109 front Final drive unit, ATB and 4.1:1

I have completed the refurbishment of the second hand final drive unit and 4.1:1 gear set with the Ashcroft ATB.  The axle will get Ashcroft 24-23 spline shafts to make it all fit later.

Rover 4.1 ring and pinion

Used Discovery final drive

The used assembly was sourced locally and was from a late Discovery with the Rotoflex coupling. The centralising spigot in the drive flange wasn’t coming out without a major struggle, as is common, and ultimately the three-holed flange was cut up with a grinder to get access to the securing bolt into the pinion as it was destined for scrap anyway.

Pinion, bearing caps and adjusters cleaned

After that, the main bearing caps and main housing were stamped to ensure they were reassembled in the correct orientation and the whole lot was stripped down. Easier said than done when dealing with highly torqued bolts with no bench, let alone vice, and no big wrench to engage on the drive flange. But I got there with a degree of pig headedness, and got it all stripped and thoroughly cleaned with petrol.

Setting up was the big challenge. The diff main bearing preload and lateral setting comes later and is very simple with the slotted adjustment nuts. It’s the pinion depth setting that is difficult. I had two major issues here: a lack of special tools for the job, and no standard depth markings on the pinion.

4.1:1 pinion markings

The only marks on the pinion are 33k, which I assume to be the mileage the gear set did before removal from their previous unit, and 76.25, which is not a typical depth measured from any point on Rover units. Normally, the depth is measured between the face of the pinion and the lowest point of the main bearing seats, so would result in a number in the low 40s or even less. I did take a measurement of the original pinion in the housing before it was removed using a straight edge across the main bearing cap faces with a digital calliper, which is not ideal and quite fiddly to get a good reading, and that came out as a little over 78mm, suggesting the engraved markings were for setting in this manner but that the gears are not genuine Rover (as was already suspected). It’s not certain, but Dave Ashcroft suspects they were made by KAM Diffs, which I would be very happy about.

The pinion depth is set by shims under the forward (large) bearing outer race. Changing those shims would be a pig with the new race fitted, so after checking the condition of the old race was adequate for this task, I ground its outer circumference down so it could be installed in the housing by hand to allow rapid trial and error setting of the shims.

Starting with the original shim from the casing, which KAM Diffs said is correct in 95% of gear set changes, the distance measured after a lot of fiddling about to be accurate was within 0.02mm of that engraved number. That was from a different main housing, which could have the bearing caps cut at a slightly different height than the housing I have, given production tolerances, but it should be good enough to start with to run a mesh pattern later.

The 4.1 gear set pinion head radius is about 2.5mm less than the original 3.54 pinion, and the ring gear about 2mm thinner, so this dictated the use of the 5mm spacer ring I bought from Ashcroft with the ATB (comes with new longer bolts). I checked the ring gear, spacer and diff flange for burrs and only the spacer had any. These were dressed with a fine file. They were assembled and fitted to the housing and the backlash adjusted to a rough setting (I didn’t have a dial indicator yet) and a contact pattern was run in several places around the ring gear. They were very inconsistent.

Spacer ring measurements

Diff ring flange runout checked

Once the dial indicator and mount were acquired, I checked the ring gear runout while turning the diff, and it wasn’t great. The diff, ring and spacer were stripped and the bare diff checked again for runout at each bolt hole while the spacer was measured for thickness at matching intervals. The diff flange runout is 0.02mm, about 150 degrees of rotation between highest and lowest spots, the variation in spacer thickness 0.03mm. Serendipitously, when the high spot of the diff flange was aligned with the thinnest area of the spacer, the thickest part of the spacer is over the flange low spot. The combined variances are not perfectly cancelling around the whole circumference, but most of it falls within 0.01mm, so should be a very good basis for the ring gear.

With that rectified and the ten bolts thread locked and torqued to the correct setting (45 foot lbs), the assembly was refitted and a new mesh pattern run. I found it a little hard to read but it wasn’t far out. The contact was pretty central between root (valley) and face (tip) of the ring teeth, but seemed to be toe centric (inboard end of the ring teeth). The ideal is to have an oval contact patch on both sides of the teeth (drive side is the convex side, coast side is concave), centred roughly 1/3 of the way from toe to heel (outside circumference of the ring). I was getting a different combination.

Land Rover manual contact pattern guide (note contradicting pictures and condition statements for D and E)

Delving through the internet for pattern diagrams and YouTube videos presented a fair bit of conflicting information. The general consensus is that the height of the diff pinion dictates whether the contact is towards the root or face of the ring teeth, and the amount of backlash (play set between the teeth by moving the diff left or right) controls the toe-heel part of the pattern, but some guides contradict that. Adding to the confusion are Yukon (decent quality North American aftermarket diff and gear set manufacturer) saying to concentrate on the coast side pattern when setting up used gears, LR manuals only showing the pattern on the drive side and ignoring the coast side, and the LR manuals having an error in the condition statement of the backlash adjustment diagrams, talking about heel contact where the diagram shows toe, and vice versa, but the corrective action statement matching the diagram conditions. Very confusing and misleading.

The patterns I was getting were seemingly acceptable but not ideal. I decided to try replacing the pinion shims with combinations from the Ashcroft kit to see how it could be improved, but they resulted in the contact moving too far towards root or face, strongly suggesting the original shim was indeed correct, as per KAM’s statement, and that the depth measurement I had being almost identical to the engrave number was what was needed even with the housing change, so I went back to the original shim and installed the new outer race (it had been in the freezer for quite some time to shrink it in an effort to make fitting a bit easier).

With the pinion depth set, the preload shims were selected through trial and error (there is a lot of that in setting up final drive units) until a preload resistance of between 20 and 25 INCH pounds (not to be confused with the more oft used foot pounds) was set. Use 15 inch pounds for old bearings.  The pinion nut has to be done up to the full 95 FOOT pounds when checking the preload, which was challenging with only a 1/2” drive T-bar to wedge between the socket on the pinion and an M10 bolt through the drive flange bolt holes so use to stop the pinion turning with the torque wrench.

It took a few attempts, but the right combination of shims from the Ashcroft kit gave me a preload of 21”lb, a little light but not bad. With that done, another contact pattern was run and got better results. The coast side shows the marking paste (I used oil paint) squeezed out at the toe, but there was visible contact for the bulk of the tooth on that side. The drive side was much closer to the diagrams of the ideal pattern. This time, I marked the main bearing adjuster nuts to make setting them up for the last time very quick.

One last strip down was carried out to clean out all the threads for the bearing caps, fit the pinion seal (with a smear of RTV sealant around its outer edge, after using a flap sander on a Dremmel to clear the rust off the housing aperture) and repaint. The pinion and it’s bearings were given copious greasing before refitting, not so much for lubrication as much as for rust prevention, the drive flange was similarly greased for all of its seal land and the splines, it was refitted and the nut torqued back down again to 95’lb. The preload was rechecked within spec.

Drive side pattern

Coast side pattern

Then the main bearings were packed with grease, the bearing caps, seats and adjusters sprayed in WD40, all assembled and the cap bolts torqued down. The backlash was roughly set and then the alignment marks from the previous contact pattern testing were used to set the main bearing adjusters, but bizarrely this gave a smaller backlash than previously. A very small adjustment was carried out and I got a decent pattern on the smallest end of the backlash range (0.10-0.17mm). No matter; the pattern is good, with contact on both sides mid way between root and face and the drive side having contact from toe to between 15-20% from the heel end and central on the coast side with about 10% unmarked on both toe and heel.

Complete final drive with ATB and 4.1:1 ratio

The adjuster locking roll pins were hammered in, the innards and drive flange exterior sprayed heavily in WD40 and the whole lot wrapped thickly in cling film to protect from damp and dust while it is stored.

Be Sociable, Share!

By continuing to use the site, you agree to the use of cookies. more information

The cookie settings on this website are set to "allow cookies" to give you the best browsing experience possible. If you continue to use this website without changing your cookie settings or you click "Accept" below then you are consenting to this.

Close