Differential Chatter Loud on Slightest Turning, 83 LSD
 

Hi All,

I have a 1983 GSL, ~40K miles, replacement drive shaft (non staked u-joints, Mazdatrix part), small axle LSD diff. Clutch is super smooth, replaced ~4K miles ago.

I've had these rear axle symptoms over the last 3 years:

- when starting off with a turn, left or right, a slight stutter/chatter in engagement
- whirring noise at highway speeds on acceleration and downshifts, quiets on coasting
- occasional pretty loud groan coasting through a sharper turn at 10-15 MPH

I tried to address these with fluid change (Valvoline 70W90 mineral with LSD additive) and adding additional friction modifier, 1 or 2 ounces swapped in 2 times. That helped a bit at first, but after a few hundred miles it would return.

Now, in the last couple of days the noise progressed to chattering/clattering noise when driving in anything but a straight line, even a slight turn coasting or powered makes a pretty loud chatter.

I've removed both axles, and they look great, oil seals in good shape, no play in/out before removing. They came out with a tug on the flange, no puller needed. Both axle bearings are pretty smooth, can hear a very slight blip on the right bearing with a stethoscope. Left axle bearing has some resistance to spinning freely. They are worth replacing but are not the source of the sound IMHO.

What are your opinions on what to do next? I am guessing pull the diff out and look at it. Can I just disconnect the drive shaft at the diff flange, remove the bolts on the housing and pull the diff out? I'd prefer not the drop the whole axle if I could avoid that.

Thanks so much for reading!

IMO, start with the axle bearings, they are pretty easy to do, so if it fixes it, you're good.
if it still makes the noises after that, then you need to pull the diff and see what is going on in there.

you might also try a different oil.

i'm in the same boat, before i rip my FC diff apart, i've been told i should just do wheel bearings.
i'm 99% sure its the diff, but its a lot easier to do wheel bearings, so i might as well start there

I would try MORE additive first.

If you alrady have the axles out though you are 1/2 way there. Pull it and check the cluth plates IMO.

It's tricky AF to get it all set up properly if you tear the whole thing down. Getting the pinion depth and ring lash correct is tricky.

you might also check the pinion torque before digging into it. About 95lbs if memory serves correctly.

Thanks you guys, I'll take a measured approach. Without the axles in place and just spinning the diff by hand, it seems really smooth, no bearing noise. And fortunately no metal shavings or chunks on the magnetic plugs, 2000 miles on drained fluid and it looked clear. I've got Lucas mineral 70W90 LSD to replace the Valvoline.

For the axle bearings, I'm assuming I can reuse the spacers and just replace the bearings and collars?

always got my bearings and spacers as a pair from mazdatrix. not expensive, and cut mine off if I remember correctly....cut off wheel/angle grinder....but don't cut too deep.

I suppose a large press would work, but that pressing two tight fit items at once...a lot of pressure.

bought me a a harbor freight press for 200 for just this kind of stuff. about the price of one machine shop visit. and I can sell it and get most of my money back.

Have you tried GM LSD Additive?

https://www.volvette.com/Merchant2/g...01/s/SU60U.jpg

I previously used CRC Posi-Trac and added about 4 oz in addition to whatever the Valvoline had in it. That GM stuff looks similar to what Mazdatrix recommends.

I've ordered the spacers, bearings, collars, and the LSD additive from Mazdatrix. I'll update when I get it all sorted.

If you would like a condensed version of the archived LSD build thread with my notes on the few I’ve setup so far (low to no miles, but everything measures out) also with Mazda part numbers along with a poc at that can still get the parts let me know. It’s not that difficult if you don’t mind taking the time to measure up everything multiple ways and set up the stack and do the math.
https://cimg5.ibsrv.net/gimg/www.rx7...df1e425f5.jpeg
https://cimg0.ibsrv.net/gimg/www.rx7...1b52f38ae.jpeg
https://cimg3.ibsrv.net/gimg/www.rx7...07beb8c02.jpeg
https://cimg6.ibsrv.net/gimg/www.rx7...aac417cb8.jpeg

https://cimg5.ibsrv.net/gimg/www.rx7...0a375cf54.jpeg

Thanks! If I do end up needing a rebuild I'll take you up on it. That is the kind of thing I like to do...is your guide applicable to the 83 GSL LSD?

In any case I'm thinking your guide would also be really helpful to people here in general.

I love what you did with the wood. Pain taking it in and out of the vise. My LSD is set, but i do have mine torn down for R&P setup. Do you have anything specific the the ring and pinion setup, lash etc?

I have read the other thread,,,,, I am just in a holding pattern now as I am down to just doing the R&P.....but I screwed it up once and don't want to again. As you know the more infor the better...just like here with that board. gonna make that for finishing mine.

That board is a life saver. I tried to do a test run with the casting rib in a vise on a used crush sleeve and it 1. Fell out and b. ripped up the freshly painted third member.

Yes, I think all the internal measurements and clearance of the different year LSD cores are the same (double check me to be sure). The primary difference is that GSL friction plate stacks go the way I set it up…
outer-inner-outer-inner-spider-i-o-i-o
the -se for extra torque slip (guessing here) has a different stack arrangement that didn’t use the inner most disc.

My measurements on the post it note are shorthand for thousandths of an inch. Got tired of writing .0xxx the accruals are on the cardboard.

I took all of the archive thread and copied over the posts that had info in them and made a word doc. I can post that tomorrow.

Some pointers…
-there’s a torque setting for the pinion nut… that does not take into account the large amount of force needed to start crushing a new sleeve.

This is how it was held to crush the new sleeve. That’s a four foot jack handle…

https://cimg9.ibsrv.net/gimg/www.rx7...571866206.jpeg

-the amount of torque needed on the circular internal spline adjuster nuts is actually pretty low. Those are only holding the caps down in place enough to get the bearings seated and to set the lash and other items. It’s very easy to over torque them and cause the bearing caps to open up over the spec… hence the mic.

-that spec for the spread across the bearing cap flats… I think it’s there to make sure that during operation, the pinion causes the ring to push them out and the housing’s internal ribs are what actually takes the load of the pinion to ring gear. If those have been ground down in the past by an over exuberant prior user… then I’m not sure, short of welding the rib ends and machining them back down in a mill. Any type of shim on the rib could fall out… maybe some careful weld on the bearing cap and very careful grinding 🤔 (i do not recommend this). The reason I say those ribs are machined that way is that I did measure and when I installed the third member, it would only drop down dead vertical. Anything off center it would bind and those ribs are the only things down there.

re: pinion lash. As long as you reused the pinion shim that came with that housing… your tooth pattern engagement should be good. I sharpie’d or grease pen 4 locations, 90 degrees apart, where I check the ring lash. Check each location once, adjust one cap, adjust a few more times until you get a feel for lash and then check the cap spread. Tap the caps with a brass or dead blow to set the bearings. Rotate a few times and then check the lash again. I was shooting for the middle of the range and I ended up a little on the tight side with a couple corners and where I wanted on the others. The cleanliness of the ring gear to carrier is very important. I hand filed all the mating surface before loctite and setting the ring gear for good. These don’t need heat to fall in place. The gear material likes to break off in slivers. So double check threads and everything for clean. Once you are close on the lash, remember both bearing cap nuts work together, if you move the left spline nut in without moving the right spline nut the same amount, you are spreading out the caps. I worked with the caps more or less snug about half torque. The locking washer tang thing can be moved a decent amount once it’s into a spline slot. My final pass for lash was after I torqued the bearing caps down fully with the tang in a slot, then intentionally moving the spline nut so that the tang gets snug and doesn’t allow further rotation. Do that tang preload on one side then check the lash. Whichever side you do first try and predict how much lash you’ll get from the next one, taking into account the cap spread and the rotation needed on the next one. I think I did this tang preload each side by unscrewing the spline nut.

They all adjust and work together, you’ll get a feel for the changes needed after a few rotations of measurements.

Hope this helps.

Also…
https://cimg8.ibsrv.net/gimg/www.rx7...e39a5a14a.jpeg

Also, also…

Finally got some time yesterday to remove the old collars, bearings, and spacers. I used the FSM recommended method to grind a slot and whack it with a chisel. It worked perfectly for the old collar and bearing, the spacers just slid right off. It's pretty much impossible not to damage the spacer if you're cutting the bearing to remove it, so you will need replacement spacers. If you have the very long puller you might get away with it. They are cheap anyway.

I took the opportunity to clean up all the parts, and a local guy pressed everything back together today, for $20 (!), before and after photos below. I cut a slot in the brake line mount so I can remove these in the future without disconnecting the brake lines just to get the dust shield off.

I'm also waiting on new brake lines all around, sticking with OEM style rubber all around, I don't track it and I'm happy with the pedal feel as it is on the 41 year old lines.

https://cimg0.ibsrv.net/gimg/www.rx7...5ef31a3c12.jpg https://cimg6.ibsrv.net/gimg/www.rx7...8cf88ff01d.jpg

Thanks @j9fd3s @ATC529R and @j_tso for the suggestions. I put it all back together today, new Lucas 80W90 with all 4 ounces of the GM additive in the diff. That seems to have done the trick, no more stutter on launch and no chatter on tight corners or figure 8's.

​​​​​Flushing the left rear caliper showed a lot of rusty fluid! I had bought a seal repair kit this week just in case and I was glad to have it. The caliper cleaned up really well, piston was not damaged at all. I wouldn't have caught it otherwise. It's the Raybestos seal kit from Rockauto, perfect fit.

Now I can hear all of the other weird noises the car makes :p

And thanks @swbtm for all of that diff documentation.