My DIY differential gear change

[Allan]

Thanks for the heads up on this apparently joyful process.

BTW, here's what I'm planning to do:
1) Install the pinion w/o the seal or crush sleeve. Torque enough to seat the bearings (50 ft-lbs?)
2) Install the carrier and check backlash and the mesh pattern. Iterate on the 3 spacers until it looks good.
3) Remove everything, install the seal and crush sleeve, torque to a minimum of 200 ft-lbs and then continue torquing the pinion nut (in very small increments) until I get ~30 in-lbs of rotating torque (this sets pinion preload)
4) Install the carrier, set backlash to 0.008-0.010
5) Set carrier preload to [pinion preload + (8-10) in-lbs + 3 in-lbs (to account for the seal)]
6) Verify the mesh pattern
7) Install output seals and output shafts

You mention the "art" part of this and I know I'm trying to define it as a science, as I'm new to this. Anything you can point out here that you think will help me would be great!

Bryan

Uh,...50 Ft Lbs sounds like a bit much for the pinion with no crush sleeve. You may flat spot the bearings. the 'mock-up' pre-assembly to check things out is a good idea, but the pinion needs to be able to turn. This part (well the whole thing actually) is a 'feel thing'. You're in for a suprise with the final assembly on the crush sleeve. The 200 Ft Lbs ain't even gonna move it. You've gotta pull harder than that to get that sleeve to "crush". You have this all well thought out, and again, are obviously not dumb, but this particular repair is more art/feel/experience than theory/measurements/science. I think you'll be alright, just remember....if it feels jacked-up, it probably is. It should feel similar to how it was before you took it apart, only a little tighter. ....You are gonna freak-out on the force required to sqeeze that crush sleeve!

 

[cowger]

Uh,...50 Ft Lbs sounds like a bit much for the pinion with no crush sleeve. You may flat spot the bearings. the 'mock-up' pre-assembly to check things out is a good idea, but the pinion needs to be able to turn. This part (well the whole thing actually) is a 'feel thing'. You're in for a suprise with the final assembly on the crush sleeve. The 200 Ft Lbs ain't even gonna move it. You've gotta pull harder than that to get that sleeve to "crush". You have this all well thought out, and again, are obviously not dumb, but this particular repair is more art/feel/experience than theory/measurements/science. I think you'll be alright, just remember....if it feels jacked-up, it probably is. It should feel similar to how it was before you took it apart, only a little tighter. ....You are gonna freak-out on the force required to sqeeze that crush sleeve!

Got it. 50 ft-lbs is too tight; 200 is too loose.

Seriously, Allan, thank you for all your input, feedback, and suggestions. It's truly helpful and this is the type of expertise I was hoping to tap into when I decided to share this process with this forum.

Here's everything, clean, ready to go, waiting for the new Wavetrac and bearing/seal/rebuild kit...

You must be registered for see images

 

[cowger]

Okay, time to start putting this thing back together. It was a beautiful day here in Northern California and I miss driving the car!

Yesterday I got home from a business trip to find all this waiting for me:

You must be registered for see images

Unitrax provides everything, right down to the shim kits, the new straps/bolts for the driveline/u-joint, and even the correct RTV.

The Wavetrac itself is a piece of art, beautifully machined and made right here in the USA:

You must be registered for see images

I sorted out the various shims included -- 8 pinion shims and 5 carrier shims.

You must be registered for see images

I warmed up the ring gear in the oven to about 250 degrees. It nearly pressed on while warm, and it was easy to use the bolts themselves to snug it in the final distance. Then with red threadlock, I torqued them all to 100 ft-lbs (the note included with the ring/pinion said 95-105).

The Wavetrac comes with a couple of big notches in the case making it very simple to mount in the vice while you're torquing the bolts.

You must be registered for see images

Tomorrow I'll press on bearings and start the search for the correct set of shims...

 

[Allan]

...............excellent!

 

[cowger]

Back at it this morning. First thing was to put the aluminum case into the oven and warm it up to 200 degrees.

You must be registered for see images

While that was baking, I used my new $20 crock *** to warm up the main pinion bearing and the two carrier bearings, making sure I kept track of which set went on which side (so I could match them back up to their races):

You must be registered for see images

As you can see, I'm a fan of heating things up -- seems to make installation much easier.

To press the carrier bearings on, I pulled one of the old ones off the carrier that I'm tossing, cut off the cage / bearings, and used that, inverted, to press the new bearings on:

You must be registered for see images

I used this kit:

You must be registered for see images

To help press in the two pinion bearing races:

You must be registered for see images

 

[cowger]

Next, I built a little tool to hold the pinion in place. It's just a piece of wood that uses the cap bolts and another bolt to make contact with the face of the pinion:

You must be registered for see images

That temporarily holds it in place while I invert the housing...

You must be registered for see images

...and drop the front pinion bearing into place. With it heated up, it slips over the pinion shaft with finger pressure.

You must be registered for see images

Then I installed the flange and tightened (using the old nut for now) it down until all the play was gone from the pinion bearings. As Allan had suggested, it didn't take a ton of torque to do this and I just did it by feel. It was pretty obvious when the play was gone.

I used the case spreader to open it up by 0.010", which allowed me to drop the carrier in place, install the gear-side shim, and then use a rubber soft-blow mallet to tap in the opposite side shim. For now, I used the same ones I had pulled out of the old setup.

You must be registered for see images

On the pinion, while I had pulled out a 0.050" shim, for the new setup I started with an 0.046" shim, as recommended by Unitrax.

Install the bearing caps and tightened them enough to clamp things down:

You must be registered for see images

To measure backlash, here's how I used two woodworking clamps to immobilize the pinion:

You must be registered for see images

And here's what my first try measured, 0.008" of backlash:

You must be registered for see images

Not too bad, right at the high end of the range.

Then I painted on some Prussian Blue to the ring gear teeth:

You must be registered for see images

I used a portable drill to drive the 1-5/16" socket on the pinion flange and run the setup for several rotations in either direction. I also used a piece of wood that I could gently lever inside the case, against the ring gear, to provide some load while it was turning.

Here's what I came up with for the first contact pattern. Coast side:

You must be registered for see images

And drive side:

You must be registered for see images

I was pleasantly surprised -- it seems closer than I had been expecting. I think the pinion depth looks decent, but I might try 0.048" (slightly thicker) to push the contact down toward the root of the tooth more. I also think the contact should be closer to the toe on both directions. Maybe I'll try to tighten up the backlash and see what that accomplishes.

But I'd really like to hear your thoughts on how it looks. Thanks!!!

Bryan

 

[JAY]

Nice Work !!!

 

[MtnBiker]

Your ingenuity and creativity are impressive.

 

[ACRsnake]

wow ........very impressive

where did you learn to do that ??

 

[Allan]

Yeah, to get things perfect, it's a put it together / take it apart deal. Inspect and measure, repeat as necessary. The more of these you do, the less you have to 'repeat as necessary'. I agree on the pinion shim change, get that just a little deeper and it should be perfect. -if that were a regular car or light truck, that would be close enough for horse shoes. Your work so far looks great! .....I can't wait for the episode on the crush sleeve.........................

 

[cowger]

Nice Work !!!

Your ingenuity and creativity are impressive.

wow ........very impressive

where did you learn to do that ??

Thanks guys! Where did I learn to do this? I'd say I'm learning as I go, with help from this forum. I've watched a few youtube videos and read whatever I could get my hands on.

 

[cowger]

Yeah, to get things perfect, it's a put it together / take it apart deal. Inspect and measure, repeat as necessary. The more of these you do, the less you have to 'repeat as necessary'. I agree on the pinion shim change, get that just a little deeper and it should be perfect. -if that were a regular car or light truck, that would be close enough for horse shoes. Your work so far looks great! .....I can't wait for the episode on the crush sleeve.........................

Thanks Allan! I'll post more pics once I get some more configurations tested out.

I just bought a 3/4" breaker bar today. You have me looking forward to / dreading this crush sleeve...

 

[21bubba]

For your first try you are amazingly close. Good job.

 

[Allan]

Thanks Allan! I'll post more pics once I get some more configurations tested out.

I just bought a 3/4" breaker bar today. You have me looking forward to / dreading this crush sleeve...

You better also eat your spinach AND your Wheaties. ..........you may also need to drink a Monster, the BFC,...not just the regular size can. And get plenty of sleep the night before......

 

[ViperTony]

Excellent write up so far Bryan. Love the crockpot idea. My wife gave me grief when I took her turkey baster from her fluid changes. But the slow cooker is next. It's gonna go missing sometime this evening.

 

[N2 Vipers]

Use a 3' piece of pipe to fit over the breaker bar and you will have plenty of power

 

[Allan]

Use a 3' piece of pipe to fit over the breaker bar and you will have plenty of power

Yes, the cheater pipe may be required. Also hope that there is no pre-existing heart condition or high blood-pressure problem before this.

 

[N2 Vipers]

If not enough power, use a 4 or 5' piece of pipe!!!

 

[Fatboy 18]

Great thread, but I have to say this is way beyond my capabilities Ive got to replace the output shaft on my Gen 2 Diff and am in the slow process of removing everything under the car to enable me to get the diff out. Im rubbish at maths so reading all that Backlash data and shim sizes, I think I will take my diff to someone who knows what their doing!
I will have a go at most things, but like Clint Eastwood once said "Every man should know his limitations"


Good luck with the rest of it
Mark UK

 

[The_Greg]

Wow, Awesome work. To my knowledge, my '06 diff has over 50k on it. It will be interesting when I pop it off to see what it looks like.

 

[Allan]

Wow, Awesome work. To my knowledge, my '06 diff has over 50k on it. It will be interesting when I pop it off to see what it looks like.

That's a lot of miles for a gen III diff to survive. Mine has 42,000 on it since being replaced. They will last if you don't let the car wheel hop. You can spin the rear tires up and it won't hurt anything as long as you keep it smooth. I've been hanging the tail out at the track a lot for the last 3 years with no problem other than tire wear. The car doesn't wheel hop in a drift, just from a hard launch, or heavy roll-on in first gear on cold roads. I also change the fluid every 5/6 track days.

 

[cowger]

Wow, Awesome work. To my knowledge, my '06 diff has over 50k on it. It will be interesting when I pop it off to see what it looks like.

For those wondering what their spider gears might be looking like...

I mentioned on another thread that for the 3k miles I've driven my car, I suspected absolutely nothing wrong from my differential. No noise and no sign of any gear trauma inside.

During that 3k miles, I experienced wheel hop once (on heavy launch in 1st gear, exactly as Allan describes above) but backed off immediately. Of course I have no idea what the car experienced during the first 27k miles of its life...

Devin @ Unitrax did suggest that perhaps I had a "bad batch" of gears. We'll probably never know.

I think it might be possible to look for gear fragments during an oil change. Here's what was in mine after pulling it apart:

You must be registered for see images

You can see small pieces of metal there on the side (which is normally the bottom), plus one chunk about the size of a kernel of corn. I think you might be able to push a rare earth magnet up inside the case (use the lower hole, driver side), push it as far as you can with a wire or something, then retrieve it with a piece of string glued to it. Maybe, don't know for sure. Finding stuff attached to it, though, might suggest it's time to dig inside...

Bryan

 

[roller]

Gear pattern looks good. Check again after you preload the bearings.

Frank

 

[cowger]

Gear pattern looks good. Check again after you preload the bearings.

Thanks, Frank. I'm definitely planning to check again once things are in there permanently. I do have a couple of crush sleeves on hand (one came with the gear set, one came with the rebuild kit), but hopefully won't need both...


What I'm finding so far is that the original shim set appears to be the best place to start. From this, I'm assuming it means that most variability comes from the case itself, not the other components (pinion, ring gear, etc) that are probably easier to machine to tighter, more consistent tolerances.

So right now I'm back to the original 0.050" pinion shim + the original 0.132" gear-side carrier shim, I have backlash dialed into 0.006-0.0065", and I'm verifying the mesh pattern with Unitrax. Rather than flood the forum with endless mesh pattern pictures, I'll get it finalized and share what I end up with, hopefully soon!

And then Allan will get my sleeve crush saga that he's waiting for...

Bryan

 

[Allan]

.006-ish backlash, if everything else is perfect, and it all turns smooth on the bench, is great. the tighter the initial set-up, (without being too tight) the longer it will live. The looser 'rule of thumb' spec I posted previously of .008-.012 gives an error margin. If you break it in easy, you'll be OK with the .006......but I wouldn't go too much tighter. If it's too tight, you'll hear the gears scream ugly when they get hot. Then you'll be changing the ring & pinion again. -don't ask how I know.

 

[CEJ]

I'm impressed with your shop...looks as if you have everything and then some. I'm more impressed with your attention to detail. I'm doubtful that many shops would do a rebuild as carefully as you are doing.

Did you wife know about the bearings in the crock ***?

 

[cowger]

.006-ish backlash, if everything else is perfect, and it all turns smooth on the bench, is great. the tighter the initial set-up, (without being too tight) the longer it will live. The looser 'rule of thumb' spec I posted previously of .008-.012 gives an error margin. If you break it in easy, you'll be OK with the .006......but I wouldn't go too much tighter. If it's too tight, you'll hear the gears scream ugly when they get hot. Then you'll be changing the ring & pinion again. -don't ask how I know.

Okay, perhaps I'll shoot for 0.007 or so. Unitrax's set-up sheet said 0.005" - 0.008" for Dana sets and 0.004" (!) for American Power Train (may be mis-remembering that latter manufacturer's name.)

They also mentioned something about "2-cut" vs "5-cut" gears and how the setup is different between the two. Not sure I understand all what that means (beyond the obvious number of milling operations)...

In any case, I am planning on taking it very easy (no high-torque situations) for the first 500. Will be tough, but it's good that it this coincides with winter driving, anyway.

Bryan

 

[cowger]

I'm impressed with your shop...looks as if you have everything and then some. I'm more impressed with your attention to detail. I'm doubtful that many shops would do a rebuild as carefully as you are doing.

Did you wife know about the bearings in the crock ***?

Thanks! I did buy a few new tools for this job, including the 12 ton bearing press. I am doing this project both because it gives me a good excuse to buy new tools (now I have that for any future needs) and because it allows me to see how these things really work inside and know exactly how it's done. I do enjoy it, though I realize this much work (and time!) is not for everyone.

Thanks also for your concern, but I'm safe on the crock ***. $20 at Lowes, it's mine forever, and she keeps hers safe and happy in the kitchen...

 

[CEJ]

Every project should be an opportunity to buy new tools!

$20 is cheap for domestic peace and harmony.

 

[Allan]

If your little paper that came with the gears says that, shoot for the .006........If you miss by .001 either way, run it. If you are on the tight side, just take it real easy at first, and maybe extend the break-in period a little. you'll be fine. With as clean and precise as you are being, you can push for the tighter side. It's the sloppy guys that need the 'error margin' more.