Final Drive Shimming

[rxcrider]

Does anyone have a link to any discussions on final drive shimming? I've found lots of final drive threads, but none that pertain to shimming, measurement techiniques and the potential for incorrect shimming from the factory.

I pulled apart my spare 1996 R1100RSL final drive yesterday. It is relatively low mileage at just over 20k, but it was stifff to turn relative to the one in my 1995 R1100RS. It also had a slight leak at the output seal, but turned smoothly and the bearings seem to run fine with the cover removed. My first stab at measuring the preload came out to 0.20 - 0.25mm of preload which is a bit much in comparison to 0.05 - .010mm called out in the manual. I intend to repeat the measurement using a different technique that I found linked from ADVRider.

http://advrider.com/forums/showpost.php?p=11550207&postcount=66
http://www.advrider.com/forums/showpost.php?p=16598689&postcount=18

There was also discussion of using soldering wire to compress in place of the shim(s) to get an accurate measurement of the gap between the bearing and the housing from the deformed solder wire.

The links took me to a video from CharlieVT on the bmwlt forums:
http://www.bmwlt.com/forums/showthread.php?p=519256#post519256

I was hoping to find some comments here letting me know whether or not I was headed down the right path, but my search results didn't get me anywhere. Any thoughts or links?

 

[David13]

BMWLT was having a somewhat spirited discussion just now about that issue, or the whole final drive issue, with the shimming for the preload. I think you have a preload unit there.
Charlie VT aka Curtis was involved in it. He may well know who knows what about that.
I have the 'evo' unit, which I am watching like a hawk. But for the shimming unit I would have him do it, or Anton in VA.
They seem to have perfected a procedure.
dc

 

[ragtoplvr]

I have used the solder method, after a failure at 36K. I am at 77K now (and counting) I think mine had 2 defects. The outside of the hub where the big ball bearing is mounted is oversize. This reduces radial clearance in the bearing, and makes the balls slide when side load is applied. I did sand the area some, I may have removed .0001 not much. My preload was correct by solder deformation. I have not had a further failure.

Rod

 

[roborider]

You can also use "mold a gauge" from Cycle works.

 

[rxcrider]

thanks for the replies. I've also found some more links / discussions

http://www.bmwra.org/otl/finaldrive/

http://www.bmwlt.com/forums/showthread.php?t=59416

http://bmwsporttouring.com/ubbthreads/ubbthreads.php?ubb=showflat&Number=776665&page=1

 

[m_stock10506]

LT Final Drive video is good source

The LT video from CharlieVT is very good. It is quite long but it is a very thorough look at a final drive bearing R&R and shimming.

I just used that video along with Paul G's write-up to do the big bearing replacement without pulling the entire final drive off of the bike. My crownwheel gear appears to also have been slightly oversize; the video and instructions indicate that the bearing comes off and on with a modest amount of heating and cooling. I couldn't do that with mine and ended up trashing one replacement big bearing ($85). I ended up having my local dealer put the next bearing on the crownwheel for me - for 1/2 hour of shop time - fair enough.

When I put the new oil seal on the crownwheel and cover, it sank about a millimeter below "flush" with the cover. Once in, there did't seem a way to pull it up so I left it. Don't see any leak - should this be a problem?

 

[rxcrider]

After reading as much as I could, I came up with my own little spin on shimming. It was a bit tedious, but I believe the results are reliable. I stacked four washers together that were ~9.9 x 38.5 x 3.4mm. I bolted them together and then wrapped them with electrical tap until they were 40.55 - 40.60mm OD. I made four of these stacks and bolted them in place around the big bearing. The tedious part was rotating the stacks and adding / removing small pieces of tape until I got measurements from the outer race of the bearing to the mounting face for the cover that were all withing 0.03mm of each other.

Once I got to that point. I took four measurements for each of the dimensions called out in the service manual and averaged them. I then did the calcualtion and came out to using the same shimming that was already in from the factory for a 0.06mm preload.

I buttoned it up, added a new seal and found that the hub turns easier than it used to, on par with the one in my 1995 bike with over 80k on the clock. I'm guessing that the drag I was feeling was coming from the leaking seal.

Now I have another question. The crown wheel hub is hollow on the 1996 FD that I just put back together, as you can see above. The one on My 1995 has a post in the center like the one shown on Anton's site:

Is one any better than the other. I'm guessing that the 1995 unit weighs more, but other than that... Anyone have an answer or opinion on the subject?

 

[gsaddict]

I have used the solder method, after a failure at 36K. I am at 77K now (and counting) I think mine had 2 defects. The outside of the hub where the big ball bearing is mounted is oversize. This reduces radial clearance in the bearing, and makes the balls slide when side load is applied. I did sand the area some, I may have removed .0001 not much. My preload was correct by solder deformation. I have not had a further failure.

Rod

What diameter of solder did you use for your test?