As we all know, we have a weak differential. I'm getting ready to put together a couple of differentials with REM and cryo treated parts. I'm just brainstorming here, so bear with me. Everyone who has built a differential knows that one step in the process is to set the pinion gear preload which involves crushing a pinion spacer which is designed to collapse under a load. I'm thinking that one possible mode of failure is that under extreme driving conditions (as in high rpm clutch dumps) this collapsible spacer has enough give that it compresses further, allowing the pinion gear to lift away from the ring gear, resulting in destruction. If I'm all wet here, let me know. I'm bulletproof after three glasses of champaign.

So how about when a differential is set up and the proper gear wear pattern is established, it is taken apart and the pinion spacer length measured and replaced with a spacer of the same exact length that will not crush? If you haven't rebuilt a differential, you are probably lost, but if you have, give it some thought and let me know what you think.

 

Chrysler 8 3/4 diffs where built with 2 styles/sizes pinions. ( this is a common diff on muscle car era cars) One version used a crush sleeve, the other used a solid spacer with shims to set the pinion bearing preload. On high HP cars the crush sleeve would tend to deform and you would loose the bearing preload. The solution was to adapt a solid spacer similar to the early model.

You'd only have to measure the O.D. of the pinion and have a spacer made, then have shims made to get your pinion bearing preload set.

 

Building the spacers is no big deal. So am I thinking in the right direction here?

 

It wouldn't hurt, as I've seen crush sleeves get further "crushed" from to much power. I think the ring and pinion are to small to live at high HP/torque #'s

An 8 3/4 Chrysler diff could take over 300ft lbs stock with out any issues. They start to break at the 450ft lbs range, especially on slicks, in a heavy car.

A Dana 60 axle is considered bomb proof with a 9 3/4" ring gear.

I'm not sure of the Ring gear Dia. of the S's, but I think it's under 7"

 

From what I've read, one common cause of failure is the flexing of the stock caps and the weakness of their bolts. Even a stock S2000 can cause the caps to flex slightly under heavy acceleration.

Also, the spacer giving more would not cause the pinion to lift away from the ring gear. It would only cause the pinion to move further in or out from the center of the ring gear. I don't think that temporary flex would be enough to cause failure, but that's simply an educated guess.

 

A moving pinion gear is bad no matter what direction it moves. In, out or sideways.

To verify the bearing preload is changing, you'd have to remove and inspect the diff frequently before it fails.

 

What is it about the S2K diff that makes them prone to failure and how can we prevent it BEFORE it happens?

If the diff fails, how much does a new or rebuilt diff cost?

Warren

 

Oh good point. That's true, no matter which way the pinion moves, even a hundreth of an inch drastically changes the contact patch.


Warren,

I'm not sure what the main failure is, but I do know from reading that a common problem is the flexing of the caps under heavy acceleration and launching. This can cause other kinds of failures as it can move the ring gear in relation to the pinion and destroy the diff. One solution to prevent this is the comptech reinforced diff which is what I got. It reinforces the caps and replace the bolts with much stronger studs. But keep in mind no diff is bulletproof, and I'm only relaying what I've read on the forums.

As for replacement cost? Well my best guess would be in the $2000 range plus labor.

 

the main problem is weak pinion gear materials. many members have sheared teeth off of the pinion gear.

 

I would like to know who did the rating? I have read this a few times, posted by several. I don't think this statement can be supported with facts.

Facts:
Honda changed the carrier design on the MY04s
The oil galleries in the carrier have been increased.

Everything else is opinion, and conjecture.