V2 Diff rebuild + Gear Ratio Swap (4.10)

[maupineda]

This is how I measured the pinion to confirm if I needed to change the depth shim, and by how much.

First I setup the pinion on a precision block, and align the micrometer so that I could have the exact same setup for both pinions, I started with the old one


I also rotated the pinion to verify the runout, it is incredible that the micrometer did not move at all as I spun the pinion.

Here is the original one



and here the new, same, no runout as I spun the pinion, they are cut exactly the same.



then it was all about getting the pinion into the carrier and adjust the preload





Edit: spent time in ISTA last night and reviewed the specs for e36 chassis, and found that the pinion preload should be 2.5nm or 250ncm. I will redo this as I fear I may have screwed up the process.

I also found the carrier preload values which are bearing brand dependent. Will add here.

K = kompact or small
M = medium
G = grob or large

[maupineda]

To close the carrier I use the same method I used to open it, get to as close as I could, the. Zip tie it and use a hammer to tap the tool and lock it in place.


the final torque was a bit lower than what I expected, but it is very similar to how it was before, this makes me think the unit was still ok, also the clutches looked fine!


I was able to close the unit to where it was. I also shimmed the side gear with the play, and it is much reduced now that the unit is closed. I used 2 0.1mm shims with RC40 harness for durability.

Edit: took a photo of the carrier closed to the original position with it set screw to lock it.

[maupineda]

Continuing with the rebuild I made progress this weekend. Basically finished the carrier assembly as well as replacing the bearings and races, and setup the pinion preload.

I also installed the 4.1 pinion and made sure I did not have to change the pinion depth shim, RD say is very unlikely that the pinion shim needs changing, and was pleasantly surprised that it is true, I guess it has to do with BMW tolerances on these parts.

so before closing the carrier I wanted to see if I could make the ring close to the original position, to my surprise it goes to the same place with and without clutches







moving on, I removed the old bearings, for this I got a bearing puller and it worked great, it really removes the bearings in a breeze.



I did not take a photo of the pump side bearing but you must remove that with the pump off the carrier as the tool does not fit otherwise.

[sapote]

I have been playing with the shimming for the short output shaft side gear, I started with a 0.1mm shim, and I reduced the backlash some, but by feel it did not seem to do much. Then I tried 4 shims and with 0.4, the locating shafts don’t go in

The short output flange backlash is not just reduced by its own shims, but also depends on the plays of the other side-gear, because the 4 spiders still free to rotate due to their backlash with the longer flange side-gear. Imagine if the longer flange side-gear is removed, then the short flange side-gear would have infinite backlash.

[maupineda]

I have been playing with the shimming for the short output shaft side gear, I started with a 0.1mm shim, and I reduced the backlash some, but by feel it did not seem to do much. Then I tried 4 shims and with 0.4, the locating shafts don’t go in, with 3 you can get them in if forced, so the happy place is 0.2mm. You can feel thing a bit tighter, there is still lash, but all spins freely.

i will post here the shims I used, they were a bit expensive, but they are hardened to 40 RC which is needed due to the inherit bearing load they will see under use.

[maupineda]

The missing clutch and steels came in today. I was surprised to see how nicely they packed them

[maupineda]

A fellow member mrgizmo04 made a very educative video on it.

https://www.youtube.com/watch?v=O0jaRcg7T9o&pp=ygUNIzFvdmVybWUyMHJ1bg%3D%3D

Enjoy

[sapote]

What is the difference between the V1 and V2 M-Track diff?

V1: right output flange is shorter
v2: right flanges is longer . Has more friction clutch plates than v1, handles higher load.
they have total different LSD carriers.

[ac427]

What is the difference between the V1 and V2 M-Track diff?

[sapote]

i Also measure the axial displacement of the bottom side gear and it was 8thou, or 0.2mm. I will add. 0.1mm shim.

8thou doesn't sound like a lot for side-gear to spiders clearance, but I wonder if you can measure the rotational plays on the side-gear. To cause the infamous M clunk, I would expect to see more than 30* free rotation.

[maupineda]

I reassembled the carrier unit without the clutch pack to measure the space between the pump piston and the base plate that supports the clutch stack.

As I had suspected, when removed, the pump piston extends to what I assume is it most extended position. As you can see, there is only ~13mm in between, which means that any clutch stack of 13+ mm will have the piston in its operating range. Without drawing is impossible to know what is the full travel, and although I could press it on the hydro press until it bottoms out and measure, I don’t want to break it.

This explains why there were no shims, possibly not needed, and although I saw in one of the recent RD videos, at least my car had no shims between the piston and the first steel plate.

with my OE clutch pack the pump’s piston would be pressed in around 1.4mm.

RD sold me some extra steel plates and 1 clutch, so my stack will be 0.4mm thicker (14.8 v 14.4) which should be ok assuming the pump has enough travel.



i Also measure the axial displacement of the bottom side gear and it was 8thou, or 0.2mm. I will add. 0.1mm shim.

[sapote]

as mentioned above, the clutch pack gear (what I called spacer gear before) has an effective height of 16mm! More than that and you will likely have a lot of preload and risk the first clutch plate to crack by the piston pump, or wont be able to close the unit at all.
....
OE stack is 14.42mm

The clutch stack cannot be higher than the splined spacer as the clutch above the spacer no longer function as its teeth don't engage the splined spacer. It is important to install the pump exactly to the original stock position in order to have the same preload friction on the output flange bearings.
Another reason to keep the pump at the factory position: lower or higher position will cause the R side-gear to have less or more backlash from the stock.
This means if you don't use the spring washer on the new RD clutch stack then you have to select the right flat shim to have a proper break away torque, through trial-error. Selecting the proper shim is a long process with many thin shims, or many different shim thickness. I think the factory has a magic total clutches + shim thickness and they just need to sort this out without the need to trial-error for the breakaway torque. This magic number is a little higher than 14.42mm based on your measurement of the old worn stack.

Those rebuilt diffs using RD parts with the spring washer will not provide the same locking characteristic as the OEM diff, because the spring absorbs part of the force from the pump instead of transfer 100% of the pump to press on the clutch stack.

[maupineda]

Took a photo of one of the clutches. All look like so. They seem to be fine.

[maupineda]

I think RD's deal here is the steel/clutch stack will be thicker so it acts like a mechanical clutch diff. The reason being that the viscous pump wears out and no longer engages to the press the clutch stack together. Try reaching out to RD and see if that's normal?

I am not claiming this to be an issue with the kit, just an observation. the RD kit is designed to use the spring washers, and as such the stack height of the clutches and plates alone is much different to the OE ones, but since there is only 6 and 7 respectively they stack height is less, so you must use those spring washers, or get more plates and clutches to match the OE setup.

My plan is to go with an OE based layout, that is, clutches and steels, with no spring washer. so I need to stay within reason when assembling my clutch pack.

If you check post #42, you will see that when the washer provided is fully compressed, the clutch stack height calculates to 14.67mm, and about 17.3mm when uncompressed.
OE stack is 14.42mm

There is no information on this, but the pump's piston may have a total travel of 1.5 - 2mm.

If you had a failed viscous pump, even with those spring washers, you wont have a lot more clutch pack material to generate lock VS OE. Without special equipment is tough to say, but I bet the pumps piston takes less force to be pushed in than the force needed to compress the spring washer, if that is the case, the piston will be fully pushed in, and the spring washer will be compressed a bit (but not fully). The issue I see with the RD's approach is that some (I assume most) of the force that the pump generates is used to compress the washer and not to bind the discs & clutches. that is why I wont use any of those spring washers.

I am trying to source 1 more clutch and 2 more steels. if I can do that, this will be my array

Thickness (mm)	Qty	Height (mm)
1.1	6	6.6
1.03	6	6.18
2.02	1	2.02
	Total	14.8

If not, I plan to reuse 1 clutch, and 2 steels from the OE setup. This will leave me at 15.2mm.

We'll see.

[bigjae46]

If I were to use all RD items (same number of steels and clutches), this would give a theoretical value of 18mm!!! no way we can fit all of that in the unit. as mentioned above, the clutch pack gear (what I called spacer gear before) has an effective height of 16mm! More than that and you will likely have a lot of preload and risk the first clutch plate to crack by the piston pump, or wont be able to close the unit at all.

I think RD's deal here is the steel/clutch stack will be thicker so it acts like a mechanical clutch diff. The reason being that the viscous pump wears out and no longer engages to the press the clutch stack together. Try reaching out to RD and see if that's normal?