Last year I had my differential rebuilt by a “specialist” and am now beginning to regret it because the pinion oil seal is still leaking. For reasons I won’t go into, I’m not going to return the car to the specialist - I would rather fix it myself.
Am I right in thinking that replacement of the seal can be done with the cage in the car and all that is required is removal of the flange and old seal and pressing in a new one? As I won’t be disturbing the bearings will it be sufficient to simply re-torque the flange nut to the previous setting or do I need to follow the workshop manual procedure to re-seat the bearing and adjust the torque required to rotate the flange?
Frankie, I have not done this myself but one way to do this is use a punch on both the Pinion nut and the flange
that it draws up against to avoid using a torque wrench to reset it. This assumes the old setting was correct.
The punch marks are used as witness marks to reset the pinion nut to its former location, and by inference
its old torque value. Pete
the 1971 should use a crush sleeve. The manual states it is a one time use. However, if changing out the seal, it is hard not to reuse as there is no torque spec for used bearings, especially with the diff still together. You can mark the nut so it goes on in the same position. The big assumption is that it was properly set up by your “specialist”. Setting it back to that spec might not be a good idea if that was the problem. You do mention it is leaking and do not mention a bearing issue so probably OK. To replace the seal in the XKE is really tight. The 1971 XJ6 is not that tight and a doable job. Make sure you polish the input flange so it does not tear up the new seal. Also, there is a spring on the perimeter of the sealing surface to hold it tight. Make sure it is there on the new seal and make sure it does not jump off when installing the new seal. There is a special seal installer from Jaguar for this job that holds everything in place.
My factory Service Manual says this job cannot be done without taking the unit out of the car and dismantling it, because of the need to replace the collapsible spacer every time, and to ensure proper backlash and tooth meshing; that’s the factory approved methodical approach.
The Haynes manual says otherwise, their method being to not replace the spacer (aka crush sleeve). Haynes describes duplicating the original torque and position of the nut.
Either way, the goal is to get the backlash in the gears down to a very small amount, but not zero.
Since you are starting with a situation where the work performed by the outsider is unknown, you should probably assume that the crush sleeve is questionable.
A crush sleeve acts like a spring of very short travel, like only a few thousandths of an inch, pushing the inner roller bearing away from the outer race.
When you get the flange and seal off, you may find the tapered roller bearing is too tight on the shaft and won’t just come out like a front wheel bearing.
But if you can get it out, you may want to try to remove the crush sleeve if you can get some kind of hook tool behind it.
The torque on the nut is 120 to 140 foot pounds while rotating the flange to ensure correct seating of the bearings, but that’s with a new crush sleeve.
With an old sleeve, you will be limited to feeling the backlash and stopping before you think you will go too far.
BTW earlier models have an axle with a solid spacer tube and shims rather than the crush sleeve.
All very helpful - thanks. It will be interesting to see if I can tell whether the seal was replaced when the unit was overhauled last year and/or why it is still leaking.
There is no noise from the diff to suggest that the bearings are not 100% - just the oil leak - so I will take the other work on trust.
There is one instruction in the workshop manual (Page Hs.12 item 6) which I don’t understand. It says:- “Lightly score oil seal recess with tang of file”. Why?
**
As you outline; it is straight forward in-car, Frankie - a fairly routine procedure…
The very important thing is to note the torque required to loosen the nut, and to mark the precise setting of the nut position as Dick says. At initial installation the torque is applied to compress the crushable to get the bearing play correct - indicated by the force required to turn the pinion, as the manual describes.
If the nut is overtorqued, the sleeve too compressed; the sleeve must be replaced - which requires diff out and dismantling. Which is why the crushable is used instead of shims - which would require several dismantling/reassembling to get the lash right, but is indeed possible. At the rear wheel bearings, also adjusted by measuring lash; easy shim access allows the use of shims…
Rob’s warning about the unknown state of the sleeve is very pertinent. In-car testing of the lash, using spec torque readings, is likely futile. The torque specified, 5 lbsft, refers to a ‘free’ diff, not connected to the drivshafts. Hypothetically, the lash should be detectable by push/pull on the nut - but unlike the rear wheel bearings; no specs are given. On the latter, the axial loads on the bearings are large, hence more play is required - the axial loads on the pinion shaft is minimal, requiring less lash…
However, the lash adjustment on the pinion axle bearings are critical for proper crownwheel/pinion mesh - so it would be very remiss of any ‘specialist’ to get it wrong. Are the reasons for not going back related to diff noise or misbehaviour that imply doubts as to overhaul quality…?
Otherwise - the leak being the only symptom; replacing the seal as described should likely be OK…
The pinion nut is tight, and if u dont have strong impact driver, make a steel bar, and bolt it to the flange to jam it while you undo the pinion nut, even by putting a trolley jack under a breaker bar
To remove the recessed seal, use a self tapper in a slide hammer
You can get a double-lipped Viton seal, but it is expensive
It is possible to use a electronic fish/luggage scales to measure pinion pre-load in inch pounds
Lightly scoring the recess might be to remove any remaining rust and bits of rubber from the outside of the old seal. The new seal outer surface is usually coated with rubber. You want to push it straight in, no cocking. I use a big socket held on it and tap lightly with a small hammer.
Here is a tool I use to pull out old oil seals.
I got it from an old retired Viking marauder who called it his Saxonator.
Yes, care as to the depth you drill. although, unless you have a Cobalt drill or better and you hit the cage, the rest of the bearing is super tough and an average bit will not bite into it.
when installing the new seal, wipe the lip with a bit of lube. A prelube of sorts. A normal weep will take ver from the Initial turns.
The relation of gears and their beatings in all differentials is almost a black art.
Generally speaking, it takes several hundred foot/pounds to compress, or to further compress, a crush sleeve. So if you had noted the loosening torque required, simply approach, but DO NOT EXCEED, that same amount of torque when re-assembling. Also, use a “medium strength” thread locker on threads. IF the pinion nut removal torque is not known, I would feel comfortable with between 120 and 150 ft/lbs AND thread locker as a re-assembly procedure.
New bearings always have a higher “turning torque” spec than used(bedded-in) bearings…figure about 60% turning torque as compared to new. So, if the “new” bearing preload spec is, say 15 in/lbs, 8-10 in/lbs should be fine for old ones…your situation.
BTW, there must be NO wiggle in the pinion, either forward/rearward or up/down, just a slight backlash between the pinion gear and the ring gear. But since the ring gear is installed, estimating turning torque can only be felt within the backlash measurement, a bare fraction of a rotation, until the pinion tooth contacts the adjacent ring gear tooth.
No, the hook seal puller works fine, you just turn it sideways.
Of course you understand it wrecks the seal as it pulls it out.
The Lisle tool in the video may be intended to pull the seal out without ruining it. Though usually the reason for pulling it out is its no good.
Funny the guy didn’t know how to pronounce the name Lisle, which rhymes with file and style.
Not to some of us that enhance vocabulary by reading as opposed to by hearing. I’ve found that is me, many times!!
Lisle, an old and respected name in tools. Not sure if I still have tool to depress valve springs in a side valve engine. Not Frd’s V8, Odd means of valve installation and removal. A big bar.
A unique tool that used hydraulic pressure was available to the professionals.
Carl
