I am working on putting a limited slip from a RX-7 in my Miata. It is the clutch type from an 88. I measured the breakaway torque before I tore it apart at about 10ft-lbs. So pretty much shot. I added two beer can shims (0.1mm each) and measured the breakaway torque again, this time it was about 17ft-lbs, still pretty much shot.
I added the shims at each end between the cone spring and the outermost friction plate. Is this the right place to put them and I should just add more or should I put them in a different location?
You know, LSD can mean multiple things, right?
I do. 
That's why I didn't put it in off-topic.
What did the plates measure compared to the specified thickness?
Uhhh...not bad?
Original spec is 2mm, minimum is 1.9. Here are the measurements I got measuring at two opposite sides of each disc:
Ring gear side:
1.97/1.98
1.98/2
1.99/2.01
1.97/1.99
Opposite side:
2.03/2.02 (don't know how I got that, buildup of stuff on the disc maybe?)
1.99/1.97
1.99
1.99
So adding that all up I am about .1mm or so off new? I shimmed it .2mm.
Here is a diagram of the assembly:
I put the shims between 3 and 4.
Measure the plates and thrust washers too?
Hopefully someone who's actually been through this process will chime in.
Those measurements were for the friction plates and friction discs (4 and 5 in the diagram).
The thrust washers also measure within spec, 1.55-1.57mm.
Hmm, I just noticed in the diagram that the friction plates and friction discs are in alternating order: plate-disc-plate-disc.
When I took mine apart they were lined up plate-disc-disc-plate. I just put them back in the same order they came out. Maybe that has something to do with it.
Yea I figured. Someone must have had this apart before and put it together wrong.
I'll redo it after class tonight and see what it does.
Stacking them different will give you different lockup percentages (I dont know what it is off hand) - I've got a tran X rebuild sheet somewhere that had the info.
Also note, the 10 ft lb you get at zero rpm dry will be a lot different when the wheel is spinning a few hundred+ RPM. The RX7 FSM may have rebuild instructions with pass / fail breakaway torque numbers.
Kendall
From what i have read the way to measure it on the car is with one wheel in the air and a torque wrench on the axle nut. I think the factory specs is 30-40ft-lbs. I'm shooting for 100+.
...it's no wonder those clutch diffs are a bit loose after nearly 25 years, most of those cars came with space saver spares.
The design in part uses the angle between the spider/axle gearteeth to add pressure to the clutch packs. The more power you put thru them, the harder they grab. Not much good for an autocross car with little weight on the inside tire, but for drag style use the worn clutches usually still work just fine. With some clutch diffs, roadrace guys sometimes complely eliminate the pre-load springs and rely solely on the internal geometry to load the clutches...the diff grabs on power, opens up when you are off.
Post #12 HERE has some info about the clutches/shims in the ford 8.8. Should be very similar. They talk about the steel vs friction disc order and such. I think they usually try to put in as many shims as they can and still get it back together.
On the trashed 12A diff that I have (two clutches per side, not just one) I have a .010" on each side. Breakaway torque used to be "infinite", tires howling when trying to turn on the street, a real riot. Just like a spool. After a while it started to pop and creak and crack and actually start to break away, so I put a tube of LSD goo in there, which quieted things down as well as making it more progressive. I can still torque the lug nuts without the diff breaking away, mind you.
OTOH, having this much breakaway means the car understeers a lot more. Like, if it's raining, I have to use the handbrake to turn the car in a parking lot, since the front wheels will just slide. When there's grip, there's a weird rubbery sensation to the steering... just like driving a car with a spool. You may not necessarily want this much. In fact, I probably don't want this much, but I figure it just gives me more time before I finally have to give Mazda a good chuck of change for new clutch pack parts.
The diff WILL still grab when under throttle, that's what the funky ramps are for.
And YES, alternate the discs! Mazda put them in like that so the diff is weak and wimpy and Joe Plebian doesn't bitch about weird rearend issues. You're not Joe Plebian.
Well I put the discs in alternating order and added two more shims for a total of .4mm of shims.
Measured the breakaway again and now it is at 40ft-lbs so that is definitely an improvement.
Are you measuring the breakaway torque with the ring gear bolted down? The four little screws do not cut it, and the diff will compress further when you tighten down the ring gear.
I was before, this time I was lazy and wanted to test it out before bolting the ring gear back on.
I'll put it on and try that.
Ok, ring gear bolted on and it is now much better. I got to 80 ft-lbs without it breaking away and I wasn't really comfortable torquing the little bolt any higher.
weedburner wrote:
...it's no wonder those clutch diffs are a bit loose after nearly 25 years, most of those cars came with space saver spares.
I ruined an LSD diff with a space saver spare. I knew I shouldn't have put it on the back.. but it was POURING down and I did not want to be out in the weather any longer than I needed to be.. Sadly, toasted the diff in 20 miles of driving
EvanB wrote:
Yea I figured. Someone must have had this apart before and put it together wrong.
I'll redo it after class tonight and see what it does.
I've had one of these apart recently and the plates did not alternate like the 1st gen Rx7 LSD does. My understanding is the 2nd gen came from the factory like this to reduce complaints from customers about the noises a LSD can make. I reassembled alternating the plates and did not shim as there was very little wear. So far have not heard the inside wheel spin uselessly on corner exit at an autocross unlike it did with the standard differential.
