Fitting an Ashcroft ATB to my Salisbury axle

[Peaklander]

This is for the rear axle of my 1996 110SW 300TDi which is at about 190,000 miles. I have been suffering a 'clunk' for years and finally decided that it is at the rear axle diff. I managed to find play in one of the upper arms, at the chassis end but after checking the outer driveshafts at the flanges, I decided that the Salisbury final drive needed 'a looking at'. This is detailed in my vehicle thread which is here. The result is that I am fitting the relatively recently released Ashcroft ATB, which requires the crownwheel to be moved over from the original diff.

So now I am at the point where the casing is stripped, except for the pinion outer races and the suspension brackets have been replaced and the axle has a coat of red oxide primer and one coat of chassis black.

With acknowledgements to @ejparrott and his thread here, plus Mike at Britannica Restorations, with his series of videos on this very subject, I am going to document my efforts in this thread. The reason for writing as I go, is so that I can be sure that I have thought it through and won't make an error. I am NOT writing it because I know what to do. I am a novice and a complete Salisbury virgin.

I have gathered together some assembly / adjustment items. The pinion bearings will be changed and so I need to check the pinion height, relative to the lowest point of the diff carrier bearing surface. This requires a height block of 30.93mm. This was made for me by a local machinist. I also have a pair of dummy bearings for the carrier and a new pinion inner bearing, outer race that has been ground on the OD to be a push fit into the axle casing. These dummies are needed because the adjustment shims for the carrier are placed before the bearing inner races are pressed onto the shafts and for the pinion, they sit in between the outer race and the axle case.

 

Pinion first.

The dial gauge is zeroed at the nominal height and then the change from the nominal height is measured, with the base on the end of the pinion and the gauge probe sitting in the lowest point of the carrier bearing surfaces. This measurement is taken at both sides.

This arm has one tightening knob that simultaneously tightens all three ends of the two red arms, so the whole adjustment and positioning is done with one final turn on that knob. The two bearing surface measurements need different arrangements of the arm, which needs to be first set up on the pinion. The whole thing is carefully moved to the height gauge to be zeroed and then back to the pinion where the relative height of the bearing surface can then be measured.

I removed the magnetic base as it rocked on the pinion and instead used a flat non-magnetic base that is just about heavy enough to hold the gauge unless the arm sticks out too far.

 

 

On the carrier side opposite to the crown wheel, I measured +4 thou. (0.004") from nominal.

 

On the crown wheel side it was +8 thou.

 

 

A pinion is matched to its own crown wheel and the pinion manufacturing tolerances result in varying pinion heights. I don't know where this machining variation is but it is measured with the pinion nut tightened to give the correct turning effort at the propshaft flange. This variation is marked on the end; mine is  +9 thou. (0.009") and has be be allowed for when selecting the shims to place between the pinion inner bearing race and the axle casing. In this case the +9 thou is subtracted from the measurment with no shims in place, the result being the shim thickness required.

The other number (4912) is a serial number that should also be found on the matching crown wheel.

 

 

So as the shims are still in place because the inner bearing race is still there, the measurements for this pinion should show that it is 9 thou below that nominal height. Well it isn't quite but I have already loosened the big nut and not correctly re-tightened it so I suppose it isn't accurate but at least I know that I can make these two measurements.

Edit: something doesn't feel right about this. I am reading the manual as I do it and it does have one error that I have marked. However the main question is, in the example they give for a +3 thou pinion, why should I see 0.003 below nominal (instruction 88 shown below), when it has been set correctly? If, following the measurement, +0.003" of shim is removed to account for the pinion, why isn't the dti required to read 0.000"?

 

 

 

Once I have drifted out the old race, I can recover whatever shims are there. The temporary race can be inserted into the casing. Then I will pull the bearing off the pinion and replace it with a new Timken. I don't know the brand of the old inner bearing; I had confused myself by saying it's Koyo but that is the outer bearing. I won't be able to see the marking on the inner until it has been removed.

I feel that replacing the main pinion bearing is "a good thing". I don't want to go through this again later, just do it all correctly this one time. Do I sound nervous?

Edited February 25 by Peaklander

[Peaklander]

This pinion end-marking that denotes a 'variation from nominal' is a bit strange. If there is no variation, so no marking, then the pinion is shimmed to the value closest to that obtained when comparing the height block of 30.93mm with the height between pinion top and the curve in the carrier bearing surface.

As it says in the good book: 

"The figure marked on the end face opposite to the serial number indicates, in thousandths of an inch, the deviation from nominal required to correctly set the pinion. A pinion marked plus (+) must be set below nominal, a minus (- ) pinion must be set above nominal. An unmarked pinion must be set at nominal".

If there is a 'variation from nominal' marked, as there is on mine of (plus) 9, the pinion is set 0.009" lower. When the shims are placed correctly, this 9 thou is what is seen on the dial gauge indicator - a deliberate deviation from 'nominal'.

What I don't understand is what this difference is, between different pinions, that requires the top of the pinion to be set higher or lower than the desired point. It can't just be a measurement because that is captured anyway. Is it for differing expansion coefficients? What is going on? 

I don't know that it actually makes a lot of difference because of what comes later in the adjustments:

1) The end-float of the diff carrier is found, the necessary pre-load added and then the carrier has that total amount shimmed to the side that pushes the crown wheel away from the pinion.

2) The backlash between CW and pinion is found. If it outside the allowance (0.006 - 0.011") then the shims are moved little by little across to the other side of the carrier.

3) Then using 'blue' the mesh pattern is obtained and if it isn't correct then the pinion height is possibly going to be adjusted anyway. The nominal height plus/minus the marked number on the pinion, is only a start!

 

I hope I have got this right.

[fmmv]

I think the pinion marking reflects its desired position relative to the CW as a matched pair. Presumably that as a result of manufacturing variations.

[Hybrid_From_Hell]

My advice on this would be to remove the race and find the maker

 

IE if Timken replace with Timken,

if SKF replace SKF etc.

This way you can just change the bearing - reshimming the head height is not only not needed............. but unadvisable !

Nige

[Peaklander]

8 hours ago, fmmv said:

I think the pinion marking reflects its desired position relative to the CW as a matched pair

Yes that makes sense. A relative position adjustment rather than something about its absolute size.

 

8 hours ago, Hybrid_From_Hell said:

reshimming the head height is not only not needed............. but unadvisable !

I will measure the existing once more to check that my 4 thou and 8 thou are correct and then punch out the old race and see what it is. I do have the undersize Timken outer race ready to use, so that I can fit and fiddle with the pinion shims and I already bought a new Timken bearing to use, so I could go further if I have doubts about the existing set-up.

[Hybrid_From_Hell]

 

 

The thing with pinion height is LR were / are very poor on tolerances - the tolerances are sometimes quite staggering

so - the diff is built at the factory "Within tolerance" but in real terms is actually a bloody long way off what it COULD / SHOULD have been ................

 

This means the pinion being "technically" WRONG.. ( but actually from LRs point of view "WITHIN TOLERANCE" )  means quite simply its more of a factory 'that will do near enough'

and has then spent its entire life not quite in the correct position / or what could have been better - and the pinion and the crown wheel have meshed from birth it these positions

and have worn themselves for however many x,000,000 of miles. 

 

If you now start setting pinion heights and in effect make the pinion with the new bearing 'Closer' the where the pinion 'should' have been at day 1 - its will probably howl its head off !!!!!!!!!!

 

And this is why I say use the same Brand head bearing as to exactly whats in there.

Yes, the pinion head height will have changed as the bearing has worn, but that's still only fractional, and, all the time it has had that bearing in the wear has been in a small range .

if you put in the same bearing replacement and from the same manufacturer then the pinion will still be asking to work in the same range again...

 

We did a test once set up a pinion to 0.0 +/- and then swapped to SKF NTN - and the readings were different !

There is a huge risk of you over thinking this fitting of the ATB, the pinion isn't being changed, and whilst Head / tail / seal / drive flange is an excellent idea replace like with like and leave alone head height !

Nige

[Peaklander]

OK @Hybrid_From_Hell I follow that advice. Thanks

Just one more question then. I use the same brand bearing and shims as found and the pinion sits in there as before. The crown wheel is being moved across to the ATB but with new, same brand carrier bearings. Won’t I need to look at the new mesh pattern and maybe need to adjust pinion height anyway or is that all going to be tuned at the carrier shims?

[Peaklander]

I knocked out the race and recovered the single 0.030 shim in there.  It is a Koyo bearing, now made under the Jtekt banner. They are not available on the shelf locally and I half wanted to admit defeat and put it back. The bearing suppliers I rang have each told me that there's no need to match the brand.

I have found one on-line that will be sent today, Royal Mail 1st class - so I am committed now. I also took delivery of the Vevor bearing puller that will be used. It's an expensive way to pull a bearing but I will give someone else the chance to use it once I have finished! I'll practice on the diff carrier first, just to see how it behaves.

[Snagger]

12 hours ago, Hybrid_From_Hell said:

My advice on this would be to remove the race and find the maker

 

IE if Timken replace with Timken,

if SKF replace SKF etc.

This way you can just change the bearing - reshimming the head height is not only not needed............. but unadvisable !

Nige

The problem is that this assumes the shims were reset correctly when the bearing was swapped out to the Koyo.  I think this is unlikely - I imagine the shims were left as installed in the factory, despite the bearing brand change.  However, it should still mean the shims are correct.

[Snagger]

20 hours ago, Peaklander said:

This is for the rear axle of my 1996 110SW 300TDi which is at about 190,000 miles. I have been suffering a 'clunk' for years and finally decided that it is at the rear axle diff. I managed to find play in one of the upper arms, at the chassis end but after checking the outer driveshafts at the flanges, I decided that the Salisbury final drive needed 'a looking at'. This is detailed in my vehicle thread which is here. The result is that I am fitting the relatively recently released Ashcroft ATB, which requires the crownwheel to be moved over from the original diff.

So now I am at the point where the casing is stripped, except for the pinion outer races and the suspension brackets have been replaced and the axle has a coat of red oxide primer and one coat of chassis black.

With acknowledgements to @ejparrott and his thread here, plus Mike at Britannica Restorations, with his series of videos on this very subject, I am going to document my efforts in this thread. The reason for writing as I go, is so that I can be sure that I have thought it through and won't make an error. I am NOT writing it because I know what to do. I am a novice and a complete Salisbury virgin.

I have gathered together some assembly / adjustment items. The pinion bearings will be changed and so I need to check the pinion height, relative to the lowest point of the diff carrier bearing surface. This requires a height block of 30.93mm. This was made for me by a local machinist. I also have a pair of dummy bearings for the carrier and a new pinion inner bearing, outer race that has been ground on the OD to be a push fit into the axle casing. These dummies are needed because the adjustment shims for the carrier are placed before the bearing inner races are pressed onto the shafts and for the pinion, they sit in between the outer race and the axle case.

 

Pinion first.

The dial gauge is zeroed at the nominal height and then the change from the nominal height is measured, with the base on the end of the pinion and the gauge probe sitting in the lowest point of the carrier bearing surfaces. This measurement is taken at both sides.

This arm has one tightening knob that simultaneously tightens all three ends of the two red arms, so the whole adjustment and positioning is done with one final turn on that knob. The two bearing surface measurements need different arrangements of the arm, which needs to be first set up on the pinion. The whole thing is carefully moved to the height gauge to be zeroed and then back to the pinion where the relative height of the bearing surface can then be measured.

I removed the magnetic base as it rocked on the pinion and instead used a flat non-magnetic base that is just about heavy enough to hold the gauge unless the arm sticks out too far.

 

 

On the carrier side opposite to the crown wheel, I measured +4 thou. (0.004") from nominal.

 

On the crown wheel side it was +8 thou.

 

 

 

A pinion is matched to its own crown wheel and the pinion manufacturing tolerances result in varying pinion heights. I don't know where this machining variation is but it is measured with the pinion nut tightened to give the correct turning effort at the propshaft flange. This variation is marked on the end; mine is  +9 thou. (0.009") and has be be allowed for when selecting the shims to place between the pinion inner bearing race and the axle casing. In this case the +9 thou is subtracted from the measurment with no shims in place, the result being the shim thickness required.

The other number (4912) is a serial number that should also be found on the matching crown wheel.

 

 

 

So as the shims are still in place because the inner bearing race is still there, the measurements for this pinion should show that it is 9 thou below that nominal height. Well it isn't quite but I have already loosened the big nut and not correctly re-tightened it so I suppose it isn't accurate but at least I know that I can make these two measurements.

Edit: something doesn't feel right about this. I am reading the manual as I do it and it does have one error that I have marked. However the main question is, in the example they give for a +3 thou pinion, why should I see 0.003 below nominal (instruction 88 shown below), when it has been set correctly? If, following the measurement, +0.003" of shim is removed to account for the pinion, why isn't the dti required to read 0.000"?

 

 

 

 

 

Once I have drifted out the old race, I can recover whatever shims are there. The temporary race can be inserted into the casing. Then I will pull the bearing off the pinion and replace it with a new Timken. I don't know the brand of the old inner bearing; I had confused myself by saying it's Koyo but that is the outer bearing. I won't be able to see the marking on the inner until it has been removed.

 

9thou is a lot of variation, especially as the literature says there are three specs x +3thou, 0 and -3thou.  
 

Did you run a contact pattern as I suggested before removing the diff centre to check the mesh was correct before stripping?  If not, it’s worth putting it back together and having a go, just to work out if whatever is going on with the pinion is right.  Once that is ascertained, then you can take the actual pinion depth, swap to Timken bearings (for ease of replacement now and in the future) and set up the pinion to be the same depth afterwards.  The pinion shims are easy to adjust because the outer race is easily removed and refit in the casing with a hammer and drift - it’s nothing like as difficult as removing the inner race from the pinion or the races from the diff centre (swines, the lot of them).  You don’t need to worry about the crush sleeve or seal for doing the contact pattern test, just refit the bearings, flange and tighten the nut to 80 foot lbs or so to get everything nipped.

Once you have ensured the mesh is good (and pinion depth right), measure the depth, swap the bearings with shims unaltered and if the depth is the same, leave it.  Adjust as necessary only if depth changes.  You can run another pattern check if you wish to be thorough.  You can then finish off the pinion with a new crush sleeve, seal and so on before you go onto the diff.

When you do the diff bearing shim checks, start with the diff naked - the ring gear will hit the pinion and limit the endfloat, giving false readings for total shims required.  Once you have that total naked endfloat, you can add the ring gear and then try again using the method in that video link I posted to work out how much of the shim pack should be put on the side with the back face of the ring gear.  Remember the total will be the naked endfloat + 5thou.

 

[Peaklander]

3 hours ago, Snagger said:

9thou is a lot of variation, especially as the literature says there are three specs x +3thou, 0 and -3thou.

I'm not sure to what you are referring. That 9 thou number is the one on the pinion, not a measurement that I have made. Also I haven't seen those specifications for it.

Yes, the diff end-float is to be found without the pinion (or crown wheel as I haven't moved it over yet) so there's no clash. I have just done that, using my dummy carrier bearings and measured 0.100". I doubt that I can get the existing diff carrier bearings off without a fight but then I would recover those shims, or I can get more from Paddock's (£1.32 each) with a bit of a drive. I have 2 off each of 3 / 5 / 10 / 30 thou., so 0.096" in total - not quite enough. I haven't started to split them across the sides yet which might make it more fiddly without a big choice of thicknesses and will do that once I'm ready to measure the backlash.

 

 

I didn't run a contact pattern when you mentioned it as by then the diff centre was already out and I was looking into the cross-pin wear. I did have a go doing it four years ago with these two pics as evidence. I don't think that I took any advice about the results, as it was on 11 March 2020 and we were about to go into Covid lockdown.

 

I have a pinion inner bearing cup that has been ground on the OD to allow it to be used during set-up, so experimenting with pinion shims is straightforward. There was only one pinion shim in there, 0.030".

First thing today I again measured the pinion height and this time got around 4 - 5 thou on both sides. I still find this strange as the pinion is marked 9 thou and so I should measure -0.009" from the height block nominal. If I am correct then my pinion was about 0.013" too high.

I am definitely a little bamboozled by the approach needed on the pinion because surely it all comes down to mesh. Maybe the factory set-up did not include a mesh check and they used a known process of measurement and set-up the shims as a result of that, with no check of 'actual'.

That measurement approach to pinion height setting appears to get things into the right ball park but it is really the contact pattern that is to be used as the final judgement, plus the necessary backlash.

If that is correct then I can do as you say @Snagger - but why would I do it with the existing diff, as the ATB is going surely going to affect it? Can't I just go direct to the ATB?

Edit: putting it back and checking the original mesh with blue. 

Edited February 26 by Peaklander

[Snagger]

The large variation of 9thou that I’m referring to is the number on the pinion, not any of your measurements.  The LR manuals all suggest you should only find a -3, +3 or 0.

You can run a pattern with the existing diff as it has the factory shims and Timken bearings, so removes the diff shim variables and limits the errors to the pinion shims/bearing.

I think the old pattern you have with the yellow goo looks a bit toe-centric on both sides, but mostly the convex side.  From interpolating these charts, it suggests to me the pinion height was a little high , but not bad, and the backlash too tight.  Hard to tell with that kind of paint, though, as most of the teeth don’t have a clear imprint.  I found it a bit tricky when I built that front diff a couple of years ago, trying a few different paints to get a clear pattern as engineering shops are hard to find here.  I ended up using oil paints from an art shop.

[Snagger]

I think you have (or had) a mix between these two conditions with the red lines.

[Hybrid_From_Hell]

6 hours ago, Peaklander said:

The bearing suppliers I rang have each told me that there's no need to match the brand.

 

Yeah was told same with SKF and NTN vs Timken  but if the pinion & Crown wheel have done loads of miles its not a crucial as if all new !

If you leave the shim as is and add a new bearing the pattern will be what the pattern is - its the backlash and run out to work on basically the pinion is what it is / and will be, you are just adjusting the crown wheel in / out to mesh with the pinion and get the right level out run out and back lash.

Don't go bonkers in trying to set backlash as if the CW&P is new, be a bit more generous, the likelyhood is it go go too tight you will have big run out be a bit more generous and you'll get a bigger backlash number but way lower run out and its run out that can make diffs hum !!!!!!!!!!!!

Nige

[Peaklander]

I did run some 'today' checks this afternoon. Maybe I should have painted it a bit thicker but I have at least got more evidence of where it is with the original diff carrier in there.

I am learning all the time about the contact pattern - there's advice all over but thank you Nick for your screenshots.

The pinion cup that has been ground down slipped-in nicely so that will help. With everything back, no crush washer but the pinion tight with a little pre-load (I didn't measure it), I have these patterns...

Drive side (convex)

 

Coast side (concave)

 

Thanks @Hybrid_From_Hell, my head is a little less scrambled now that I know if the mesh is correct then 'it is all good'.

Can you tell me how I change the run-out? I thought that is something that I can't change, save for ensuring that the CW and the mating surface are as flat as can be and the bolts are evenly torqued-down.

 

[Nonimouse]

Did you use Somerset Bearing Supplies? From FB? As suggested...  Best bearing suppliers in the UK

[Hybrid_From_Hell]

Correcting run out on a salisbury is tricky at best - from another angle whats the back lash at 12 o clock 3 6 and 9 in numbers old school thou would be good if you can if mm I'll convert and advise

 

Nige

[FridgeFreezer]

If it's not already been said, can I vote this thread be added to the technical archive?

[Snagger]

The new pattern looks pretty good, Tim.  A bit hard to read on the convex side, but a few of the teeth suggest a mid to toe patch which is good.  It helps to wedge something between the edge of the ring gear and the casing to make plenty of resistance so the teeth really push hard against each other.

There isn’t really any adjustment for runout.  It’s a matter of ensuring the faces are spotless and using a wet stone to smooth off any burrs around the bolt holes.  I had a very small amount of runout on my ATB and 4.1 (possibly KAM) ring gear, but I was able to use the tiny variations in the Ashcroft spacer ring to cancel the runout - the variations coincided very well if I set the ring one way around (having measured it between and at every bolt hole with verniers after using the whetstone on that too.  Since you aren’t using a spacer, the only option you have is to fit the ring gear, check for run out and then rotate it on the diff flange one bolt hole at a time until you are satisfied.  Slow, but may yield a result.

[Peaklander]

I am trying to remain focussed but having started the engine top-end check and overhaul some weeks ago and forcing myself to get to the bottom of the drivetrain clunk 'once and for all', I seem to be doing little else. Driving it to Ripon on Sunday seems a tall order, as everything I do takes longer than I expect.

So this morning I wanted to clear my head and so Mrs P and I went for our weekly swim (normally Friday) and I emptied my head of pinion thoughts and filled it with counting strokes. 2,000m later I am back and want to get this over!

 

Here's a  summary of where I am and please excuse the ramblings, it helps to get my thoughts onto 'paper'. So far nothing has changed and those yellow and blue contact patterns are with the original set-up. All I have done is plenty of pinion height measurements, I removed the inner pinion bearing cup, measured the shim in there for interest (30 thou) and replaced the shim together with the ground-down Timken cup so that I can adjust the shim thickness easily.

So the blue isn't a new pattern Nick, it is just the most recent.

In the post this morning (it never seems to stop), I have red Loctite 263 for the crown-wheel bolts and an inner pinion bearing (Koyo). Both of these were delivered overnight in 1st class post so thanks Royal Mail. I got the Koyo in a rush yesterday when it appeared that I was going to preserve the current setup, either by not changing the pinion bearing or by staying with the existing brand (Koyo). Now I think I am going to change things so much by fitting the crown wheel to the ATB, that I can consider moving back to Timken and relying on wear patterns to get me something like OK.

The pinion height is defined by the shim and is way off where I believe it should be, as I am measuring  4-5 thou above nominal and the pinion, marked +9, needs to be at 9 thou below. That's what I understand from the workshop manual. However, as you can see, the existing contact patterns (nothing at all has yet been changed) are OKish.

I now need to decide, change the pinion bearing or not. I have a new Timken and now also a Koyo.

19 hours ago, Hybrid_From_Hell said:

If you leave the shim as is and add a new bearing, the pattern will be what the pattern is - it's the backlash and run-out to work on. Basically the pinion is what it is / and will be, you are just adjusting the crown wheel in / out to mesh with the pinion and get the right level of run-out and backlash.

Don't go bonkers trying to set backlash as if the CW&P is new, be a bit more generous, the likelihood is if you go too tight you will have big run out. Be a bit more generous and you'll get a bigger backlash number but way lower run-out and its run-out that can make diffs hum !!!!!!!!!!!!

 

With this in mind I will pull the Koyo off the pinion using the Vevor clamshell puller and then put the shaft in the freezer. Later I will put the Timken into the Aga (there's a choice of ovens at 100C or 220C) and then I think the new bearing will go on without too much bother. I do have a press but the jack is cr@p and it doesn't always want to exert enough pressure.

Then, when it has all got back to ambient temperature, I will mount the pinion and measure the height with the 30 thou shim still in place, just to see what has changed.

I only have 90 thou of carrier shims and this won't be enough, so I will cut the existing bearings off the Salisbury and hope I can recover the shims from it. There's no way that the Vevor puller can get under those side bearings - there isn't any space but it might pull the outer race off - a less sparky method maybe. If there are enough shims then I can load-up the ATB on the crownwheel side.

The crown-wheel has to be removed, and the mating surfaces of it and the new ATB checked and then I can fit it and check the run-out in various places, as Nige says. I like the method that you suggest Nick of rotating it one segment at a time to try to overcome excessive run-out.

Then I will fit the ATB and see where we are, using the dummy bearings on that, again so that I can easily adjust things. Well as easy as it will be with the spreader - I hope that bit works.

Then I will check the backlash and move the shims from the Crown wheel side to the other. Finally I will get the blue out and report back.

Edited February 27 by Peaklander
typos

[Snagger]

You can give that runout process a go, but it’ll only make a difference if the ring gear is out of true, which seems unlikely.  I think it’s worth checking the diff ring flange for runout with the dummy bearings fitted, but that should be pretty good as it hasn’t had an opportunity to be buckled.  Still, it only costs time to try.

I think the pattern is pretty close, but the contact does seem to be a bit toeward on the concave side.  The convex side looks spot on.  The blue pattern is much easier to read than the yellow, and it’s result do seem to fit the top line of the guide I posted with the added red lines on the left.  In the remarks of that top line, it does say the contact may be towards from centre on the coast side, as long as it is not near the ridge of the teeth, and you have just that, so I think you can leave the pinion shims are they are.

[Peaklander]

Thanks @Snagger - I will be spending lots of time very soon looking at more blue marks. I will apply it more carefully too.

This afternoon I had a great time with the Vevor puller and thanks to @hurbie who mentioned it in my vehicle thread and to Mick at Britannica Restorations who produced a short 'first use' video here.

I started on the pinion bearing. You need to use the appropriate size clams, there are three to choose from.

It's interesting how it works, not able to get under the inner race, it lifts the cage but the bearing is clamped together with the two knurled discs that are wound down the main shaft. Then a retaining sleeve slides over and is secured with a large screw against the clam.

 

 

 

 

Then all that's needed is for the centre 'pin' to be wound down to push against the top of the pinion. It was time for my China impact through a 19mm socket and.... nothing happened. So then as my only and last resort, I dusted-down my Aldi middle aisle Workzone air impact gun and cranked the air pressure up to 9Bar and to my slight surprise and relief, the shaft started to turn and the assembly moved up!

 

Bingo! One Koyo pinion bearing is now free.

 

I was so confident and now it was clear that the bearings will move with this method, I moved to the diff carrier after all. This has shims that I wanted to save. I decided to use the black clamshell pair as that seemed to be a little larger and kept away from the shims underneath the race.

 

 

These two bearings were not quite as tight as that on the pinion and they were soon off and the shims recovered.

 

That is 105 thou (0.105") available which happens to be the ATB (end-float + preload) that I measured the other day, or not far off. So with the 3 / 5 / 10 / 30 pairs that I bought sale or return from Paddocks, I am going to be OK.

Then I released the crown wheel. These bolts were tight and needed a breaker bar, sequentially released just a little at a time. For a bit of light therapy I then cleaned off the red Loctite, which needs plenty of heat to soften it and then a stiff wire brush.

 

 

Then it was onto the ATB, just for a quick look but I need to check how to fasten them in, when I might need to remove them again in order to rotate the ring.

Do I torque-up fully without thread lock, check run-out and if OK then remove one bolt in turn, apply the red stuff and re-torque that?

Or do I have to do the full sequence which might affect run-out if it is fully released?

That will be the morning's job.

 

 

Edited February 27 by Peaklander

[hurbie]

oeps ... make sure u use the outer race on the bearing when pulling , this prevents the cage from distortion !!

[fmmv]

I do wonder if the case spreads and relaxes much in use and alters the settings. The standard cover won't give it much stiffening.

[Hybrid_From_Hell]

some casings at 30+ years do not like being stretched - not enough and it won' come out too much and thats a shagged axle tube, and rust and old age do not help

Its an old trick but before you tighten the carrier bolts up fully give them a big clout with a hammer, sometimes you can hear a click 0 thats the race just settling which can defo help

 

Whats the backlash  in the 4 quadrants ?

nige