Quick thought, the only difference between how the diff is mounted in an mx5 & your roadster, is the brace that goes from the diff forward. I know this is too bulky, but, if you used a bar with female ball joints on each side, this should stop the problem. I think that the diff is twisting under strain, snapping the arms, hopefully the female ball joints should allow enough movement to take the load off the diff when you are pushing on. :rolleyes:

Ahahahahaha!!! Whoops! :D

I think we all have this to look forward too, guess the mx5 diff just ain't made to be not solidly mounted! Might keep hold of that 1.8 diff as a spare for when mine goes.

Opps, that must be really frustrating when its pretty much your first ride. Still whatever happens, it can't be as bad as it was going through the IVA. Sure you will have it fixed in no time.

First up, glad to hear it sounds like you're OK Skov, that could really hurt if bits of metal get loose!

Second of all, the issue of mounting the MX-5 diff has been discussed, at extreme length, in the past when Nathan suffered a diff failure during testing. Now, the reason for his failure was different - it wasn't bolted in correctly. But the discussion that followed concerned the diff nose mounting specifically. Keith Tanner's build used a box section frame, with captive nuts, and a part of it was welded to the transmission tunnel. In other words, the nose was solidly mounted. After a couple of years, he removed it during a diff change IIRC and found no issues with it whatsoever.

Every kit car manufacturer that uses the MX-5 diff has solidly mounted it AFAIK, including the diff ears (or in the case of GBS, cut them off altogether and bolted the top part of the diff to the chassis).

Now, I'm not a trained engineer, but it appears you need to replicate the PPF. The PPF is solidly mounted to the diff and the gearbox tail. There are no compressible mountings between the two. This fits in with what the kit car manufacturers have done.

So do I think that a broken diff arm (Mazda deliberately engineered a failure mode in the arm) is an inevitability? No. On my newly acquired build, I have a set-up identical to Johno's design, where there are 2 rubber Land Rover V8 engine mounts, fitted so that 1 mount would always be in compression. I'm planning on substituting these mounts for 1 piece solid versions in a material such as ali or nylon. The diff ears will either be polyurethane or solid again.

Only my pride that's been hurt so no real damage done :o
Luckily I'd only just pulled away from a junction at the time, so wasn't going very fast.
I do recall all the discussion following Nathan's diff fail, but I think I'd already mounted my diff by then. I've always had it in the back of my mind that it'd probably fail at somepoint, but thought I may as see how long it lasted... I'm a bit of a lazy git like that!
I think what happened is that the landrover mounts failed and at that point the diff could move enough to rip itself apart. I suspect if I'd done it properly with them in compression it would have been fine.
Saying that I am going to solid mount the next one.

http://www.mpsgarage.com.au/shop/pro...or-mounts.html

Mx5 boys check out the diff mounts here on the above link. They are proper diff mounts for the mx5 arms to stop the snapping of the arms. :cool: they are proper Mazda speed items.

They are engine mounts.

Mazda deliberately engineered the O/S arm to break in an accident.

Sorry about that. I found them on a forum that the drift boys were using, as one bloke was complaining of the same problem, and that was what somebody suggested. Do you think Phils idea will work ? I want to see somebody try it before me, I hate being a guinea pig. :D

Blinkin eck, that's something you don't want happenning too often. When I first started building my new monster I was very concerned about mounting the diff. The forces involved in an IRS set up are very complex. The assembly has to be able to handle the torque from the engine, the torque driving the wheels and the lateral forces involved in accelaration and retardation. It basically seems like every part experiences forces in every direction under one condition or another. Any weakness in any direction can result in failure. There is a web site in the US of A dedicated to irs setups its this one here: http://irsforum.boardhost.com/viewforum.php?id=1 There may be some useful info buried within that will help get to the bottom of this.
Best of luck.
Dave.