Diff pinion and oil seal replacement

Care needed!

When replacing the pinion oil seal you need to be aware that you could disturb the settings on the diff of some models.

The reason for this is that the pinion bearings are set up with what is known as pre-load, which can be thought of as the bearings are squashed together a bit rather than having play.   The purpose of the pre-load is simple: it better braces the pinion in place against the side thrust from its contact with the crown wheel. With no (or inadequate) preload the pinion would flop about a bit and result in wear and noise.  Once the pinion nut is released, the pre-load will disappear so it is important to set it correctly when reassembling.

The preload is achieved by means of a spacer between the two pinion bearings.  Tightening up the pinion nut clamps all three items together (inner bearing, spacer, outer bearing). The shorter the spacer, the more the bearings are pushed together and so the higher the preload.  This provides the way of adjusting the preload to the correct value.

Early cars (up to about mid 60’s) had a solid spacer with adjustment via shims.  This gives very good, accurate control – but is quite time consuming to set up.  Hence a collapsible spacer was introduced to both the tube axle fitted to MGB’s and also to Midget diffs.  (I’m not sure the exact change point in the Midget, but all 1275 diffs have it and some of the earlier ones may too.)

It is the collapsible spacer which can provide a bit of variability when changing the oil seal.  Solid spacer ones go back exactly as they were.

The preloading of the bearings results in drag on them when you turn the pinion flange. This gives a method of measuring the preload, and so it is expressed in lb-in of torque required to turn the pinion flange. BMC had official tools to measure the preload torque, but there are other means:

  • I have fashioned a bar to bolt to the flange so it holds it when doing up the pinion nut so that it doubles as a preload measurer via its weight providing the required torque.
  • Another way of measuring it without the official tools is to wrap a piece of string round the flange and pull on it with a spring balance … for example, at a radius of about 1.5″ 4-6 lb-in translates to 3-4lb pull on the string to get the pinion to move.

Seal replacement:

The seal replacement procedure for cars fitted with a collapsible spacer is quite simple in principle:

Measure the torque needed to move the pinion (string and spring balance) beforeundoing anything, then when replacing do tighten a little at a time until the original torque is reached, or bit more if it was less than 4-6 in-lb.

The steps involved in doing this are set out in the MGB manual as:

  1. Remove brake drums (to eliminate drag for step 2)
  2. Record the torque required to rotate the pinion
  3. Remove and refit seal much as you would expect to do
  4. Tighten the nut gradually until resistance is felt
  5. Rotate the pinion a few times to settle the bearings
  6. Measure the torque needed to rotate the pinion
  7. Tighten a little more if required
    Repeat 5,6,7 until a torque reading is achieved which is the GREATER of
    (i) what you measured in step 2 OR
    (ii)4-6 lb-in

(To give you a feel for what to expect, 100 ft-lb is usually not enough to collapse the spacer further whereas 140ft-lb usually will.)

The manual further notes in bold type:

CAUTION: Preload build-up is rapid, tighten the nut with extreme care. If you exceed the target figure the diff must be dismantled and a new spacer fitted.

Important: The Midget manual instructions are the same as those for the B which had the solid spacer and shims so appear to not have been updated when the collapsible spacer was introduced.  Following the procedure for the solid spacer – which is that outlined in the Midget manual – is likely to result in further crushing of the collapsible spacer and excessive pre-load on the bearings and thence premature bearing failure.

There is a simpler method of replacing the seal:  Mark where the pinion nut was in relation to the pinion shaft before removing it, then when reassembling do it up to there.  This is a simple means of restoring it to its setting before the seal was replaced. What it misses compared with the manual procedure above is checking whether the required minimum preload has been reached. It’s worth doing while you are there as over time the preload can reduce due to gradual wear in the bearings.

Notes about pre-load figures:

12-ish lb-in of preload is used when setting a diff up with new bearings etc. It is measured with only the pinion installed – no crown wheel or cage in place. Once the set-up is installed and gets a bit of mileage on, the pre-load gradually reduces. Hence:

The 4-6lb-in figure is what the (MGB) manual lists as a MINIMUM figure if replacing the pinion seal without stripping the diff and replacing bearings etc.   What the manual actually says is to tighten to the existing preload (that’s the mark-the-nut method) but if that pre-load is less than 4-6 in-lb then tighten it till that minimum figure (4-6) is achieved.

Notes on tightening the nut:

The spacer starts to buckle before the bearings are in contact. Once the bearing clearance is taken up and some pre-load sets in, this will account for some of the effort going into the torque wrench.
So the spacer collapses before the tightening torque of 140 lb-ft, but once there is also a bit of resistance from the bearings nipping together then the torque needed to further compress the assembly gets larger. Usually, but not always, 140 lb-ft and the requisite torque figure on new bearings come up about together. But not always, so I always do a bit at a time until the pre-load figure comes right. So in fact you could do that bit without a torque wrench if need be.

However, a torque wrench is useful to give reassurance that the nut is suitably tight:

If you get the pre-load below 140 lb-ft, then don’t fret about it coming loose unless it’s under the 100 lb-ft mark. If less than 100, I would renew the spacer. A bit more work, but ultimately less and cheaper than renewing the bearings and possibly the CW&P as well if the nut undoes itself.

The spacer

On occasion, when in a hurry (imminent event) and being out of stock of the spacer and not having time for another to wing its way halfway round the world, I have recycled the spacer with a thick-ish shim (20 thou-ish) between it and the bearing. While this has proved perfectly OK in practice, my preference would be to fit another spacer if I possibly could. But it’s a handy thing to know if you are desperate. (But not to save a few cents!!)

What happens if the nut comes loose:

We had once an issue with the diff of our K-series engine car Widget that took some while to track down as the symptoms were propshaft out of balance. It was diabolical to the point of scary. Three propshafts rebuilds and re-balances later and no improvement we were still scratching our heads when we were given a tip-off that a simliar symptom that turned out to be the diff.

And yes, it was pinion bearing preload gone. It appears what happens is the spacer collapsed further under extreme load (in this case a decade of autotesting behind a K engine) which means that the flange nut has nothing firm to grip against any more. So inevitably it loosens off, which if bad enough gives that gross out of balance feel.

We chucked out the collapsible spacer and set it up with a solid one and shims to give better control in this respect. More work but problem solved.

I have come across other hard working diffs where the pinion had loosened off, but the symptoms had always been a sudden development of diff whine rather than no whine and severe vibration.

Interesting footnote:

Due to the usual time constraints of an imminent event, when we found the K Midget’s little problem just outlined we ignored the bruised-looking wear pattern on the crown wheel & pinion (none in stock) and just fitted a new spacer with the gears set up to give the best contact pattern we could -promising to do it properly later. It turned out that (amazingly) it ran almost quietly – just a faint murmur. A year later it is still as quiet and this discussion has reminded me we have yet to do that!

Recent Posts