A customer brought in the rear wheel of a Princeton WAKE6560 for servicing.
I'd previously replaced the front wheel bearings (→here),
and the rear wheel had play in the hub axle.
From what I observed, rather than just lateral play,
the hub rotation itself had a grinding sensation,
and it seemed at least one of the four bearings—
the left and right sides of the hub body, plus the front and rear of the freebody—was failing.
WAKE is the model name,
and 6560 means it has a wave-type shape with 65mm height and 60mm height.
Recently Fulcrum has been releasing wheels that look like they might get Princeton in trouble,
but I wonder if it's really okay.
The customer mentioned that when trying to lubricate the freebody,
they disassembled the hub and found an O-ring had broken,
and asked me to source a replacement.
From the linked front wheel work, I still don't fully understand some things,
but the ultra-thin spacers and this O-ring turned out to be critical functional parts.
Tactic Racing is a hub brand, and on Princeton's manufacturer website
when selecting wheel specifications,
it's quietly listed among well-known hub makers like
DT, Chris King, and White Industries,
but actually it's
Princeton Carbon Works' own hub brand.
Product-wise, rather than being robust and durable,
it prioritizes light weight without worrying about durability,
and smells a bit like ExtraLight hubs—
the vector is somewhat different, but it's the same kind of thing.
The problems that occurred with this hub
would almost never happen with DT or Chris King hubs in the same conditions.
If I knew this hub's structure before ordering a Princeton wheel,
I wouldn't choose Tactic's hub.
It's true that Tactic's hub is light,
but I'd probably go with DT instead.
The gap around the left cone end is already showing rust color.
I said "cone end," but
the dimensional tolerances between the hub axle's outer diameter and the cone end's inner diameter
are tight enough to be called a press-fit.
Unless removed immediately after installation,
cone ends often can't be removed without proper tools.
I removed the left cone end.
There's rust transfer on the surface that contacted the bearing.
↑Hub body left side with end removed
Rust, oil, and sand had accumulated as clay-like deposits.
In the image above, I've wiped only the left half.
～About Non-Standard Bearings～
The bearing in the image above is standard bearing 6803.
Using the notation for non-standard bearings where you write inner diameter/outer diameter/width in millimeters, it's 17265.
"Sports cycling," which frequently uses bearings up to about 30mm outer diameter,
has various non-standard bearings that differ from standard bearings by just slightly modifying the inner or outer diameter.
Most industrial products choose their dimensions to fit existing standard bearings for cost reasons,
but in sports cycling, non-standard bearings are sometimes made
when there's demand to justify the cost or because the dimensions are particularly unusual.
Shimano HG freebodies use a structure that extends the boss freewheel concept on Shimano hubs
(or rather, that's its origin),
but most other makers use two cartridge bearings inside.
Initially these were often 6902, or 15287,
but with 28mm outer diameter, the meat thickness of the spline valleys in the freebody becomes extremely thin,
so someone thought "wouldn't it be nice if there were a bearing that was 6902 but with 26mm outer diameter instead—
though nothing like that exists in standard bearings,"
and that's how 15267 was created.
This definitely originated from demand in the cycling world.
Since a 12mm through-axle passes through the rear hub axle,
the hub axle has a 12mm inner diameter,
and while 15mm outer diameter isn't impossible,
17mm is preferable,
so 6903, or 17307, is often used.
But on Mavic's Instant Drive 360 hubs,
the out-butted section of the hub axle is 18mm,
and a bearing 18307—6903 with just the inner diameter changed to 18mm—is used.
By the way, 18mm inner diameter doesn't normally exist in any standard bearing.
In early Shimano Hollowtech II bottom brackets, what's inside is
6805, or 25377.
A plastic cover is placed over this 25mm inner diameter
reducing the inner diameter to 24mm, and the 24mm outer diameter crankshaft passes through
without metal-to-metal contact.
On Campagnolo Ultra-Torque,
the crankshaft outer diameter is 25mm and appears to pass directly through 6805,
but the bearing at the root of the Ultra-Torque crank arm
is "6805 but with width reduced to 6mm,"
or essentially 25376.
Design must have required the thickness to be exactly 1mm thinner.
I don't think this is intentional obstruction to prevent generic repairs.
Recently I was surprised to see
T47 bottom bracket bearings where they're using 24377—6805 with just the inner diameter changed to 24mm—
with the Shimano crankshaft contacting it from the inner race.
BB30 gets its name from the 30mm crankshaft outer diameter,
and Cannondale's original BB30,
which pressed bearings directly into the frame and retained them with C-rings,
used the 6806, or 30427 bearing.
So this Tactic hub uses
a bearing 17287—6903 with just the outer diameter 2mm smaller—
on the right side of the hub body.
It's quite specialized, but obtainable.
The other three bearings are 6803,
with the 17mm outer diameter hub axle passing straight through.
If the bearing outer diameter were 1mm larger in radius, or 2mm larger in diameter,
the ratchet structure around here wouldn't work.
The face ratchet spring is single-acting,
and is housed in the hub body
(by the way, DT's EXP ratchet is on the freebody side).
The face ratchet contact surface is tapered.
↑This is a Novatec freebody, with a spacer that protrudes just enough
so the freebody doesn't make surface contact with the hub body bearing
(so it doesn't push anything but the inner race).
↑Extracted, it looks like this
The equivalent part on Tactic was plastic.
The freebody bearings are not both pressed in from the outside;
the inner bearing is pressed in from the inside.
Only this bearing had loose dimensional tolerances
and could be installed and removed by hand.
It doesn't appear the freebody side was worn to cause this.
The spacer supporting the inner races of the freebody bearings is omitted.
Newmen hubs had an example where the cone end is flush-tight to the hub axle (stronger against through-axle overtightening),
but the freebody's outer side and hub body's left side bearings failed quickly (→here).
Viewed from the outside, the freebody looks like this
For the bearings, the three 6803s besides the 17287 on the right side of the hub body
showed varying degrees of damage and needed replacement.
I removed the thin spacers inside the left and right cone ends.
In the image, the dirty side is the left one facing outward,
↑The face pointing inward looks like this.
I removed as much rust as possible, but
the left side actually had two stacked together.
The hub axle has several holes,
which are probably more about air vents to prevent telescoping pressure
than water drainage holes.
There are two at 180° phase,
so looking straight through you can see they're aligned.
The same holes appear on cup-and-cone Shimano complete wheels,
which don't use cartridge bearings, where the bearing inner diameter and hub axle outer diameter
aren't a tight fit.
Even then, the through-axle shaft and hub axle relationship is tight telescoping,
so the holes might be considering the through-axle and hub axle relationship
rather than the hub axle and hub body.
Both ends are flush-fit to the hub axle,
so the hub axle plus right end shown above should theoretically be separable,
but they're pressed so tightly it's not easy to remove.
↑Cleaned left end viewed from the hub body contact face.
There's a protrusion that pushes only the bearing inner race,
but actually there's also a separate groove where an O-ring is installed,
and the ultra-thin spacer also fits in between,
so this end never directly contacts the bearing inner race.
Inside the right end,
the original O-ring remained unbroken.
When I worked on the linked front wheel,
I didn't realize its importance,
but it's a functionally critical part in this hub design.
↑The original in the center came from the right end,
on the right is a generic equivalent I found with the same wire diameter and inner/outer diameters,
and on the left is another one I stocked as backup with the same wire diameter
but inner/outer diameters 1mm larger.
In the image the original appears to have thinner wire,
but originally they're the same.
The ultra-thin spacer had dimensions that just barely didn't touch the bearing outer race.
↑Shift it slightly and it looks like this
I installed the generic O-ring with matching inner/outer diameters into the groove.
It looks like it's sticking out,
but it settles properly in the end.
I installed the ultra-thin spacer.
Since this is the right side, just one as before.
↑The spacer blends into the background making it hard to see, but
when the O-ring isn't compressed, it looks like this.
With this hub, the cone ends on both sides are flush-fit to the hub axle,
so that even with through-axle compression force,
the left-to-right end spacing doesn't change.
This lets them omit the spacers inside the freebody and hub body that normally push the bearing inner races,
making it possible to live with the stupid design
of using a plastic spacer to prevent freebody flush-fit.
The left-to-right end spacing is longer than
the bearing outer-to-outer dimension from the hub body left side to the freebody right side.
If the former were smaller, it would over-compress the bearing inner races,
making the hub rotation abnormally heavy or preventing rotation entirely.
So the O-ring and ultra-thin spacer work like the wave washers
on a crankshaft tightened via bolts to one crank arm,
providing just-right pressure on the bearing inner races—
that's the stupid structure of this hub.
When the bearing, especially the outer two, wears on the inside and the inner race can move side-to-side relative to the outer race
(i.e., lateral play),
if that amount exceeds what the O-ring and ultra-thin spacer can absorb,
the hub axle develops play.
When I received the hub, I was more concerned
about the grinding rotation than the customer's reported axle play,
so I didn't really examine the play carefully.
But with this structure, just the O-ring breaking
could introduce play.
From the front wheel work, the customer had received replacement bearings from Tactic Racing saying
"if any bearings look bad, use these,"
and I also have those on hand this time.
The 17287 on the hub body right side had perfectly smooth rotation,
so I didn't replace it this time.
I replaced all the other 6803s.
The hub body left side, which was in the worst condition, was the last to be replaced,
but I ran out of customer-supplied 6803s,
so I used one from our shop's stock.
↑The hub body left side bearing that was in the worst condition.
The interior is completely rusted.
When removing the hub axle, the freebody bearing felt sticky,
and I thought either rust had frozen it or
the tolerances were very tight,
but with fresh bearings, there's no such issue.
There's definitely some resistance from tight tolerances,
but the freebody slides smoothly on the hub axle.
The ratchet part pushed up by the spring
↑when pressed down, looks like this
I applied DT ratchet grease generously.
The left end side has two spacers as before.
Maybe "two on the left" is standard and
you install two, and if the hub feels too heavy to rotate, reduce to one,
relying on the O-ring's reasonable compression—
maybe each hub gets individual tuning this way.
By the way, using a spacer of this thickness on just one side
offsets the wheel center by half that amount,
something a dishing gauge might pick up.
Fixed.
I've also done truing and dishing.
I used the dishing gauge you can apply without removing the tire.
↑This is the tire sidewall viewed from the right.
The tread pattern is reversed,
opposite from the manufacturer's specification anyway.
When I asked the customer, they said they installed it themselves.
Since sealant and such seem like a hassle,
I left the correction to the customer.
↑These are the bearings the customer provided,
with my handwriting on the bag.
Some are replacements from the front wheel work I returned.
I just returned them as proof they were replaced,
so if the customer wants to discard them, that's fine...
In the image, upper left is the original front hub left side,
upper right is the original front hub right side (with rust),
lower left is new but had gritty rotation,
and if individual bearing unit spins gritty even when not pressed in,
that's usually pretty bad.
These are all 6803,
and only the lower right 17287 remains as a healthy new bearing.
The intact right side and the broken left side
original O-rings are being returned to the customer.
For this job, I changed the label marker color to blue.
I'm providing extra generic O-rings
in addition to the ones installed in the hub.
After taking the photo, I wrote the O-ring dimensions on line three.
Anyway, the Princeton wheel bag
has a long zipper section and
opens up like a scallop shell.
I'd previously replaced the front wheel bearings (→here),
and the rear wheel had play in the hub axle.
From what I observed, rather than just lateral play,
the hub rotation itself had a grinding sensation,
and it seemed at least one of the four bearings—
the left and right sides of the hub body, plus the front and rear of the freebody—was failing.
WAKE is the model name,
and 6560 means it has a wave-type shape with 65mm height and 60mm height.
Recently Fulcrum has been releasing wheels that look like they might get Princeton in trouble,
but I wonder if it's really okay.
The customer mentioned that when trying to lubricate the freebody,
they disassembled the hub and found an O-ring had broken,
and asked me to source a replacement.
From the linked front wheel work, I still don't fully understand some things,
but the ultra-thin spacers and this O-ring turned out to be critical functional parts.
Tactic Racing is a hub brand, and on Princeton's manufacturer website
when selecting wheel specifications,
it's quietly listed among well-known hub makers like
DT, Chris King, and White Industries,
but actually it's
Princeton Carbon Works' own hub brand.
Product-wise, rather than being robust and durable,
it prioritizes light weight without worrying about durability,
and smells a bit like ExtraLight hubs—
the vector is somewhat different, but it's the same kind of thing.
The problems that occurred with this hub
would almost never happen with DT or Chris King hubs in the same conditions.
If I knew this hub's structure before ordering a Princeton wheel,
I wouldn't choose Tactic's hub.
It's true that Tactic's hub is light,
but I'd probably go with DT instead.
The gap around the left cone end is already showing rust color.
I said "cone end," but
the dimensional tolerances between the hub axle's outer diameter and the cone end's inner diameter
are tight enough to be called a press-fit.
Unless removed immediately after installation,
cone ends often can't be removed without proper tools.
I removed the left cone end.
There's rust transfer on the surface that contacted the bearing.
↑Hub body left side with end removed
Rust, oil, and sand had accumulated as clay-like deposits.
In the image above, I've wiped only the left half.
～About Non-Standard Bearings～
The bearing in the image above is standard bearing 6803.
Using the notation for non-standard bearings where you write inner diameter/outer diameter/width in millimeters, it's 17265.
"Sports cycling," which frequently uses bearings up to about 30mm outer diameter,
has various non-standard bearings that differ from standard bearings by just slightly modifying the inner or outer diameter.
Most industrial products choose their dimensions to fit existing standard bearings for cost reasons,
but in sports cycling, non-standard bearings are sometimes made
when there's demand to justify the cost or because the dimensions are particularly unusual.
Shimano HG freebodies use a structure that extends the boss freewheel concept on Shimano hubs
(or rather, that's its origin),
but most other makers use two cartridge bearings inside.
Initially these were often 6902, or 15287,
but with 28mm outer diameter, the meat thickness of the spline valleys in the freebody becomes extremely thin,
so someone thought "wouldn't it be nice if there were a bearing that was 6902 but with 26mm outer diameter instead—
though nothing like that exists in standard bearings,"
and that's how 15267 was created.
This definitely originated from demand in the cycling world.
Since a 12mm through-axle passes through the rear hub axle,
the hub axle has a 12mm inner diameter,
and while 15mm outer diameter isn't impossible,
17mm is preferable,
so 6903, or 17307, is often used.
But on Mavic's Instant Drive 360 hubs,
the out-butted section of the hub axle is 18mm,
and a bearing 18307—6903 with just the inner diameter changed to 18mm—is used.
By the way, 18mm inner diameter doesn't normally exist in any standard bearing.
In early Shimano Hollowtech II bottom brackets, what's inside is
6805, or 25377.
A plastic cover is placed over this 25mm inner diameter
reducing the inner diameter to 24mm, and the 24mm outer diameter crankshaft passes through
without metal-to-metal contact.
On Campagnolo Ultra-Torque,
the crankshaft outer diameter is 25mm and appears to pass directly through 6805,
but the bearing at the root of the Ultra-Torque crank arm
is "6805 but with width reduced to 6mm,"
or essentially 25376.
Design must have required the thickness to be exactly 1mm thinner.
I don't think this is intentional obstruction to prevent generic repairs.
Recently I was surprised to see
T47 bottom bracket bearings where they're using 24377—6805 with just the inner diameter changed to 24mm—
with the Shimano crankshaft contacting it from the inner race.
BB30 gets its name from the 30mm crankshaft outer diameter,
and Cannondale's original BB30,
which pressed bearings directly into the frame and retained them with C-rings,
used the 6806, or 30427 bearing.
So this Tactic hub uses
a bearing 17287—6903 with just the outer diameter 2mm smaller—
on the right side of the hub body.
It's quite specialized, but obtainable.
The other three bearings are 6803,
with the 17mm outer diameter hub axle passing straight through.
If the bearing outer diameter were 1mm larger in radius, or 2mm larger in diameter,
the ratchet structure around here wouldn't work.
The face ratchet spring is single-acting,
and is housed in the hub body
(by the way, DT's EXP ratchet is on the freebody side).
The face ratchet contact surface is tapered.
↑This is a Novatec freebody, with a spacer that protrudes just enough
so the freebody doesn't make surface contact with the hub body bearing
(so it doesn't push anything but the inner race).
↑Extracted, it looks like this
The equivalent part on Tactic was plastic.
The freebody bearings are not both pressed in from the outside;
the inner bearing is pressed in from the inside.
Only this bearing had loose dimensional tolerances
and could be installed and removed by hand.
It doesn't appear the freebody side was worn to cause this.
The spacer supporting the inner races of the freebody bearings is omitted.
Newmen hubs had an example where the cone end is flush-tight to the hub axle (stronger against through-axle overtightening),
but the freebody's outer side and hub body's left side bearings failed quickly (→here).
Viewed from the outside, the freebody looks like this
For the bearings, the three 6803s besides the 17287 on the right side of the hub body
showed varying degrees of damage and needed replacement.
I removed the thin spacers inside the left and right cone ends.
In the image, the dirty side is the left one facing outward,
↑The face pointing inward looks like this.
I removed as much rust as possible, but
the left side actually had two stacked together.
The hub axle has several holes,
which are probably more about air vents to prevent telescoping pressure
than water drainage holes.
There are two at 180° phase,
so looking straight through you can see they're aligned.
The same holes appear on cup-and-cone Shimano complete wheels,
which don't use cartridge bearings, where the bearing inner diameter and hub axle outer diameter
aren't a tight fit.
Even then, the through-axle shaft and hub axle relationship is tight telescoping,
so the holes might be considering the through-axle and hub axle relationship
rather than the hub axle and hub body.
Both ends are flush-fit to the hub axle,
so the hub axle plus right end shown above should theoretically be separable,
but they're pressed so tightly it's not easy to remove.
↑Cleaned left end viewed from the hub body contact face.
There's a protrusion that pushes only the bearing inner race,
but actually there's also a separate groove where an O-ring is installed,
and the ultra-thin spacer also fits in between,
so this end never directly contacts the bearing inner race.
Inside the right end,
the original O-ring remained unbroken.
When I worked on the linked front wheel,
I didn't realize its importance,
but it's a functionally critical part in this hub design.
↑The original in the center came from the right end,
on the right is a generic equivalent I found with the same wire diameter and inner/outer diameters,
and on the left is another one I stocked as backup with the same wire diameter
but inner/outer diameters 1mm larger.
In the image the original appears to have thinner wire,
but originally they're the same.
The ultra-thin spacer had dimensions that just barely didn't touch the bearing outer race.
↑Shift it slightly and it looks like this
I installed the generic O-ring with matching inner/outer diameters into the groove.
It looks like it's sticking out,
but it settles properly in the end.
I installed the ultra-thin spacer.
Since this is the right side, just one as before.
↑The spacer blends into the background making it hard to see, but
when the O-ring isn't compressed, it looks like this.
With this hub, the cone ends on both sides are flush-fit to the hub axle,
so that even with through-axle compression force,
the left-to-right end spacing doesn't change.
This lets them omit the spacers inside the freebody and hub body that normally push the bearing inner races,
making it possible to live with the stupid design
of using a plastic spacer to prevent freebody flush-fit.
The left-to-right end spacing is longer than
the bearing outer-to-outer dimension from the hub body left side to the freebody right side.
If the former were smaller, it would over-compress the bearing inner races,
making the hub rotation abnormally heavy or preventing rotation entirely.
So the O-ring and ultra-thin spacer work like the wave washers
on a crankshaft tightened via bolts to one crank arm,
providing just-right pressure on the bearing inner races—
that's the stupid structure of this hub.
When the bearing, especially the outer two, wears on the inside and the inner race can move side-to-side relative to the outer race
(i.e., lateral play),
if that amount exceeds what the O-ring and ultra-thin spacer can absorb,
the hub axle develops play.
When I received the hub, I was more concerned
about the grinding rotation than the customer's reported axle play,
so I didn't really examine the play carefully.
But with this structure, just the O-ring breaking
could introduce play.
From the front wheel work, the customer had received replacement bearings from Tactic Racing saying
"if any bearings look bad, use these,"
and I also have those on hand this time.
The 17287 on the hub body right side had perfectly smooth rotation,
so I didn't replace it this time.
I replaced all the other 6803s.
The hub body left side, which was in the worst condition, was the last to be replaced,
but I ran out of customer-supplied 6803s,
so I used one from our shop's stock.
↑The hub body left side bearing that was in the worst condition.
The interior is completely rusted.
When removing the hub axle, the freebody bearing felt sticky,
and I thought either rust had frozen it or
the tolerances were very tight,
but with fresh bearings, there's no such issue.
There's definitely some resistance from tight tolerances,
but the freebody slides smoothly on the hub axle.
The ratchet part pushed up by the spring
↑when pressed down, looks like this
I applied DT ratchet grease generously.
The left end side has two spacers as before.
Maybe "two on the left" is standard and
you install two, and if the hub feels too heavy to rotate, reduce to one,
relying on the O-ring's reasonable compression—
maybe each hub gets individual tuning this way.
By the way, using a spacer of this thickness on just one side
offsets the wheel center by half that amount,
something a dishing gauge might pick up.
Fixed.
I've also done truing and dishing.
I used the dishing gauge you can apply without removing the tire.
↑This is the tire sidewall viewed from the right.
The tread pattern is reversed,
opposite from the manufacturer's specification anyway.
When I asked the customer, they said they installed it themselves.
Since sealant and such seem like a hassle,
I left the correction to the customer.
↑These are the bearings the customer provided,
with my handwriting on the bag.
Some are replacements from the front wheel work I returned.
I just returned them as proof they were replaced,
so if the customer wants to discard them, that's fine...
In the image, upper left is the original front hub left side,
upper right is the original front hub right side (with rust),
lower left is new but had gritty rotation,
and if individual bearing unit spins gritty even when not pressed in,
that's usually pretty bad.
These are all 6803,
and only the lower right 17287 remains as a healthy new bearing.
The intact right side and the broken left side
original O-rings are being returned to the customer.
For this job, I changed the label marker color to blue.
I'm providing extra generic O-rings
in addition to the ones installed in the hub.
After taking the photo, I wrote the O-ring dimensions on line three.
Anyway, the Princeton wheel bag
has a long zipper section and
opens up like a scallop shell.