Lotus Elan and Lotus Elan Plus 2 Forums

Thanks for all the help so far, and I dont even know your name.

Tomorrow I will repeat the blueing with a heavier dose and also view the resulting marks in better lighting.

Ian

No problem - will keep a look in on this thread.

And to quote a Paul Simon song - You can call me Al !

Al

Today I did some more testing of the diff. I took your advice and applied more blue. I ran the diff in both directions with an electric drill and the blueing indicated 'heavy face contact' as shown in the manual. There is one tight spot where and at that position there is no backlash. I have not yet checked the backlash figure in multiple positions but I think is no more then 5 thou at the widest.

I noticed that the marking was not identical on every tooth and I traced this to runout on the crown wheel so I dismantled the wheel from the cage to investigate. The mounting face of the cage wobbles about 1.5 thou although the diametrical register is dead true and does not even move my 2 micron resolution indicator.

Is 1.5 runout acceptable?

I was going to machine the mounting face and thought I had found a way of setting the cage up in the lathe so that it was very accurately located but I could not get it to run true enough to replicate the identical runout so decided not to take a chance and make it worse. I now have another plan to skim the face whilst it is revolving in its own bearings.

Ian

I use black magic marker instead of blueing ----works for me

You won't get exact markings on each face - very very rarely does this happen. It's a compromise when setting one up. Even the absolute best one i've seen didn't give an exact same contact patch marking with engineers blue. The 1.5 runout you mention will affect the marking given by the engineers blue in that it will look more inconsistent.

Don't start machining the faces - i would be very surprised if it is Ford's machining at fault - they were very very tight on tolerances - the machine suppliers that supplied Rolls Royce had a hard time selling to Ford because they were so demanding on how good tooling had to be! The reps used to hate visiting Fords!! It's not something i ever check as i've never had a problem and needed to go back that far as i know Ford's work is good. Having said that there is a first for everything but i'll be surprised if it's not in the set-up of the diff given the other details you've mentioned. There is no figure for runout on the face as it?s unlikely to distort!

Most likely culprit is that the bearing cups are not seated properly - either down to swarf/burrs under the bearing seat or that the worn housing that you mentioned means it's not holding place.

First step would be to remove the bearings and see if there's anything not right underneath the bearings. Then you could check the runout using the actual surfaces where the bearing race sits as the rotation point. If you don't have any "v" stands then just take the bearings off, inspect the areas they slide over and seat on, tidy any areas up that aren't right and until you're happy, and then re-seat the bearings. Then check the runout. If it's changed then i would say it's the bearings. If not then maybe i would agree the mounting face isn't true.

I would be tempted to get a complete new diff casing (not the part that holds the planet gears but the actual diff nose) and swap everything over into that.

If you've got movement either side of 5-7" backlash then it's out of spec - again, with worn crownwheel and pinions the figure will not be the same in every place - that' why there's that range given. But outside of that is a problem - to tight is likely to cause overheating in the diff. If you don't think diff's get hot then try putting your hands on a diff casing after it's been out for a run! Too slack can allow chipped/broken teeth in extreme circumstances.

I have made some progress on my diff rebuild, actually progress is not the right term as its no nearer being put in the car yet.

I found I had quite a wide variation in backlash over one rotation of the crown wheel. To investigate I removed the crownwheel from the cage, put the cage back in position and put a dial gauge on the face where the crownwheel sits. There was nearly 2 thou (0.05mm) of lateral runout, although the diametrical register was truly concentric. 2 thou does not sound much but because of the roughly 'wedge shape' of the mating gear teeth changing the gap between the gears by that much easily accounts for the backlash variation.

After trying to set up the cage in the lathe to machine the face I eventually decided to true the surface whilst the cage rotated on its own bearings. I used a toolpost grinder on a dovetail slide which I rigged up with bits of brackets and other hardware to the diff casing itself. I initially rotated the cage by hand but as I was only taking very light cuts the process was going to be very slow so I arrange the position of the diff on the bench so that I could rotate the cage with my pillar drill running very slow.

The crownwheel now runs perfectly true and the variation in backlash is less than half a thou over one rotation. The bad news is that checking the tooth contact pattern with engineers blue shows that the pinion needs to be shimmed nearer the crown wheel.

Now I am stuck, it seems that the correct shim could be determined straight off with the aid of the pinion setting tool which I don't have, or by trial and error, which involve numerous dismantlings and reassemblies using new crush spacers each time. Maybe I should quit whilst I'm ahead and take it to a specialist.

Ian