Fixed Various Issues on the Rovaál CLX50 Rear Wheel

I received a rear wheel from a Rovaál CLX50 from a customer.

Even though it had been overhauled at another shop,
the rotation became sluggish and developed abnormal noise.
It's been used only in fair weather with less than 2000 km of distance.
Given the considerable wear, I suspect the 2000 km
is the distance since the overhaul, not from when it was new.
However, that's not really the main issue—
they'd done so many inexplicable things that
after finishing various repairs, I called the customer and
asked: "I'm not accusing you of anything, but
you didn't do this yourself, right?"
What I learned afterward was that this wheel had been left at that shop three times,
and each time it came back, something or other was wrong with it.

Regarding hub rotation, when holding the quick-release end and turning it with my finger,
the rotation wasn't cloudy so much as it had
a catching sensation like a safe dial, or
a grinding, gritty feel to it.


This Rovaál hub is a DT-made star ratchet style hub,
and the bearing on the right side of the hub body requires a special tool to replace.
That bearing had indeed been swapped out.
Since few individuals purchase that tool,
it's certain that this was done at the shop.


I removed the left quick-release end.
Oil and metal particles had mixed to form a paste-like grime
around the bearing seal.


This is the inner surface of the left quick-release end.
There are wear marks on areas other than the step that presses just the inner race,
but I'll post the cleaned image later.


I removed the right quick-release end.
The orange color of the freewheel body's bearing and the flush surface
looks unusually shiny because there's oil on it.


This also had paste-like grime stuck to it.


On the star ratchet's ratchet section,
there are no traces of the special grease being applied.
If properly applied during overhaul,
it's impossible for it to completely vanish after just 2000 km in fair weather use.


Even with the freewheel body removed—in "front hub mode"—
the rotation still grinds.
So the cause remains on this side.


Using the special tool to remove the star ratchet spline,
I loosened the spline.


Before the spline comes out, a ring with a U-shaped groove
—which forms a labyrinth seal between the freewheel body—
comes out first. The Japanese distributor called it
the "water and dust seal ring."


I removed the spline.


I pressed out the right side hub body bearing along with the hub shaft.
There's also a special tool for pressing it out.
From this original overhaul charade,
a gray, dull-looking grease was applied generously.
Maybe it's their shop's "special grease" (laughs).


This shows the state after the right side hub body bearing was removed.


I inserted the hub shaft from outside through the left side hub body bearing
and turned it by hand—there was grinding.
This is the main cause of the bearing trouble.

With standard methods, this bearing won't budge at all.
Once the tool is set up, it should come out with the impact load from a plastic hammer.
This bearing sits deep in the setting position, but
it seems like it was driven in forcefully while tilted,
and there were burrs or raised edges in the cylindrical hole.


To avoid damaging the hub body,
I threaded a spare lock ring onto the center lock seat,


and aside from the DT hub shaft method of pressing out,
I attached a bearing removal jig and
struck it repeatedly with an iron hammer.


It came out in pieces, but I recovered it.


With DT hub bearings, one is kept straight as a guide for impact pressing,
and the special tool sets it with impact load.
For a hub body without a hub shaft passing through,
this special tool can be used as a static load jig
for press-fitting installation.
In this case, the left side of the hub body is the better choice.


I press-fitted the left side hub body bearing.
Before that, I sanded away the burrs with sandpaper.


I wiped the contact surface of the left quick-release end.
There are wear marks even on the surface below the step that presses just the inner race.


The gray grease had been smeared all over even in the grooved cutout areas
of the left quick-release end.
There's something with a red paint-like coating stuck on it,
but I'm not sure what that is.
This quick-release end doesn't have rubber O-rings or other
water-resistant features, so if water gets on it,
the bearing seal immediately becomes the final line of defense.


The four bearings total—two for the hub body and two for the freewheel body—
that were swapped in that original overhaul charade
all had metal seal bearings installed.
If you operate these with oil instead of grease,
and if you don't care about lifespan, you do get lighter rotation.
I asked the customer on that initial call whether they knew
the bearings were set up that way,
and they didn't.
With the customer's consent, I explained that this is a trade-off with lifespan,
and that such a tune-up is an option. If they'd gotten approval beforehand,
it would be fine, but they did this without asking.
Maybe they thought, "You don't need expensive ceramic bearings;
with these bearings, you get lighter rotation more cheaply"—
what absolute idiots.

The left side hub body bearing had no water intrusion or rust,
so flooding wasn't the cause,
but the fact remains that the rotation seized up
in just 2000 km or less of fair-weather riding—
an incredibly short timeframe.


The bearings are both NTN,


the metal seal version is 6902Z,
the non-contact rubber seal version is 6902LB.
The notation varies by manufacturer:
Z or ZZ means metal seal, LB or LLB means non-contact rubber seal,
LU or LLU means contact rubber seal.


For contact rubber seals, there's a convention that the seal color
should be orange or brown,
though it's not always strictly followed.
NTN uses orange.


↑In the 6902LU, you can see ridges where it contacts the inner race.
This time, I replaced all of them with non-contact rubber seal bearings.
Metal seals are also non-contact, but
within the non-contact category, the Z's water resistance
is inferior to the LB's.


The hub shaft threaded through the left hub body bearing
serves as the guide for straight pressing,
so I positioned the special tool and


set the bearing.
It protrudes slightly from the back face of the hub body,
which is the normal state.


↑This is the face of the star ratchet spline
facing the hub body, and


a washer needs to be placed in the recessed area.


If you tighten the spline with it shifted,
it becomes a real disaster (though I've never done it).


The tool tip has a step that matches the washer's inner diameter,
but as you tighten,
it doesn't automatically guide it into place, so


before you tighten with force, you must verify
it's positioned correctly.


This is the face of the water and dust seal ring
facing the hub body, and


the side with the U-shaped groove faces outward.



There's also a special tool for straight pressing it in,



and I set it this way.
It's strange that this original overhaul charade had the tool
but lacked the star ratchet special grease.
Well, I don't need to understand or bother with it anyway.


I reassembled it into front hub mode again.
When I turn the hub shaft by hand, it rotates smoothly without cloudiness.


The freewheel body spacer, which contacts the inner race of the right side hub body bearing,
has wear marks that shouldn't appear in such a short term.


Directly under this spacer, the star ratchet parts
are set up like this, but
if the hub shaft has no play or eccentricity,
wear marks like those shouldn't appear in the short term.


The star ratchet part had the lightened version with cutouts.
When buying a fairly high-end DT-brand wheel or
ordering this part separately,
you get this version.


↑This one has more ratchet teeth,
so it's an even higher grade and naturally has cutouts.


In budget-tier hubs,
a version without cutouts is installed.
You can't buy this part alone.

In the cutout areas of the star ratchet part,
even with heavy long-term use,
the pink star ratchet special grease
remains in its clean color.
This wheel had none of that left,
just gray grease worked into it.


DT has a multipurpose grease (left side of the image above)
separate from the star ratchet special grease,
but this is a different grease altogether.


The freewheel body bearing was also swapped to a 6802Z.
It showed no internal or external problems,
but after consulting with the customer, it was decided to replace it anyway.


I removed the inner bearing and the spacer that presses the inner race.


The gray grease had mixed with grime to become cotton-like,
and I picked it out with tweezers.


I removed the outer bearing.


This grease irritated me so much that



I thoroughly cleaned it along with the cutouts in the star ratchet part.


The U-grooved ring pressed into the hub body and
the protrusion on the freewheel body side form a labyrinth seal gap.
With rubber O-rings, deterioration can cause them to snap off suddenly,
but the labyrinth seal is formed entirely from metal,
so the parts don't deteriorate.
Dust can jam in and create resistance, and
a high-pressure washer's water can penetrate it, though.


The two freewheel body locations and the left side 6802 bearing
also use NTN bearings.
The NTN marking is barely visible,
around the bottom left of the image above.


I applied star ratchet special grease and
installed the freewheel body.


I wiped the contact surface of the right end bearing,
but unlike the left end, there were no uneven wear marks.

It appears there was eccentric runout with the left end
as the apex of a cone.
Anyway, that gray grease was still in there, so


I cleaned it.

After finishing the bearing replacement work,
I noticed the spoke tension on the freewheel side was slack.
This isn't about re-lacing—
the tension is below the as-shipped spec.
Someone who does such shabby bearing swap nonsense
surely would've botched wheel truing too,
so I decided to inspect the wheel as well.


I deflated the sealant-filled tubeless tire and
pushed the tire bead downward to release the bead lock, when


there were areas where the double-layer rim tape
was peeling up, and


areas where it was fused to the sealant and peeling.


↑This is where the tape ends, but


sealant had worked its way in between
the lower and upper tape layers for quite a distance from there, and


more than one-third of the rim had tape peeling sideways.
I later figured out why they'd swapped out the Rovaál-original tape.


The tire sealant is probably Caflatex, and


it had formed quite strong gumming.



The rim was centered to the right side initially,
and there was definitely radial runout they'd created,
the wheel was in pretty bad shape,
but what bothered me most was
the slack spoke tension on the freewheel side.


↑This nipple has damage marks on the inside grip area,
but when I gripped and turned the outside grip area,


the area around the tooth profile pocket
was rubbing against the spoke, not clearance,
creating heavy drag, so


I removed it.


↑The deformation looked like this.


Our shop has reluctantly built up
a pretty good setup for restoring Rovaál wheels to factory condition
apart from complete re-lacing.
So we keep spare nipples in stock.


↑This is a nipple from a different area for comparison—
no problems with this one—but


the nipple at the neighboring rim hole
was clearly non-original.


I removed it.
It was a Sapim 14mm aluminum nipple.

After replacing this with an original nipple too
and doing a rough lateral true,
when the rim position and spoke tension
roughly matched the surroundings,


the spoke had about two thread pitches protruding
beyond the nipple.
This is the freewheel side of the CLX50 rear wheel.
Investigation revealed these spokes were 249mm,
not shop-modified pieces like those on Aerolight or CX-RAY,
but genuine Rovaál original spokes.
249mm is the freewheel-side length for the CLX50 rear wheel—
but for rim-brake wheels!
The hub flange diameter and rim inner diameter are roughly the same,
but disc-brake rear wheels are 16H on the freewheel side
while rim-brake wheels are 14H, so the spoke length differs.
The correct length in this case is 252mm.
My guess: they damaged the nipple threads and the spoke got bent too,
so they threw in a 3mm-shorter spoke they had lying around
that had enough thread to grip.
By switching to a 14mm Sapim nipple,
they deliberately hid the exposed threads—
malicious intent, in the legal sense.

With Rovaál wheels, if a nipple accidentally
drops inside the rim or needs replacing,
I'd grab 2–3 nipples using a magnet
on the Zonda method without peeling tape,
but they seem to lack the skill or ingenuity for that,
so they probably peeled the tape.
This is just speculation, though.



Combined with the original centering error,
over-tightening the freewheel side pushed in lateral and radial runout.
Now the rim was centered far to the right with almost no runout.



From that state, I pulled the center back.
In the final fine runout adjustment,
I slightly loosened the freewheel side a bit and such,
but mostly relied on tightening the opposite side.


It's fixed.


The two taped areas are where I replaced the crushed nipple and
where I swapped the wrong spoke and nipple.
I used a genuine Rovaál 252mm spoke.


↑Replaced parts—
I'm handing these to the customer as evidence.
As for me, I'd like that other shop to cover
this time's parts and labor...