A customer brought in the front and rear wheels from a Cannondale
Hologram 35 Carbon wheel set for service.
The front wheel's bearings were grinding and needed replacement,
and they wanted a general inspection as well,
but both wheels turned out to be in pretty bad shape.
Let me start with the front wheel.
As I mentioned before, the front wheel is laced 2:1 only,
with the front being 24H and the rear being 28H.
That's fine, but this front wheel was laced XI pattern.
So there are others besides Roval with this lacing pattern.
It's not my first time seeing a Hologram wheel,
but I hadn't been paying close attention to it.
When I grip the spokes on both sides,
the deformation is noticeably larger on the multi-spoke side.
In fact, the front wheel of the Nomu Lab Wheel #8 that I built today
with left-right different diameter, same number lacing
(in the unlaced state) shows less difference in spoke deformation.
That said, even with Nomu Lab Wheel #8's front wheel,
you can see that the side with flatter spoke angles
has greater deformation.
That one will have lacing applied later.
Compare that to Speed 25's front wheel—
if I grip the spokes without looking at the wheel,
the deformation is so similar between left and right that I can barely tell them apart.
With a 2:1 lacing on 24H,
if you skip holes on a 32H hub and lace 2:1,
you'd get a repeating XI pattern: RRLRRLRRLRRL,
which could potentially be converted to a Zh lacing repeat
where the final crossing angle on the multi-spoke side becomes obtuse:
RRLRLRRLRLR,
but the downside is you'd end up doing 2:1 lacing on a nipple-less hub,
and the valve hole positioning would fall within the Zh pattern
rather than between Zh and Zh, which doesn't look great.
I'm not getting into that level of detail today though.
The customer mentioned there were abrasion marks on the spokes,
but if I don't judge them to require replacement,
they said to leave them as-is.
For the rear wheel,
I haven't done a spoke replacement "just because of abrasion marks."
The right side's quick-release end came out easily with hand pressure,
so I removed that one first.
There's secondary corrosion where the bearing's inner race contacts.
The bearing's inner contact surface with the hub body has a shape
that looks like a bearing puller wouldn't fit,
but that's just because the sleeve supporting the bearing's inner race from inside
is butted and ends up with this shape.
If it were just a sleeve, I could insert a rod inside the hub
and apply force at a slight angle to shift the sleeve within the hub body,
but since this sleeve-like part also supports the bearing's inner diameter on one side,
it also functions as a hub axle element.
The right bearing from the earlier image I left alone,
and I tapped out the left bearing-equipped sleeve toward the left side.
The right side's quick-release end, unlike the left,
has a portion recessed into the sleeve.
It seems to be biting in a bit—
I can't rotate the black quick-release end with one hand
while holding the silver sleeve fixed with the other hand.
I managed to pull out the quick-release end.
↑On the left side, the end is inserted into the sleeve (hub axle)
by this much.
By the point of the image above, I had also removed the right bearing from the hub body.
Directly below the right bearing is just the quick-release end,
but directly below the left bearing is a sleeve—
quite an unusual structure.
The Hologram wheel is made by XERO (Zero),
and the bearing seals also bear the XERO marking.
A through-axle shaft with 12mm outer diameter passes through
a hub axle with 12mm inner diameter, and the outer diameter of that axle can be 17mm.
In that case, the standard-size bearing 6903—
with 17mm inner diameter / 30mm outer diameter / 7mm width—
is commonly used.
However, when you want the hub axle outer diameter to be 18mm,
there are non-standard bearings
with only the 6903's inner diameter changed to 18mm,
designated as 18307, and both front wheel bearings in this case were that size.
18307 is just a name formed by stringing together the numbers 18/30/7mm.
A similar example would be the 15267 bearing,
which takes the standard bearing 6902 (15/28/7mm)
and changes only the outer diameter to 26mm—this is also a commonly-used size.
When you fit a 6902 into an HG freebody,
the thickness of the spline areas (non-tooth sections)
becomes quite thin, so 26mm outer diameter works better than 28mm
for design purposes. (For example, Tni's Evo Hub uses 6902,
but the Evo Lite Hub uses 15267.)
When it comes to sourcing 15267 bearings though,
whether it's more cost-effective than 6902 is debatable,
but examples of adopting it in hub bodies have increased.
As for 18307, it appears in Mavic's InstantDrive 360
rear hub as well, but that hub is extremely prone to water ingress,
and it's common to find that the internals have rusted
and turned into an 18307 endlessly weeping brown liquid.
The XERO logo is even printed on the seal.
There's even a stylish message included.
DT does the same thing with their hub bearings,
where the XERO bearing has a contact-type seal facing outward
and a non-contact seal facing inward toward the hub.
If you look closely, the orange-marked seal has ridges between it and the inner race,
but the gray one doesn't.
Getting bearings with this specification isn't impossible,
but with DT, regardless of size, steel ball bearings
run about ¥2,700 before tax / ¥2,970 after tax per piece—quite pricey.
While I'm at it, DT doesn't make ceramic bearing versions of 18307,
but ceramic bearings also have a fixed price
of ¥12,200 before tax / ¥13,420 after tax per piece, regardless of size.
With DT's steel ball bearings,
the outer seal is orange and inner seal is black,
while with ceramic bearings,
the outer seal is yellow and inner seal is black.
These are XERO-brand wheels,
and they even put the XERO logo on the spoke heads.
The bearings and spokes aren't made in-house—
they just order them from manufacturers.
The right bearing, whose inner race was in contact with
non-anodized aluminum on the end (no anodizing), shows
corrosion marks from the rubbing,
while the left bearing, whose inner race was in contact with
non-anodized aluminum on the sleeve, is relatively clean.
However, the left bearing's rotation was the grittier of the two.
The hub body with both bearings removed.
I pressed in the left bearing first.
Then I inserted the sleeve (or hub axle) from the right side.
↑The hub body's right side looks like this at that point.
If I'd pressed in the right bearing first,
I'd end up having to press in the left bearing with its protruding sleeve.
That's not impossible, but it's annoying.
This jig
doesn't contact the protruding part of the sleeve,
and with slightly negative tolerance, it fit just right
to maintain the centering guide function.
I pressed in the right bearing.
I'm showing photos taken with the original bearing I removed,
but this is sized to press exactly on the outer race.
This is speculation, but the reason both 18307 bearings—
such a large size—went bad is most likely
through-axle over-tightening.
However, this wheel was presumably installed on a Cannondale
frameset and ridden.
Cannondale framesets are equipped with
Mavic Speed Release–standard two-start threaded holes,
and Speed Release is designed with a mechanism
that prevents over-tightening—one that users aren't really aware of
but is essentially built in.
Given the bearing size and Cannondale's specs,
over-tightening seems unlikely, but...
Specialized's X-12 standard through-axle, promoted by Syntace,
can still be tightened another quarter turn or so
even after the wheel is sufficiently fixed,
and doing that causes abnormally rapid bearing damage.
Additionally, Roval's front wheel uses a standard 6802 bearing,
or in 18307-style notation, 15245—a small bearing—
which partly explains why you see so many Roval front wheels
with gritty bearings.
I don't know what frameset the customer has installed this wheel on,
but for through-axle installation,
once there's no lateral play in the hub
(grab the rim with tire from above the wheel and shake side to side—
if there's no play, you're good),
try going with "as loose as possible while maintaining no play."
Finally got it to temporary center.
The rim has shifted toward the left side (the rotor mount side, multi-spoke side)...
At the phase with maximum lateral runout,
the spoke comes just barely touching the gauge on the truing stand,
but at most other phases there's this much clearance.
And the runout isn't just at one spot—
there are roughly three major wobbles.
On top of that, there's also radial runout that I'd say
wouldn't pass a manufacturer's quality check.
Some of that might be because of the XI lacing,
but the multi-spoke side was pretty loose,
so knowing that the center shift would initially increase, I went ahead
and focused on tightening the multi-spoke side to address
lateral and radial runout, then brought the rim from
"runout-free but offset to the left" back to center
by tightening the single-spoke side.
In other words, there's not a single nipple that was loosened
from the start.
Now for the rear wheel.
It's left-right same number lacing, 28H.
Both front and rear spokes are built with
all-compe equivalent round butted spokes.
Of the front and rear left-right ends, only the rear right
is threaded rather than quick-release.
When there's a note saying which direction to loosen,
it's usually reverse-threaded like this.
To tighten and loosen this, the method for preventing rotation is
to insert a 4mm allen key into the hub axle
after removing the left quick-release end.
That's fine, but dirty grease has leaked out from the bearing
through the contact-type seal.
The area around the shoulder where the quick-release end presses
the inner race was also dirty.
The freewheel body's pawl area is gunked up with dirty grease,
but the bearing on the hub body's right side seems okay.
Since it wasn't as bad as the front wheel,
the customer didn't suggest replacing it,
and the hub rotation was better, but
the rear wheel also had pretty gritty rotation—
just better than the extremely gritty front.
I judged the freewheel-side bearing to likely be in better condition,
so after tapping out the left bearing with its hub axle,
Hologram 35 Carbon wheel set for service.
The front wheel's bearings were grinding and needed replacement,
and they wanted a general inspection as well,
but both wheels turned out to be in pretty bad shape.
Let me start with the front wheel.
As I mentioned before, the front wheel is laced 2:1 only,
with the front being 24H and the rear being 28H.
That's fine, but this front wheel was laced XI pattern.
So there are others besides Roval with this lacing pattern.
It's not my first time seeing a Hologram wheel,
but I hadn't been paying close attention to it.
When I grip the spokes on both sides,
the deformation is noticeably larger on the multi-spoke side.
In fact, the front wheel of the Nomu Lab Wheel #8 that I built today
with left-right different diameter, same number lacing
(in the unlaced state) shows less difference in spoke deformation.
That said, even with Nomu Lab Wheel #8's front wheel,
you can see that the side with flatter spoke angles
has greater deformation.
That one will have lacing applied later.
Compare that to Speed 25's front wheel—
if I grip the spokes without looking at the wheel,
the deformation is so similar between left and right that I can barely tell them apart.
With a 2:1 lacing on 24H,
if you skip holes on a 32H hub and lace 2:1,
you'd get a repeating XI pattern: RRLRRLRRLRRL,
which could potentially be converted to a Zh lacing repeat
where the final crossing angle on the multi-spoke side becomes obtuse:
RRLRLRRLRLR,
but the downside is you'd end up doing 2:1 lacing on a nipple-less hub,
and the valve hole positioning would fall within the Zh pattern
rather than between Zh and Zh, which doesn't look great.
I'm not getting into that level of detail today though.
The customer mentioned there were abrasion marks on the spokes,
but if I don't judge them to require replacement,
they said to leave them as-is.
For the rear wheel,
I haven't done a spoke replacement "just because of abrasion marks."
The right side's quick-release end came out easily with hand pressure,
so I removed that one first.
There's secondary corrosion where the bearing's inner race contacts.
The bearing's inner contact surface with the hub body has a shape
that looks like a bearing puller wouldn't fit,
but that's just because the sleeve supporting the bearing's inner race from inside
is butted and ends up with this shape.
If it were just a sleeve, I could insert a rod inside the hub
and apply force at a slight angle to shift the sleeve within the hub body,
but since this sleeve-like part also supports the bearing's inner diameter on one side,
it also functions as a hub axle element.
The right bearing from the earlier image I left alone,
and I tapped out the left bearing-equipped sleeve toward the left side.
The right side's quick-release end, unlike the left,
has a portion recessed into the sleeve.
It seems to be biting in a bit—
I can't rotate the black quick-release end with one hand
while holding the silver sleeve fixed with the other hand.
I managed to pull out the quick-release end.
↑On the left side, the end is inserted into the sleeve (hub axle)
by this much.
By the point of the image above, I had also removed the right bearing from the hub body.
Directly below the right bearing is just the quick-release end,
but directly below the left bearing is a sleeve—
quite an unusual structure.
The Hologram wheel is made by XERO (Zero),
and the bearing seals also bear the XERO marking.
A through-axle shaft with 12mm outer diameter passes through
a hub axle with 12mm inner diameter, and the outer diameter of that axle can be 17mm.
In that case, the standard-size bearing 6903—
with 17mm inner diameter / 30mm outer diameter / 7mm width—
is commonly used.
However, when you want the hub axle outer diameter to be 18mm,
there are non-standard bearings
with only the 6903's inner diameter changed to 18mm,
designated as 18307, and both front wheel bearings in this case were that size.
18307 is just a name formed by stringing together the numbers 18/30/7mm.
A similar example would be the 15267 bearing,
which takes the standard bearing 6902 (15/28/7mm)
and changes only the outer diameter to 26mm—this is also a commonly-used size.
When you fit a 6902 into an HG freebody,
the thickness of the spline areas (non-tooth sections)
becomes quite thin, so 26mm outer diameter works better than 28mm
for design purposes. (For example, Tni's Evo Hub uses 6902,
but the Evo Lite Hub uses 15267.)
When it comes to sourcing 15267 bearings though,
whether it's more cost-effective than 6902 is debatable,
but examples of adopting it in hub bodies have increased.
As for 18307, it appears in Mavic's InstantDrive 360
rear hub as well, but that hub is extremely prone to water ingress,
and it's common to find that the internals have rusted
and turned into an 18307 endlessly weeping brown liquid.
The XERO logo is even printed on the seal.
There's even a stylish message included.
DT does the same thing with their hub bearings,
where the XERO bearing has a contact-type seal facing outward
and a non-contact seal facing inward toward the hub.
If you look closely, the orange-marked seal has ridges between it and the inner race,
but the gray one doesn't.
Getting bearings with this specification isn't impossible,
but with DT, regardless of size, steel ball bearings
run about ¥2,700 before tax / ¥2,970 after tax per piece—quite pricey.
While I'm at it, DT doesn't make ceramic bearing versions of 18307,
but ceramic bearings also have a fixed price
of ¥12,200 before tax / ¥13,420 after tax per piece, regardless of size.
With DT's steel ball bearings,
the outer seal is orange and inner seal is black,
while with ceramic bearings,
the outer seal is yellow and inner seal is black.
These are XERO-brand wheels,
and they even put the XERO logo on the spoke heads.
The bearings and spokes aren't made in-house—
they just order them from manufacturers.
The right bearing, whose inner race was in contact with
non-anodized aluminum on the end (no anodizing), shows
corrosion marks from the rubbing,
while the left bearing, whose inner race was in contact with
non-anodized aluminum on the sleeve, is relatively clean.
However, the left bearing's rotation was the grittier of the two.
The hub body with both bearings removed.
I pressed in the left bearing first.
Then I inserted the sleeve (or hub axle) from the right side.
↑The hub body's right side looks like this at that point.
If I'd pressed in the right bearing first,
I'd end up having to press in the left bearing with its protruding sleeve.
That's not impossible, but it's annoying.
This jig
doesn't contact the protruding part of the sleeve,
and with slightly negative tolerance, it fit just right
to maintain the centering guide function.
I pressed in the right bearing.
I'm showing photos taken with the original bearing I removed,
but this is sized to press exactly on the outer race.
This is speculation, but the reason both 18307 bearings—
such a large size—went bad is most likely
through-axle over-tightening.
However, this wheel was presumably installed on a Cannondale
frameset and ridden.
Cannondale framesets are equipped with
Mavic Speed Release–standard two-start threaded holes,
and Speed Release is designed with a mechanism
that prevents over-tightening—one that users aren't really aware of
but is essentially built in.
Given the bearing size and Cannondale's specs,
over-tightening seems unlikely, but...
Specialized's X-12 standard through-axle, promoted by Syntace,
can still be tightened another quarter turn or so
even after the wheel is sufficiently fixed,
and doing that causes abnormally rapid bearing damage.
Additionally, Roval's front wheel uses a standard 6802 bearing,
or in 18307-style notation, 15245—a small bearing—
which partly explains why you see so many Roval front wheels
with gritty bearings.
I don't know what frameset the customer has installed this wheel on,
but for through-axle installation,
once there's no lateral play in the hub
(grab the rim with tire from above the wheel and shake side to side—
if there's no play, you're good),
try going with "as loose as possible while maintaining no play."
Finally got it to temporary center.
The rim has shifted toward the left side (the rotor mount side, multi-spoke side)...
At the phase with maximum lateral runout,
the spoke comes just barely touching the gauge on the truing stand,
but at most other phases there's this much clearance.
And the runout isn't just at one spot—
there are roughly three major wobbles.
On top of that, there's also radial runout that I'd say
wouldn't pass a manufacturer's quality check.
Some of that might be because of the XI lacing,
but the multi-spoke side was pretty loose,
so knowing that the center shift would initially increase, I went ahead
and focused on tightening the multi-spoke side to address
lateral and radial runout, then brought the rim from
"runout-free but offset to the left" back to center
by tightening the single-spoke side.
In other words, there's not a single nipple that was loosened
from the start.
Now for the rear wheel.
It's left-right same number lacing, 28H.
Both front and rear spokes are built with
all-compe equivalent round butted spokes.
Of the front and rear left-right ends, only the rear right
is threaded rather than quick-release.
When there's a note saying which direction to loosen,
it's usually reverse-threaded like this.
To tighten and loosen this, the method for preventing rotation is
to insert a 4mm allen key into the hub axle
after removing the left quick-release end.
That's fine, but dirty grease has leaked out from the bearing
through the contact-type seal.
The area around the shoulder where the quick-release end presses
the inner race was also dirty.
The freewheel body's pawl area is gunked up with dirty grease,
but the bearing on the hub body's right side seems okay.
Since it wasn't as bad as the front wheel,
the customer didn't suggest replacing it,
and the hub rotation was better, but
the rear wheel also had pretty gritty rotation—
just better than the extremely gritty front.
I judged the freewheel-side bearing to likely be in better condition,
so after tapping out the left bearing with its hub axle,
Related Products on Amazon
* Amazon affiliate links — prices may vary