos and he uses an impact wrench. I can see myself blowing past the correct torque with that and having to use a new crush washer. He seems to know exactly how far to push that impact wrench technique.

 

[Malleus]

I have watched Otramm's videos and he uses an impact wrench. I can see myself blowing past the correct torque with that and having to use a new crush washer. He seems to know exactly how far to push that impact wrench technique.

I rewatched Ryan's video again, to make sure I had his procedure in my mind as I write this. Just to make sure we're talking about the same thing here, this is the video I'm referring to: Otramm 80 series 9.5" differential rebuild

So, I understand your concern about using an impact wrench. If you don't trust yourself, lift the companion flange off the carrier housing and stick a shim in there (or two, one on each side, if you can't find a "C" shaped shim – you could cut one with tin snips out of brass shim stock); 0.020" would be sufficient. Torque the nut until you hit the shim, and then pull it out. This way, you'll never compress the crush sleeve too much.

As far as a "special" holding fixture, yes you have to have one. If you're not into making tooling, let someone who has the tools do the work for you. You have no other choice. Having said that, you do not need a $5,000 welded platform with spinning capability and multiple locking stops (that'd be killer, though).

What you do need is a means to secure the carrier housing while you work. You can make one out of wood, and throw it away when you're done with the job. You need to have a means to secure the mounting flange (this is where the carrier housing is secured to the axle), and access to the inside of the housing. This is by no means to only way to do it, you could clamp the outside of the casting somehow, but it is the conventional way to skin the cat.

Oh, and if you have a hydraulic press, don't pound bearing cups into the housing on the floor. Use the press.

 

[Malleus]

A lot to digest there. Starting from the bottom working up...

The thing about the crush sleeve that started me going down the road with the solid spacer is how to hold the companion flange while torquing the pinion nut to 250 ft-lbs or thereabouts. Most of the references I have checked use a homebuilt contraption of one type or another that looks like welding is involved and I do not weld.

I had come up with a tentative plan for the solid spacer installation. The old crush sleeve was 1.865" tall. Assume that I start with a solid spacer + shims that is slightly taller, say 1.87" - 1.88". When I tighten the pinion nut, it should just lock up at some point, and the preload should be very low. Then, remove a few thousandths in shims and repeat. At some point, the preload will reach the FSM spec and it should be ok - I think?

Unless I misread your post, you are using new gears, correct? Why would the old crush sleeve correspond to the new gears? You must measure before you build. This is the step that backyard rebuilders don't get/understand/care about. Don't be that guy. Measure, measure, measure.

This is not a trivial exercise, but it's not difficult, either. Think about what is needed and do that.

The goal in setting the pinion is to ensure that the contact pattern you painted is replicated inside the housing when the assembly is built. There are various ways to do this, even in a home shop or garage, but they all boil down to one thing: measure.

If I was going to build differentials in my shop, I'd build a setup stand that would positively hold the ring gear on a mandrel that I could measure off of, or that I knew the location of relative to some datum on the stand, and it would have a means for moving the pinion into contact so that I could measure from the back face of the pinion to the centerline of the ring. This is, by the way, the setup dimension on the gearset drawings from the manufacturer. This is where the spacer is going to seat, between the poinion and the bearings so it's the only surface that matters.

Whether you choose to do this, or like Ryan, use the hit or miss approach, which will work, you have to make sure the measurement from the pinion backface to the ring gear centerline is the same when the gearset is installed as it was when you printed the gears. If it is, you're good. If it's not, you're not. It's that simple.

If you know the dimension, and you build the differential assembly according to that dimension, you're done. No need to measure backlash or rotating torque, just like the pros do. You can measure those things, so you have a reference for wear in the future, but you don't need them to validate a good build.

 

[Gerrha]

I rewatched Ryan's video again, to make sure I had his procedure in my mind as I write this. Just to make sure we're talking about the same thing here, this is the video I'm referring to: Otramm 80 series 9.5" differntial rebuild

So, I understand your concern about using an impact wrench. If you don't trust yourself, lift the companion flange off the carrier housing and stick a shim in there (or two, one on each side, if you can't find a "C" shaped shim – you could cut one with tin snips out of brass shim stock); 0.020" would be sufficient. Torque the nut until you hit the shim, and then pull it out. This way, you'll never compress the crush sleeve too much.

As far as a "special" holding fixture, yes you have to have one. If you're not into making tooling, let someone who has the tools do the work for you. You have no other choice. Having said that, you do not need a $5,000 welded platform with spinning capability and multiple locking stops (that'd be killer, though).

What you do need is a means to secure the carrier housing while you work. You can make one out of wood, and throw it away when you're done with the job. You need to have a means to secure the mounting flange (this is where the carrier housing is secured to the axle), and access to the inside of the housing. This is by no means to only way to do it, you could clamp the outside of the casting somehow, but it is the conventional way to skin the cat.

Oh, and if you have a hydraulic press, don't pound bearing cups into the housing on the floor. Use the press.

Yes, that is one of his videos that I have watched several times. It has been very helpful along the way.

What I have been using as a lever is shown in the photo below. It works at the torque levels so far and I could clamp it in a vise and/or make an even longer version that likely would do the job. Not the vise in the photo, a real vise.

I have a hydraulic press and do not pound the bearings in place or oil seals for that matter.

 

[Gerrha]

Unless I misread your post, you are using new gears, correct? Why would the old crush sleeve correspond to the new gears? You must measure before you build. This is the step that backyard rebuilders don't get/understand/care about. Don't be that guy. Measure, measure, measure.

This is not a trivial exercise, but it's not difficult, either. Think about what is needed and do that.

The goal in setting the pinion is to ensure that the contact pattern you painted is replicated inside the housing when the assembly is built. There are various ways to do this, even in a home shop or garage, but they all boil down to one thing: measure.

If I was going to build differentials in my shop, I'd build a setup stand that would positively hold the ring gear on a mandrel that I could measure off of, or that I knew the location of relative to some datum on the stand, and it would have a means for moving the pinion into contact so that I could measure from the back face of the pinion to the centerline of the ring. This is, by the way, the setup dimension on the gearset drawings from the manufacturer. This is where the spacer is going to seat, between the poinion and the bearings so it's the only surface that matters.

Whether you choose to do this, or like Ryan, use the hit or miss approach, which will work, you have to make sure the measurement from the pinion backface to the ring gear centerline is the same when the gearset is installed as it was when you printed the gears. If it is, you're good. If it's not, you're not. It's that simple.

If you know the dimension, and you build the differential assembly according to that dimension, you're done. No need to measure backlash or rotating torque, just like the pros do. You can measure those things, so you have a reference for wear in the future, but you don't need them to validate a good build.

Measuring is fine. I have no problem with that, but is the overall concept reasonable? Start with a sleeve slightly longer than what is required and slowly reduce its length until things are where they should be. It seems starting with a sleeve too short would be big trouble.

 

[Malleus]

One last thought: don't expect that you're going to get this done in a couple of hours, and if you don't, you're seriously screwed, or something like that. Even with precise measuring tools, building a properly working gearset is an art. If this is the first time you're doing it, give yourself some time, and expect to have to do it a couple of times, before you get it right.

Ove the years, I've run three different differential production assembly lines, by which I mean I was the line engineer, not the floorwalk cracking a whip, and all of them had their own ideosyncrasies, even though they were all based on precise measuring instruments, from direct reading dial indicators to force and distance transducers feeding a PLC. None of them worked 100% all the time. I know how to build diffs because I've spent a lot of time rebuilding diffs that the assembly lines couldn't get right. Sometimes, I built them several times before I got them to work. After three times though, I tore them down and redistributed all the parts into their various bins so they couldn't be used together again, but I had that luxury and you won't.

I say all that to say this: there is an amount of skill involved in doing this work quickly and correctly, that is only gained through practice; you really only need to worry about the "correctly" part.

 

[Malleus]

Measuring is fine. I have no problem with that, but is the overall concept reasonable? Start with a sleeve slightly longer than what is required and slowly reduce its length until things are where they should be. It seems starting with a sleeve too short would be big trouble.

My thought is just that you should be able to know what length you need before you get that far. A little creative thinking will get you there. Draw pictures as you go; I do it all the time.

Based on the photo, I think you have all you need. I've seen acceptable results come out of shops that were not as nice as yours.

I hope some of that helped. Don't hesistate to ask questions if you have them. There are lots of guys on this forum who know lots of stuff.

 

[Gerrha]

One last thought: don't expect that you're going to get this done in a couple of hours, and if you don't, you're seriously screwed, or something like that. Even with precise measuring tools, building a properly working gearset is an art. If this is the first time you're doing it, give yourself some time, and expect to have to do it a couple of times, before you get it right.

Ove the years, I've run three different differential production assembly lines, by which I mean I was the line engineer, not the floorwalk cracking a whip, and all of them had their own ideosyncrasies, even though they were all based on precise measuring instruments, from direct reading dial indicators to force and distance transducers feeding a PLC. None of them worked 100% all the time. I know how to build diffs because I've spent a lot of time rebuilding diffs that the assembly lines couldn't get right. Sometimes, I built them several times before I got them to work. After three times though, I tore them down and redistributed all the parts into their various bins so they couldn't be used together again, but I had that luxury and you won't.

I say all that to say this: there is an amount of skill involved in doing this work quickly and correctly, that is only gained through practice; you really only need to worry about the "correctly" part.

For sure. The quickly part is not in the equation. I have already taken this one apart and put it back together at least 5 times and that is fine.

 

[Malleus]

You have the patience of Job, my friend.

 

[vipergrhd]

Yes, that is one of his videos that I have watched several times. It has been very helpful along the way.

What I have been using as a lever is shown in the photo below. It works at the torque levels so far and I could clamp it in a vise and/or make an even longer version that likely would do the job. Not the vise in the photo, a real vise.

I have a hydraulic press and do not pound the bearings in place or oil seals for that matter.

View attachment 4025908

Nice setup, I am in the process making similar tool but mounted on my engine stand. I used a hot plate to install bearings no pounding or need for a press. I did borrow my friends clam shell tools for bearing removal which makes it so much easier specially when you have to swap pinion shim.

 

[vipergrhd]

@Malleus I dig all the stuff you shared on here but my training was more like a crash course and you pretty much learn on the job and find a quick way of doing things I'm no engineer but I have wrenched all the years since I was 19. Did you worked for Dana or AAM? Pardon the hijack.

 

[Malleus]

Dana and Getrag. I built 70(-ish) series drive and drive–steer axles used in agricultural and industrial vehicles and the transaxles used the Cadillac CTS (I also helped design and build the assembly line, as well as managing the technical aspects of operation) and Corvette. I never had the opportunity to visit an American Axle plant.

That was another life ago.

 

[Gerrha]

Nice setup, I am in the process making similar tool but mounted on my engine stand. I used a hot plate to install bearings no pounding or need for a press. I did borrow my friends clam shell tools for bearing removal which makes it so much easier specially when you have to swap pinion shim.

Thanks!

 

[Gerrha]

After redoing everything and using less marking compound, here is the outcome. This is with same thickness pinion plate as the photos from yesterday. From my look-see, the pattern looks ok?