S2 Differential Pinion Oil Seal Replacement

I’d like to replace the pinion oil seal on a '69 S2.
The cage is out of the car.
Do I have to press out the pinion and dismantle the internal assembly?
Is there a way I can extract the seal from the front after removing the nut?
Any suggestions? I need help.

Hi Thomas,

There are two things I will not work on, because the cost of error is very high. (ie the unit must be removed again). They are automatic transmissions and differentials.

If it were mine I would take it to a qualified shop and have them do all three oil seals in the differential as well as put a new gasket on and check it out internally.

I did this 20 years ago with new bearings needed and have not touched it since!

Just a suggestion

60 OTS

I agree. I have no interest in going inside the case.
I wonder if I could remove the nut, weld two threaded rods to the seal face and extract it with a bearing puller?

The input shaft has a crush spacer that is crushed to give a preload to the bearings on initial set up by tightening the large nut…its possible to replace the seal but first punch mark the end of the shaft and the nut…count the turns needed to remove the nut…replace the seal…you may need to chisel it out…note when refitting the correct way around that the seal fits and the thin paper gasket under the seal…ensure the flange is nice and smooth…refit the nut to the exact same position…if you dont you will crush the spacer more and affect the pre load…hope this helps…Steve

Thanks Steve
There is no external flange on the metal seal casing under which you could pry. Therefore I’d have to drive a chizzle between the seal casing and the diff housing. Is this what you mean? I’d be concerned about damaging the housing.

You do what needs to be done to get it out…iv used very sharp wood chisels so it cuts into the metal of the seal…if concerned then keep away from the edge of the case…Steve

I’ve had the (mis) fortune of doing a number of differentials over the years, both the older ones where pinion preload is set by shims, and the latter ones set by crushable sleeves. I don’t know what your is - you need to do some research to find out. The parts manuals I have access to suggest early Ser II’s were shimmed - my late production '68 is shimmed. Jaguar says under no circumstances can you take the flange off and replace it on a diff with the crushable sleeve without installing a new sleeve and resetting pinion preload. My one experience in defying the manual on this point ended in a second rebuild two months later. However if you have the shimmed diff then you can remove and replace the flange and seal without issue, simply re torquing the nut. The seal is not difficult to get out. I drilled a small hole in the internal metal support of the seal and threaded a self tapping screw into the hole and pulled the seal out with pliers. .

Thomas I should add - take a carful look at wear from the seal on the flange. If you can feel any at that point I’d recommend you install a Speedy Sleeve over it


Hello Steve,
I have trouble with this statement and the one in the S3 Repair Manual where it states that if you inadvertently exceed the the 140lb/ft torque, the current crushable sleeve must be replaced; a nonsense statement in itself,as it often requires more than that to get the correct torque to rotate the Pinion Shaft, which is the correct test for bearing preload.

The crushable sleeve is positioned between the bearings. If the 140lb/ft torque were to applied in the absence of the crushable spacer, the whole force would be applied to the taper roller bearing cones and cups and great force would be required to turn the pinion shaft. Accordingly, the purpose of the crushable sleeve is to soak up great amount of the axial force being applied via the Companion Flange nut.

If you follow the process you will get what I mean.

  1. Nut is tightened, drawing the bearing towards each other until the outer bearing makes contact with the crushable sleeve. At this point quite a lot of End Float can be detected just by feel; no measuring devices required.

  2. Continue to tighten the Companion Flange nut and the crushable sleeve starts to collapse, allowing the bearings to continue to be drawn towards each other. Still no preload and still detectable End Float.

  3. The process in 2 continues until the cones of the Taper Bearings make intimate contact with their respective cups. At this point there is neither End Float, nor Preload.

  4. Beyond the point of initial, intimate contact with their respective Cups, the Cones only draw closer together by a very small amount until their firm contact arrests any further coming together. Imagine a number of flat washers between the head of a bolt and nut. Once the nut is tight on the stack of washers, the washers can’t come closer together.

  5. Once the Tapered Cones of the bearing set have made firm contact with their respective cups, apart from a small amount of distortion of the differential housing, there is no further coming together of the Bearing Cones and therefore, no further collapse of the crushable sleeve. Further load is applied to the Bearings by the stretch of the weakest component of the Pinion, that being the Threaded Journal that the Companion Flange nut interfaces with,

I’m not one to make an hypothesis. I’ve tested this a number of times, with the tests subsequent to the first to ensure that the first wasn’t a fluke. My test was:

  1. continue to tighten the Companion Flange nut past the point of optimal bearing preload, in increments of 10lb/ft torque.

  2. after each incremental increase, the preload of the bearing was measured using an lb/inch torque wrench to measure the amount of torque required to rotate the Pinion Shaft. The outer bearing was then removed so that the crushable sleeve could be retrieved and measured

  3. the same crushable sleeve and bearing reassembled and the Companion Flange nut tightened to the same torque as previous and the prelaod checked by the method described in 2 immediately above.

  4. from the point of optimal bearing preload, to a Companion Flange nut torque load of 250lb/ft, there was no change in the length of the crushable sleeve.

Its doubtful that the bearings that have been in service for some time will have the same preload as when it was correctly set when the differential was last refurbished, or assembled initially. Accordingly, tightening the Companion Flange nut to the same point as before removing the Oil Seal wouldn’t lead to failure due to excessive preload.

When setting the preload of the Pinion Bearings, its done without the drag of the Crownwheel, by measuring the amount of torque required to turn the Pinion. This amount of torque has increase from a mean of 10lb/in in the day of the S1 through S3 cars, to circa 26lb/in for much later cars (post S3s). Although not as well, but the preload of the Pinion Bearings can be measured by the lb/in torque required to rotate the Pinion Shaft method with the differential assembled. Take up the backlash in the opposite direction that the lb/in torque test is to be carried out and measure the torque required to rotate the the Pinion Shaft through the backlash zone.




Hi Bill…so are you saying that you can remove the nut and flange…replace the seal then retighten the nut untill you have preload of 26lb/in…and not just replace the nut back to its original position…Steve

Hi Bill,

I have learned much from your writeup but I will still not touch my differential! BUT I would bring it to you if you lived here in Vancouver!!

Dennis 69 OTS

You cannot set used bearings with the same preload as new ones. This will be too tight.
Assuming that the bearings feel ok before you change the seal, use a center punch to mark the shaft, nut and yoke. Reassemble to the same place or max 5° more. Keep the washer in the same orientation.

The instructions are different depending on whether it’s shimmed or uses crushable sleeves. On the former the preload is set with shims and the specified torque on the pinion nut has nothing to do with it. On the latter there is no torque figure suggested because you turn the nut in until you achieve the desired pinion preload for the crush of the sleeve, measured with a lb/in wrench (in which both the nut and the pinion are turning simultaneously). Bill is talking about the latter type of diff. Pinion preload for a shimmed E Type Ser I diff is stated to be 8 -12 lb/in, a small figure, with no differentiation between new and old. On the series III XJ12 diff which uses crushable sleeves pinion preload is stated to be 20-25 lb/in for old bearings and 30-40 for new.

From these exchanges I’m hearing that the oil seal can be removed after removing the pinion nut and compression flange. However, the relevant question is whether the ‘69 S2 diff system uses a crush sleeve or shims.
How can I confirm this?
If it uses crush sleeves I’ll have to dissemble the entire diff which is what I’m trying to avoid and frankly won’t entertain.
What’s the consensus, crush sleeve or shims??

The technique I described above will work with a crush sleve or shims. You do not need to disassemble the entire unit, simply install the new oil seal and put it back together. By putting the nut back in the same place you are going back to the same preload as before.

A very excellent and detailed answer as usual

a note on getting out the pinion oil seal…EASY job

drill 2 x 1/16" holes in the lip, make sure it is not too deep

screw in self tapper on the end of slide hammer

I have extracted many blind seals this way, and it has never failed.

when I have done them before, I did what Bill described (I think)

mark the nut

measure pre-load in inch pounds, both before and after the seal is out,

put if back together with the same pre load as before

(the pre load should be close to spec for used bearings, if it is too sloppy when you measure initially, it may have lost its pre-load)

use loctite on the pinion nut

Well the Ser II service manual, and parts manual available on the E Type Forum UK states all Ser II diffs are the same as Ser I diffs which are shimmed, and the parts manual shows only shims, so I think you are golden.

Hello Terry,
Preload is preload on similar taper roller bearing, irrespective of how it’s achieved; via shims or crushable spacer, it makes no difference. The reason for the rather large variation between the early and late cars is evolution; they learn that the greater preload worked better, but it took some time to get there. The crushable spacer was introduced as a time saver so that pre-load setting could be achieved in the one assembly, rather than ginning around with the multiple tries that may eventuate with shims.

When load is applied to the Pinion/Crown-wheel assembly under acceleration, the preload actually diminishes. If the preload was light in the first place, the Pinion will move relative to the Crown-wheel and a different tooth contact results during acceleration, as the preload drops off.

I’m lucky to have the ear of a retired national production engineer of Timken Bearings. This fellow was a globe trotter and responsible for many design changes in differentials used by the car industry.

Torque used with the shimmed arrangement doesn’t set the preload per se, but you need to have the Companion Nut tight. The 140lb/ft torque in reference to the Companion Flange Nut, mentioned in a number of Jaguar publications is not within a bull’s roar of the correct torque for the thread of the Pinion Shaft. Accordingly, that would be a very safe figure to use when using shims to set the pre-load

Indeed there is a Torque figure specified for Companion Flange Nut for an S3 E Type which used the crushable spacer arrangement, that being 120lb/ft to 140lb/ft torque. (see the extract form the S3 Repair Manual below). In my opinion, the whole explanation in the S3 repair manual is flawed, as there is no reference to drag on the bearings (pre-load) as measured with an lb/in torque wrench acting on the Companion Flange Nut to turn the Pinion Shaft, just a an upper and lower torque tolerance range. On some assemblies, I’ve used in excess of 200lb/ft torque to achieve the correct pre-load by measurement of the torque required to rotate the Pinion Shaft.

With regards to the higher torque figures for new bearings (30-40lb/in) this amount of torque doesn’t last long. The bearings, as precisely made as they are, still bed in to some degree. You would find that the pre-load on these bearings would be down around the 20-25 lb/in mark in a relatively short time.

There is no problem in re-setting the pre-load of existing, used bearings that are in good condition. As stated earlier, pre-load is pre-load. 25lb/in torque doesn’t equate to something higher, just because the bearing is used.

I do a lot of CNC machine tool spindle rebuilding. Most spindles nowadays use ultra precision, angular contact bearings, arranged back to back, or front to front, with the specified pre-load built into the offset of the faces of the inner and outer races. Correct preload results when the bearing assembly is pressed together via the retaining nut. Some earlier CNC machine spindles used Taper Roller bearings, particularly machines destined for the US market. It was common place to check and reset the pre-load on existing used Taper Roller spindle bearing, simply because is was easy compared to Angular Contact Bearings that had the pre-load built in. The Spindles and Spindle Bearings of typical CNC machines used in production will see more revs in a few weeks than a 60 year old E Type will in its life.



1 Like

A very good source of parts info is the SC parts catalogue…there were 3 typess of diff used on 6cyl cars type1-3 type 2…the 3 both use collapsable spacers…detailes here

Hello Terry,
Not so. There were three different type number of differentials used in the S2. Only Type1 used shims. Type2 and3 both used a crushable spacer. Accordingly, its not a gimme that the OP’s differential used shims. I’ve seen many S2 differentials and I’m yet to see a Type1 used with an S2. On probability, its more likely that the OP’s diff will use a collapsible spacer.

Sorry Steve for cutting across your Post. I was still typing when you made your Post. Basically, we have said the same thing.




You know you just can’t trust anything you read any more. I always believed Ser II diffs were a mixed bag, but then I went to the Ser II parts and service manuals on the UK E Type forum and there it plainly states - no only the original type (shimmed) was used. Similarly the SNG diagrams under Ser II that I looked at also shows only a shimmed type.