WH-R9270-C50-TL

A customer brought in a Dura-Ace C50 rear wheel
for me to work on.
RIMG8984amx16.jpg
It's a tubeless rim spec.
The freewheel has significant drag resistance,
and particularly when the chain is engaged near the top gear,
the chain gets pushed out so much it hangs
down from the chainstay.

Mavic's FTS-L freebody rear wheels also develop
"ghost pedaling" with age—that phantom feeling where
the crank keeps spinning on its own—but
this wheel's resistance is way more intense.
When I grab the pedal with my hand to stop the spinning crank,
it pushes back with considerable force.
It's not quite like a fixed gear, though.

The customer isn't the original owner of this wheel;
they bought it used from a shop that sells secondhand bikes.

RIMG8985amx16.jpg
Freewheel side
RIMG8986amx16.jpg
Non-freewheel side
The sticker on the hub shell shows "UA," indicating it was manufactured
in January 2022,
but the spoke paint is already peeling noticeably.

RIMG8987amx16.jpg
RIMG8988amx16.jpg
Both the left and right dropouts show marks from a cone wrench.

This hub isn't of Shimano's traditional design
where the freehub body bearings extend from the freewheel hub concept
using a cup-and-cone bearing arrangement.
With that design, if you remove the freehub body,
you're left with bearings only on the left side,
which doesn't work as a hub at all.

By contrast, this is a freewheel hub where a front hub-style bearing arrangement
with bearings on both left and right sides of the hub shell
has a freehub body added to the right side.
With this design, even if you remove the freehub body, the hub still functions.
So you can kick off the ground and do fixed-gear-style rolling,
or if you're on level terrain without uphill sections,
you can coast down from a summit without pedaling at all.

In the past, Shimano made 7800-series hubs with
"a three-pawl spring freehub body added to the right side
of a front hub-style bearing hub" for rear wheels,
but with these 9200-series hubs,
they went with "a flat-ratchet freehub body added to the right side
of a front hub-style bearing hub,"
and they've named the freewheel mechanism
"Direct Engagement."

RIMG8993amx16.jpg
RIMG8992amx16.jpg
When you pedal, helical gears engage and
the hub shell rotates in sync.
When you stop pedaling, the flat-ratchet engagement slips
and freewheels,

RIMG8989amx16.jpg
RIMG8990amx16.jpg
The internal springs are actually two
coil springs, inner and outer.
Of these, only the inner spring has directional orientation.
In the photos above, I've corrected it to the proper orientation,
but when I first disassembled it,
it was installed backwards.

Deep in the hub shell, there's a part that looks like
a fixed gear lockring or the left cone lockring of a cup-and-cone BB.
Shimano uses a proprietary tool
not meant for external use to tighten it during assembly,
and that part has six grooves
alternating between wide and narrow widths.
The inner spring's end tab—which protrudes only from one end—
is supposed to catch on the narrow grooves, and that was installed
backwards.

RIMG8995amx16.jpg

RIMG8991amx16.jpg
The freehub body can't be disassembled any further.

RIMG8994amx16.jpg
Also, there's a warning label saying "Do not oil the interior."

RIMG8996amx16.jpg
The bearing cone shows no uneven wear or pitting,
but the hub axle has strange scratches.
In the image above, the innermost spot where the anodizing has worn away—
you can feel a faint depression if you run your fingernail across it.

Actually, when the customer brought it in,
the left end's double nut was loose enough
to unscrew by hand.
The hub had bearing play, and it's been used like that over time.

There's also no grease at all around the bearings,
which is suspicious.

I have no idea what the previous owner did
or how they handled this wheel.

So I corrected the inner spring's orientation,
tightened the left-side double nut so there's no bearing play,
installed it on the customer's bike,
and checked the symptoms again.
It got slightly better, but the ghost pedaling didn't go away completely.

One thing I learned: if I loosen the through-axle tension
within the range where the wheel doesn't have side play,
the ghost pedaling decreases noticeably.
The through-axle compression seems to trigger
"something" inside the freehub body.

On a normal freehub body, even with through-axle compression,
you shouldn't get ghost pedaling—but this anomaly
is causing it to happen.

Since I can't disassemble or oil the freehub body further,
I'm limited to adjustments.
Accounting for the bearing play tightening with through-axle compression,
I loosened the bearing adjustment to the edge where
there's no side play, but ghost pedaling still occurred.

At that point the customer said,
"Maybe a shop that specializes in Dura-Ace wheels
could fix it."
So I replied, "If such a place actually exists,
by all means try them."

Since the structure is a freehub body attached to the right side
of a front hub-style bearing arrangement,
the right end nut only serves to fix the freehub body to the axle,
so adjusting the left-side double nut does shift the wheel center slightly—
but that's not the point of this story, so I didn't use a centering gauge.
The disc brake rotor and pad position relationship also changes.
This is dimensionally critical to begin with,
so there's a real possibility pads that weren't rubbing before
start rubbing after bearing adjustment.
I didn't mention these two points to the customer.
Twist-on cup (quick-release) style hubs don't have this problem
because they lack any bearing adjustment mechanism,
and Campagnolo hubs, while they do use cup-and-cone bearings,
have the adjustment mechanism built into the hub shell itself
so axle-end play doesn't change with adjustment—
they're basically trouble-free.

The freewheel coasting noise isn't a steady "whoosh"
but more like "whoosh-click whoosh-click," with a wavering quality.
And when I hold the sprocket or freehub body and rotate it gently like a safe dial,
there's some give within one ratchet tooth's worth of rotation,
but the width of this play varies from tooth to tooth,
so there's definitely an anomaly inside the freehub body.

Shimano has a consultation window for troubleshooting with Shimano products—
though it's really just a place to chat.
If I honestly mention that it's secondhand, shows signs of tampering,
the inner spring was installed backwards, and the previous owner bears responsibility,
it stays just a conversation.
If I hide all that and tell a different story,
it's still just a conversation.
They're sympathetic—yeah, yeah, I understand your feelings—
but they just suggest buying new parts or a new wheel.

My assessment is that even with the axle showing scratch marks,
replacing just the freehub body would probably fix it.
That part costs ¥30,200 with tax—
a pricey swap for what would then be a stock-condition wheel
with peeling spokes.
If you're replacing the axle too, Shimano doesn't sell just the axle alone;
it's only sold as a kit with end pieces included,
which costs ¥16,802 with tax.
Doing that much work gives you a brand-new internal hub
on a wheel with shabby spokes.

For that kind of money, you could rebuild the wheel using that rim
with a standard hub and better spokes, and I'd be confident building
a wheel that rides way better.
But I didn't suggest that because it felt disrespectful to someone
who buys Dura-Ace wheels.

One last thing: our shop runs on a no-fix-no-fee basis,
so I didn't charge the customer any labor for this.
While I was working, they looked up the cheapest source online,
and I'd suggest they just order the freehub body from there
and install it themselves, like you mentioned.

Related Products on Amazon

* Amazon affiliate links — prices may vary