And the IRS is out... with pictures

, ,

When originally assembled, Aristides; the nut is torqued in increments.measuring the preload each time. As nut torque is increased, the spacer starts to collapse - which influences bearing play. There is no specific torque on the nut - it is tightened until the spacer is compressed enough to give the correct preload, indicating correct bearing play.

So the initial nut torque may be anywhere, nominally up to some 130 lbsft. It is immaterial to refitting after a seal change - the nut is torqued to the mark made before it was unscrewed. In this position; the original bearing play is retained by the thickness of the collapsed spacer.

(In theory; if bearings are worn; further, very slight, nut tightening may be done to correct bearing play - but in practice; worn bearings should be replaced…)

The lash on the pinion shaft is the play between pinion and crown wheel. The shaft is moving freely, except for the friction between the shaft and in the bearings - and this is measured by the preload torque on the axle. The preload does not refer to the load/torque on the nut - but the force required to start the axle moving. It’s important that the axle is fully backed, taking up the lash, before the preload torques is measured…

The specs given for preload torque with old bearings is 20 - 25 lbs/in and with new bearings; 30 - 40 lbs/in.

Your ‘8’ and ‘11’ lbs/in is somewhat odd - and does indeed imply that the collapsible was not crushed enough initially to give the proper bearing play. That the preload torque increased with 80 lbs/ft nut torque does imply further crushing.

If your readings are correct; in theory; you should then further increase the nut torque until you reach a preload reading within the specs given. I would dearly like to know what the readings, including releasing torque on the nut before unbolting! But I see no other reason for the low reading - unless something is seriously amiss with the pinion shaft bearings. Or your measuring procedure/tools are plain wrong…

And if the preload torque exceeds 45 lbs/in; the bearing play is too tight - and the collapsible spacer must be replaced - going through the initial preload set-up. Which would be ‘inconvenient’…

The whole point of these operations is to get the bearing play correct - as measured by preload torque…

xj6 85 Sov Europe (UK/NZ)

1 Like

No it does not Frank.
I just did the output shafts, same principle, and the torque required to crush the sleeve was at least twice that.
I had to use a 3 foot bar and two hands… I guesstimate that it was around 200 lbs.ft.

I know that the seal has been changed before by a mechanic, so it’s unknown if it was done correctly.

The differential had a very faint whine on deceleration but otherwise it was very quiet, so I decided to play it safe and avoid unpleasant surprises by excessive preload on the old bearings.
When the IRS goes back and I can test drive the car, I will be able to see if what I did was good or not…!
I can always go back and re-torque.

From what I have read it’s 10 - 20 for new bearings… Please send me the document if possible.


Rear Brakes

Safety Wire

And an excellent tutorial:

The sleeves are different on the two, Aristides - the apparent longer sleeve on the pinion is likely more ‘crushable’…

But the main point is; whenever the nut is turned - it moves inwards. Which it can only do if the sleeve gets shorter - ie is crushed more.

The actual torque needed on the nut is in itself immaterial. It is the inward motion of the nut that influences the bearing play - and torque must be applied as required…

In principle; the bearing play allows slight for-and-aft movement of the pinion on the crown wheel. The whine may be induced by altering the contact points between the two due to excessive bearing play?

And, of course, if a PO was sloppy in setting the correct preload torque/bearing play - he may also have been sloppy in mating the crown wheel and pinion during assembly. Or mismatched crown wheel and pinion ‘pairing’. And, of course; wear in itself may introduce whining - as may incorrect fluid level/type in the diff.

Generally; excessive bearing play is less destructive than too tight - but it is not necessarily harmless…

Readjusting preload torque after road testing seems valid - With correct procedure and equipment; I can see no harm is setting things to specs…

I took it from the manuals description of the full diff overhaul/assembly 1985 version…

Which does not preclude that other versions may have different specs.

Generally; I’m a bit apprehensive about the friction ‘new/old’ seal and ‘new/old’ bearings isn’t more explicitly treated. But it may all be included in the spread of spec tolerances…?

xj6 85 Sov Europe (UK/NZ)


Shimming the Fulcrum to Diff gap.

Wishbones installed

Bottom Tie Plate and Brake lines installed.

1 Like

Half Shafts

U-joints replaced, cleaned and painted.
And an excellent video on how to properly replace U-joints:


Rear Wheel Bearings & Seals

Hub Bearings End-float or Preload ?

The manuals for early cars calls for 0.10mm (0.004") End-float but for later cars they changed it to Preload…
My hubs were a mess, someone was there before me and one side had excessive play but the other side had zero.
In my extensive research in the archives I found no definitive answer…
The fact the I saw recurring though was that bearings that were set with the desired End-float had the tendency to increase it after some time as they set in.
So I decided to go with a very slight Preload, and I have set them with my torque scale to a about 2 of drag, almost identical to the seal drag.

Next big question was if I should use Locktite on the stub axel splines or grease / anti-seize.
As mine had not any Locktite, and no evident signs of wear or clunks, I decided to go with grease for the great benefit of ease of disassembly.

1 Like

I use loctite on the splines when I did mine, with units that have seen a few miles there is the possibility of a clunk developing going from drive to reverse, just keep an ear open for that to start happening, if it does then I would recommend stripping off the grease and going with the loctite. It will require using a puller next time but I just use a stock 2 leg unit,

Unless specific preload is given, Aristides - the endfloat, bearing play should be used. In any case; having used preload, the play should then be verified as well…

The play is changed by mowing the inner race along the axle to vary the distance between the inner and outer race - the play. When computing shim thickness it’s important that the inner race is moved to a position where the shim makes an accurate displacement of the inner race. That the play changes after assembly implies that this was not done…?

xj6 85 Sov Europe |(UK/NZ)

Yes Robin, this is exactly what my rationale was.

Yes that could very well be the case.
I tried my best to be sure that the races were fully inserted to the hub, but it was done by hand as I have no press.

I actually re-thought the whole matter and went back and removed some of the preload by adding a 0.1mm shim.
This gave me a drag torque of about 2, almost identical with the seal drag torque and no play, so I think I am very close to 0 preload / end-float.
I am happy with that.
I also revised the original post to avoid confusion.


[quote=“Aristides, post:39, topic:378381”]
“My hubs were a mess, someone was there before me and one side had excessive play but the other side had zero.”
“The manuals for early cars calls for 0.10mm (0.004”) End-float but for later cars they changed it to Preload…[/quote]"

I have read this several times on the JL E-type list over the years and I believe this is incorrect.
The series 3 XJ manual specifies end float… not pre load for the hub carriers. The later XJ40 manual specifies preload for the hub carrier with outboard brakes. Perhaps this is the source of confusion.

Wheel bearing end float on the classic irs is set via a thickness spacer which resides between the oil seal track and the inner bearing, inner race. The method for achieving the desired end float is clearly explained using the special tool in the workshop manual.

Yes indeed, but the principal of operation and bearing arrangement are exactly the same.
What is more confusing is that on the XJ40 with outboard brakes there are much greater temperature differences involved and one would expect end-float to be called for as to compensate different expansion rates.
Maybe because of the presence of the disc brake, a no play arrangement was needed?

It must also be noted that preload is specified at the output shafts bearings, and it’s the exact same tapered roller bearings configuration.

My guess is that end-float was chosen because it’s so much easier to set-up and measure, much more forgiving at mistakes either by the factory either by the service mechanics, and the minimal reduced bearing life was not much of a concern.


It’s a matter of convenient access for measuring, Aristides - and indeed convenience of altering bearing play. With shims adjustment some dismantling is required to change shims. Like the diff’s pinion axle bearing requiring major, and possibly repeated, dismantling if shims were used…

Both methods aim for the proper clearance for adequate lubrication between be fixed and moving pars of the bearing - without excessive axial and lateral movement.

With shims generally; play is independent of nut torqueing - which arguably is better for keeping assemblies together…:slight_smile:

Endfloat measures play directly, when possible - and is very exact with proper tools. While preload uses resistance to turning - relating to overall ‘friction’. Which relates to bearing size and lubrication type - and other factors like masses involved involved and other friction elements. More possibilities for errors…

xj6 85 Sov Europe (UK/NZ)

Hubs - Outer Pivot Bearings.

The bearings are in very good condition.
Even though there are marks from the rollers, the race surfaces are absolutely smooth, so I don’t think they need to be replaced.

Hubs - Grease Vents modification

Plugged center vent hole and drilled two holes outside the bearings, as suggested in Kirbert’s book.
Now new grease will have to go through the bearings before it comes out from the vent holes and will actually lubricate the bearings.

Hole distance from edge of Hub: 7mm
M3 x 5 screw
Chamfered the holes and added a small O-ring.

1 Like

Very nice… Zerk is in as good a location as possible.

Curious: What brand is stamped on the bearings and seals?


All bearings in the IRS so far were Timken, differential included.