It was very tight.
I agree, 90+ftlbs or more. I needed a breaker bar with additional foot of pipe to break it free. Will update post if the other axle nut is significantly different experience.
Kevin,
I have to thank you for this thread. I check it every day and am learning a lot.
Most of all, it is not rocket science, just a matter of understanding what to do, why, the order of tasks and having the proper tools!!
By the way, was that seal you fished out leather?? It looked fibrous.
Dennis
69 OTS
1 Like
Dennis,
Yes, it seemed like old leather.
I truly could not do this restoration without the forum and am very thankful for the help provided by the members and administrators. My late brother Jon would be so pleased to see the progress.
Sincerely,
Kevin
Kevin,
I think it is important to make sure the polished surfaces that the three diff seals run on are perfectly polished so that the new seals can seal properly. The rubber seals have a knife edge seal while I would guess that the leather seals have a very wide sealing area. But thatās a guess.
At a local shop, they spin the surface and hold a fine grit paper on it. I think about 2000 grit but could be wrong. Years ago a shop had what was called crocus cloth they used.
Dennis
69 OTS
1 Like
Hi Dennis,
The original seal was leather, as Kevin confirms. Inboard of the seal is/was an oil slinger of circa 75mm diameter, to keep the majority of the oil away from the seal (to aid the function of the seal). If you need to replace this slinger, its been replaced by a small diameter disc of circa 40mm, that acts only as a spacer with the more efficient lip seal.
Brent
2 Likes
I dont know whether you have the ball bearings or opposed taper bearings, but the previous (very recent) lister was intending to follow the very confusing (in my MKX manual, and evidently his E-type ) instructions to allow .001 to .003" endplay for the output shafts
I have subsequently been convinced that for the tapered roller bearings, that pre-load of .001 to .003" should be the correct setting
I believe both types exist in the E-types, (and I do have E-type Service Manuals)
What are your instructions/intentions ?
Hi Tony,
Yes, the manual is insufficient by itās self and Iāve had to really study it and rely on forum information along with YouTube auto mechanics videos to piece it all together so far. I have no rebuild or restoration experience except my one gearbox rebuild and now the start of this final drive rebuild attempt.
The car is a S1 4.2 litre 1967 OTS and has the opposed taper bearings. I am familiar with the tolerances you referenced that are in the manual. I meant to check them before disassembly but forgot. Was too excited to make progress 
My intentions are to match all specs and tolerances in the manual and if I cannot, I will send the unit to Coventry West and request they bail me out. So, goal is 001-.003 end play or float on the output shafts.
Next step is order replacement bearings, seals, and o-rings. If I have the confidence, I will buy the 3.07:1 ring and pinion, install it, and save the 3.54-1 ring and pinion.
I intend to include all progress and set backs as updates on this post until final drive is complete including nice paint job.
Thanks,
Kevin
1 Like
Hi Kevin,
I would rethink setting the bearings up with end float, but if hell bent of setting with end float, use the bottom end of the tolerance; 0.003" is a shite load.
The end float specified in your manual is old thinking. The later type 3 differential used double row, angular contact bearings with built in pre-load, and the same opposing taper roller bearings as your differential, that were reintroduced for S2 and later XJ6 cars, plus XJS as well as other V12 saloons, are also set up with pre-load.
Should you decide to go a little more update and set the output bearings in pre-load, rather than set the pre-load by a measurement of shim stack thickness, its better to measure the amount of torque required to rotate the output shaft in its bearing assembly. The specified amount for an XJS, effectively the same differential as yourās, is circa 5lb/in torque.
Brent
1 Like
I think you will find, as Brent alluded to, that the tapered roller bearings in any application should have pre-load.
Check the Timken website, or otherwise Google the topic.
Pretty sure Dick confirmed this
One other thing, that no-one ever seems to mention.
I found the best diff ratio is 3.31 LSD from a S2 V12 XJ
Perhaps that ratio was not standard in USA, it was standard in Australia
Thank you Tony and Brent. I better understand the different methods to obtain pre-load for the output shafts. I have changed my mind.
I will remove shims until circa 26lb/in torque is needed to rotate the Output Shaft in itās bearing assembly. Will achieve this torque while ensuring alignment of the locknut and lockwasher.
Will check Timken website.
3.31:1 vs. 3.07:1 decision will be pondered.
Have you made up āsetup bearingsā yet.
They are indespensible for this job
best way, order 2 sets of new bearings, hone the respective inner or outer surface of the setup bearings to a fairly tight slip fit
Alternatively, if your original bearings appear unworn, clamp the new and old bearings lightly between 2 steel plates with g-clamps, and measure with a micrometer, if they are within .001" of each other, you can use the old ones as setup bearings
Hi Tony,
Can you explain to me, ignorant as I am but want to learn, what is the role of setup bearings?
Thanks
Dennis
The tapered roller bearings are āpressā fitted with shims and a spacer
You would be lucky to get it right the first time
Pulling of the inner bearings is almost always destructive, as the puller catches the outer cage usuallyā¦and a pita, as they are a tight press fit
āsetup bearingsā, with the inner surface honed out, just slide on, all measurements are taken, once satisfied, the bearings to be used in the final setup are pressed on
dont forget the oil seal !
This principle also applies to diff setups, especially pinion depth
1 Like
The output shafts on the 3.8 I restored had a preload. I donāt know what it was but the bearings and shafts were blue.
So I rebuilt them with new bearings and set the float to .0015". Installed new Timken bearings in the diff too. No problems in the following 18 or 19 years of use
I wish someone would point out where Timken say that opposed taper roller bearings should be set to a pre-load, including in the setting up of wheel bearings, too.
The manual setting procedure outlined in their 2016 manual says the same as it always has for wheel bearings for example. First seat the bearings, then tighten the nut until rotational resistance is felt, then back off the nut 1/6 to 1/4 turn and lock it.
This is what they say about the clearance;-
Generally, the ideal operating bearing setting is near- zero to
maximize bearing life. Most bearings are set with a cold setting
of end play at assembly. This comes as close as possible to the
desired near-zero setting when the unit reaches its stabilized
operating temperature.
The clearance reduces as the bearings reach operating temp.
I once had a front wheel bearing disintegrate at speed on the freeway because it was over tightened, so Iām pretty concerned about getting this right.
1 Like
so you measure the width of the new, and width of the setup bearings. Then assemble using setup bearings that slide in easily. Then adjust shims for the difference between the setup and the new bearings and assemble with new bearings??
Dennis
yes, thats essentially correct
setting up Pinion & Ring gear final pattern often requires many attempts to get things right.
I suggest you Google āsetup bearingsā and you will be an āinstantā expert 
I recently did this job uncounted times on a Dana 44 solid axleā¦before ascertaining the whole axle housing must have been bent, upon close examination, I could see it had taken a hard hit, must have been steel to bend a housing as tough as thatā¦it was on the diff pumpkin housing itself, so the pinion and ring gear didnt run true
very frustrating ! I could have saved a lot of time by measuring ring gear runout earlier in the pieceā¦you learn something every time
[quote=āClive_Wilkinson, post:34, topic:415817ā]
The output shafts on the 3.8 I restored had a preload. I donāt know what it was but the bearings and shafts were blue.
Bearing preload: What is it and why is it important? | Engineering360 (globalspec.com)
Interesting Cliveā¦I see the same in the Timken .pdf
When I looked up the topic years ago there was a different statement in Timkens site
(I quoted another site as well here)
I, like you, set my axle inner shat bearings with .001 endplay, as per the manual, and I, like you, have seen them upon dismantly be blue.
Part, or the whole answer, may lay in what Timken alludes to, thermal expansion, a question which arose at times when I have asked this question before
One would have to think the situation with the bearings in this setup differs from Differential Bearings, especial Pinion Bearings, which must certainly be setup with pre-load, or they fail immediatelyā¦having said that, they also fail very quickly if setup with too much pre-load.
I am most prepared to listen to the guys that do this stuff for a living, as failure to get it right means a very expensive re-work
I know a guy that does a lot of newer Ford diffs, and they have had to ignore the factory specs, and set them up with less pre-load than specced, as they kept having failuresā¦ie they were having failures in fairly new cars, do the repair to factory spec, and ot would fail againā¦he did say crap Chinese steel was a suspect
Hi Tony
What has caused the blue? Not enough preload so the balls skid across the the race and not roll?
Lots of learning for me here!
Dennis
imo, the opposite, too much pre-load, generates friction and heat, followed by eventual seizure
on the occasions I have seen bearings destroyed by what I considered too much end play, you can see how things have been flogged about, leavings marks, and eventual distintegration
I am not an engineer, nor especially experienced with bearings, but I did have some years experience as a technician, so always wanted to understand why a part failed prior to replacing it.
Iām pretty sure if you Google (or Youtube) this topic, you will find extensive information available that will increase your knowledge
Its been a number of years since I started asking about the section in the FSM that details the assembly of the diff output shafts, is it wrong ?..poorly written, in that it covers ball bearings and not taper roller bearings, or correct because thermal expansion takes care of the endplay (I find this difficult to believe)
sorry to the OP, feel like I am hijacking your thread !
(personally I would be/am happy with .001" endplay or pre-load on these bearings)
I would also accept as definitive a fatwah by Dick or Bill
would also be dammed pleased if I knew why quotes of the previous poster keep disappearing, requiring me to edit them back in
There you go Clive. The following chart is from the SKF Wheel Maintenance Manual. This is a graph I had to hand without looking it up in the Timken Manual. Timken donāt have a monopoly on Taper Roller bearing, other companies make them also.
As can be seen quite clearly, there is a distinct increase in bearing life from being in End Float passing Zero and into pre-load where maximum bearing life is achieved.
Although there is a sharp drop off of bearing life from the peak at circa -0.003, but the bearing life is still approximately equal at -0.005 as when in End Float of +0.005; -0.005" is reasonably high pre-load.
As I have said many times on this and other Forums, the configuration of the front wheel bearings of an E Type as is, does NOT lend itself to the bearing being set in pre-load. When the bearings are set in End Float, you will see clear evidence of the cone of the inner bearing in particular, having spun on the axle. Set these bearings in pre-load and it guarantees the bearings will spin on the axle from the Get Go. Put a spacer between the two bearings and torque the securing nut up tight, so that the bearings are locked solid between the nut and the Stub Axle Carrier, the bearings can happily be set up in pre-load.
If your bearings failed due to being in pre-load, I suspect that the amount of pre-load was too great, not simply because of pre-load. The graph above is just one of many that can be found telling the same story. The End Float range shown in the above chart is for setting the typical opposing taper roller bearing configuration of wheel bearings where the bearing cones are a close running fit on the axle, not interference, with End Float being specified for the reasons I mentioned earlier, that being the tendency for the bearings to spin on the axle due to their running fit on the axle.
All CNC machine tool spindles have the bearing set in relatively heavy pre-load. Many of the machine spindles we service run 24/7 and will rack up far more revs in a few months, than a E Type will in its thus far circa 50 year life. Typically, spindles are circa ten years old plus, that we have in for spindle rebuild.
Brent
4 Likes