Need help with setting wheel endplay with Timken bearings

[2017FLHTK]

Hello,
I just got a set of late 90's sporster wheels that I'm trying to use on a 2001 XL. Obviously 1999 and newer wheels use Timken bearings, while the 2000 and up units use sealed bearings. Since it's the same axle size, everything should bolt right up...

When I took the old bearings out for a quick inspection both the bearings and the races looked to be in great shape. However, I opted to install new bearings since I didn't know the mileage of the ones installed in the wheels. I found a set of shims and a "top hat", which appeared to be correctly installed in between the center axle hub spacer and one of the bearings to aid in setting end play:


I wanted to verify that the endplay was correctly set so I mounted my axle in a vice, put on a spacer to lift the wheel up / away from the vice so it could spin freely, and placed a brand new Timken bearing (Harley part number 9052) on to the shaft:


I then placed my wheel on the axle, inserted the center axle sleeve spacer (part 16 on the diagram below), installed the shims and top hat (with the lip facing up towards the bearing), slid the Timken bearing in place, and then installed some random spacers to allow me to tighten down the axle nut. have a service manual for a '95 XL which specifies that the axle nut should be torqued down to 60-65 ft-lbs.


However, when I tighten down the axle nut I get ZERO end play. My service manual designates an end play of 0.002-0.006 inches on the Timken bearings for this era of wheels. The wheel will still turn, but it's obviously dragging or binding and only spins for a few rotations before stopping. I decided to remove the shims and top hat to see what kind of end play I got, but even with those out of the equation I encounter the same issues once the axle nut is tightned (zero end play and noticeable drag when trying to spin the bearings).

My next thought was that perhaps the center hub spacer was too long. A parts catalog shows that the spacer (Harley part number 43608-82) should have a nominal length of 4.308". I have a small mini-lathe, so I decided to take a little length of the spacer and see if that affected the endplay. I turned the spacer down to 4.25", re-installed everything except the shims and top hat and encountered the same issue. With the axle nut loose I can move the wheel up and down but as soon as I put any torque on the nut (i.e. an estimated 10-15 ft-lbs) I can NOT raise the wheel and have ZERO endplay. See video below:

I then decided to completely take center the axle sleeve spacer out of the hub and encountered the same issue: as soon as the axle nut is tightened the whole assembly becomes a solid piece with no endplay whatsoever.

Any ideas what's happening here? What do I need to do in order to get at least SOME endplay when the axle nut is torqued down (which I can then adjust by adding or subtracting shims).

[fbn ent]

The space between the bearings has to be longer meaning more shims or a longer spacer to get more play. Note that there are stepped washers that contact the inner race on the bearings on both sides. A magnetic based dial indicator is your friend when doing this. BTW, I prefer the old tapered to the new sealed units. I shoot for .004".

[2017FLHTK]

The space between the bearings has to be longer meaning more shims or a longer spacer to get more play.

Makes sense.  Looks like I went the absolute wrong way with it by taking length off the center hub spacer.  However, I had the lack of endplay issues with the top hat spacer and shims installed.  You're thinking it needed more shims installed in the stack?

Note that there are stepped washers that contact the inner race on the bearings on both sides.

I was tracking that the stepped washer was ONLY on the side that had shims installed.  In the parts diagram above, Part 14 is the shim(s), Part 17 is the "spacer washer" (aka stepped washer).  Am I missing something?

A magnetic based dial indicator is your friend when doing this. BTW, I prefer the old tapered to the new sealed units. I shoot for .004".

I have a dial depth indicator, but I need to get to a point where there's at least a little movement before I try and set how much movement there actually is.

[kd]

I had this typed out but got called away and didn't send it.  FBN answered your question but I am posting it anyway.  I hate to waste the time one finger typing. 

If I understand what your describing correctly, you are going in the wrong direction. The spacer and shims are intended to add clearance when installed.  Torqueing the axle nut squeezes the Timkin bearings (cones) onto the bearing races (cups) and causes preloaded friction when the components all touch.  Removing shims etc. reduces clearance between the bearing cones and subsequently free play goes away.  Add shims and the clearance will increase causing the wheel to rotate more freely. The purpose of thin shims is to allow you to control the clearance (required to compensate for growth from heat) to achieve the factory recommended bearing free play.  (the bearing cups or races are stationary when properly seated in the wheel hub)  Now that you have shortened the main spacer you will need more shims to get your clearance back.  You should be able to comfortably make up for the minimal amount you machined off the OEM spacer tube by using shims.

[2017FLHTK]

I went back and double checked that the center spacer is proud and protrudes (slightly) through the bottom of the hub so that the shims and top hat washer have something to press against / raise the bearing off the race:


The original length of the center spacer is a nominal 4.308".  After shortening my center spacer, I now measure a length of 4.253" (4.308"-4.253" shows that I removed around 0.055" inches of length).

I couldn't get my hands on another stock length center spacer today, but I raided the local dealership for every shim that they had.  Stacked together, I get around 0.217" of height:


0.217" of shims plus the length of my cut-down center spacer (4.253") comes out to be around 4.470".  If you subtract the original length of the factory spacer, 4.308", that means that I now have around 0.162 of additional length in the overall height, as compared to a factory center spacer with no shims whatsoever. I'd like to think that that's more than enough to start seeing some end play.  In any event, the shim stack sits up high enough that the roller bearings on the Timken bearing should be lifted up and away from the race when dropped in place:


However, I'm still getting the same symptoms.  When I tighten down the axle nut anything past finger loose I loose any endplay. Where do I go from here?

[2017FLHTK]

One last question.  Is there a step or ledge that the races seat against inside the bottom of the wheel hub?   I can see that my race is NOT bottomed out flush against the bottom of the hub, and see a narrow gap between the bottom of the race and the hub:


I'd assume that's normal, as other wise it'd difficult if not impossible to drive the race out without access to the bottom of it, but I wanted to validate that this is how they normally look.

[kd]

The bearing cup (race) has a radius on the bottom outside corner where it seats into the bore.  This allows it to be driven in until it seats at the bottom of the machined bore.  It insures that (due to the machined bore surface) it is seated square and won't be pushed deeper (by the taper of the Timkin) into the machined hub bore when the wheel encounters shock from bumps on the road. If that occurs you will have a loose fit and have to readjust it.  It doesn't need a gap under the cup to remove it.  There is enough of an edge to allow a puller or even a sharp squared off tip (on a long drift punch) to catch it and allow it get manually tapped out.  I have removed many of them by carefully tapping around the radius with a long drift punch.

You can check the one you have by trying to carefully reset it deeper into the machined bore.  By carefully and I mean carefully, work your way around the circumference of the bearing cup going from one side to the other and see if it moves.  A seated cup will have a distinct hard sound when it is seated.  Of course there is always a possibility someone removed the previous race with a drift punch slipping and damaged the seat so it doesn't seat properly or it has debris under it.  In the case of self inflicted damage there kmay be a mark or marks around the circumference somewhere indicating that.

[2017FLHTK]

You can check the one you have by trying to carefully reset it deeper into the machined bore.  By carefully and I mean carefully, work your way around the circumference of the bearing cup going from one side to the other and see if it moves.  A seated cup will have a distinct hard sound when it is seated. 

I used a punch to tap around the top edge of the race.  It gave a solid "thunk", and didn't seem to move at all.  I then found a socket which JUUUUST barely fit inside the bore (a 1-1/8 deep well) and give it a couple of solid hits with a hammer.  I didn't detect any movement on the race (I smacked the race on both sides of the hub), so I'm going to have to assume that they're fully seated and is not causing the issue.

[capn]

That gap under the race is for the tool to sit when driving the race out. Don't bang on that race with a punch. Top hat spacer only needed on side where shims are.

[fbn ent]

That gap under the race is for the tool to sit when driving the race out. Don't bang on that race with a punch. Top hat spacer only needed on side where shims are.

You are correct. My bad!