A customer brought in a pair of FARSPORTS carbon spoke wheels—front and rear—for service.
The front wheel had runout, and the rear wheel had dead bearings. For the bearings, I replaced all six: both left and right on the hub shell, plus front and rear on the freebody.
Let me start with the rear wheel.
It's a 20H wheel with equal numbers left and right, laced tangent with carbon spokes.
The rim's outer edge has a Vittoria tube imprint transferred onto it—not that it matters—but aside from the valve hole, there are no other holes in the rim.
For carbon spoke wheels, the standard design uses the nipple-like part of the spoke itself as an anti-rotation keeper, turning the nipple from the outer holes. But with this rim having no outer holes, here's what they've done:
They've made the rim holes larger than standard nipple holes, creating a two-stage nipple effect. The inner square anti-rotation keeper measures 3.4mm across flats, while the outer hex nipple measures 5.5mm across flats.
After removing the freebody, the rear hub rotation was rough and grinding, so I pulled both left and right hub shell bearings.
This hub uses a surface ratchet with two bamboo-shoot springs, but with DT hubs you need to remove the ratchet ring threaded into the hub shell to access the right bearing.
I pulled the original bearing. The size is 15267. That designation means inner diameter 15mm / outer diameter 26mm / width 7mm—non-standard dimensions where the numbers simply line up the three measurements in millimeters, nice and straightforward.
The standard bearing 6902, when expressed in millimeters, becomes 15287. But the sport cycling industry had a situation where someone wanted to take just the outer diameter from 28mm down to 26mm. That need was big enough that once someone made it, it would become universal—so 15267 became standardized.
If the rear hub shell bearing had been a 6902 instead, we'd have needed to remove the ratchet ring.
The bearing's rubber seal snaps into the outer race groove. Between it and the inner race there's a fold/ridge, and it's constantly rubbing—the rotation feels a bit heavy, but sealing is excellent. That's contact-type. Non-contact has a slight gap between seal and inner race. With contact-type there's an understanding that the seal should be brown or orange—but both hub shell bearings here had orange contact-type seals on the outward-facing side.
I flipped the bearing over. The right side of the image is the right bearing. The right bearing has a non-contact black seal, while the left bearing has a contact-type gray seal with visible folds. The inner-facing side—toward the hub shell or freebody interior—sometimes uses a bearing with reduced seal performance on just one side, prioritizing rotation smoothness.
Since single-side contact or dual-contact 15267 bearings aren't readily available, I replaced both with dual non-contact 15267 steel ball bearings.
For the freebody bearings: both have rough, grinding rotation, but the inner one shows rust and is in worse shape. The size is 15245, aka 6802.
I removed the bearings. The severely rusted one on the left is from the freebody's inner position, and visible in the upper right is the 15267 from before.
Flipped them. Both appear to have gray contact-type seals.
Opposite to which one actually failed, generally the outer bearing gets water damage more easily, so I replaced the outer with dual contact-type and the inner with dual non-contact.
Most freebodies have both inner and outer bearings installed from the outside. But this freebody's inner bearing installs from the inside.
With this bearing arrangement, some freebodies skip the spacer that holds both inner races—a petty weight-saving measure. This hub kept its spacer. The failure to include one on Tactic Racing hubs caused freebody bearing wear at an abnormally accelerated rate (→here), which is why they later released a spacer-equipped freebody version as V2 (Version Two) for Newmen (→here).
I reinstalled the hub and checked the center with a centering gauge. Dead on. I did some minor truing, but no center shift appeared after work.
I applied a generous amount of Scope Cycling's red grease—the non-silent type (→here)—to the ratchet teeth, but the ratchet noise is still prominent.
Next: the front wheel. But first:
With a 12mm through-axle, if the hub shaft's inner diameter is also 12mm, the shaft's outer diameter is typically 15mm. That means shaft thickness is 1.5mm.
Since the hub shaft outer diameter is 15mm, you can use bearings with 15mm inner diameter—15267, 15287 (6902), 15245 (6802), and so on.
With a 15mm through-axle, the hub shaft outer diameter often goes to 17mm, with shaft thickness of 1mm. You can use 17307 bearings, aka 6903.
However, with a 15mm inner-diameter hub shaft you have: the black drawing (top) shows a press-on end ultimately 100mm wide with 12mm inner diameter; the red shows a 15×100mm press-on end, ultimately 100mm wide and flush with the shaft inner diameter; and the blue shows a BOOST standard 15×110mm end with 10mm added width total. By offering multiple press-on ends, a single hub shaft can serve three different standards.
Separately, the 17mm outer-diameter hub shaft gets shaft thickness down to just 1mm! The cycling industry wanted that thickness to be 1.5mm instead, so they created a bearing for exactly that need:
The 18307 (right in the image above).
↑ 17307 aka 6903
↑ 18307
Stacked together you can see the 0.5mm radius difference clearly.
Actually, no standard bearing with an 18mm inner diameter exists in any outer diameter or width combination. The 18307 is used not just for 15mm through-axle front hubs, but many manufacturers like Mavic and DT also use it in 12mm through-axle rear hubs.
Now the front wheel. Equal numbers 20H tangent-laced both sides, just like the rear.
I've already replaced it with new non-contact 6903 bearings, but the front hub shaft was inner diameter 15mm / outer diameter 17mm.
The 6903 (17307) has a simple millimeter diameter difference of 13mm between outer and inner, which isn't smaller than the 15267's 11mm or the 15287 (6902)'s 13mm. So I don't think there's a "bearing inner diameter got bigger so the internal balls got smaller, meaning shorter life" issue. But I do suspect that 15mm inner-diameter shafts with 12mm-spec press-on ends—where the through-axle is only supported at the press-on end—suffer bearing damage more easily under axle compression.
The customer hadn't mentioned it, but the front hub has what feels like a safe-dial clicking sensation—different from the sensation of rust or ball wear—which makes me think the cause is over-tightening of the through-axle.
If you took this bearing-fresh front wheel and cranked down Specialized's X-12 through-axle without hesitation—actually trying to break it on purpose—you could destroy the bearing pretty quickly, I bet.
As the customer said, there was lateral runout, but the interim center showed the rim shifted left. You'd have to turn all the nipples on at least one side, and given the nipple structure that's extremely tedious. It does fix with a low-tension-side tightening bias though, so it's manageable. There was also radial runout. Not enough to notice while riding, but it showed clearly on the truing stand.
This wheel has a nipple structure where a hack can't touch it or will leave tool marks, yet there's no sign of tampering. If it had received proper maintenance in the past, the interim center wouldn't have drifted this far, and no one competent would miss a single sharp vertical hop in the runout. It's clear this wheel has gone untouched since it was bought.
The customer isn't the original owner, but they got it from a friend—not an auction from a stranger.
It took a long time, but I chased down the lateral and radial runout with a tightening bias and got the center true. For the previous owner's sake I'll stay quiet, but the front wheel became "a proper wheel" from this moment onward.
The front wheel had runout, and the rear wheel had dead bearings. For the bearings, I replaced all six: both left and right on the hub shell, plus front and rear on the freebody.
Let me start with the rear wheel.
It's a 20H wheel with equal numbers left and right, laced tangent with carbon spokes.
The rim's outer edge has a Vittoria tube imprint transferred onto it—not that it matters—but aside from the valve hole, there are no other holes in the rim.
For carbon spoke wheels, the standard design uses the nipple-like part of the spoke itself as an anti-rotation keeper, turning the nipple from the outer holes. But with this rim having no outer holes, here's what they've done:
They've made the rim holes larger than standard nipple holes, creating a two-stage nipple effect. The inner square anti-rotation keeper measures 3.4mm across flats, while the outer hex nipple measures 5.5mm across flats.
After removing the freebody, the rear hub rotation was rough and grinding, so I pulled both left and right hub shell bearings.
This hub uses a surface ratchet with two bamboo-shoot springs, but with DT hubs you need to remove the ratchet ring threaded into the hub shell to access the right bearing.
I pulled the original bearing. The size is 15267. That designation means inner diameter 15mm / outer diameter 26mm / width 7mm—non-standard dimensions where the numbers simply line up the three measurements in millimeters, nice and straightforward.
The standard bearing 6902, when expressed in millimeters, becomes 15287. But the sport cycling industry had a situation where someone wanted to take just the outer diameter from 28mm down to 26mm. That need was big enough that once someone made it, it would become universal—so 15267 became standardized.
If the rear hub shell bearing had been a 6902 instead, we'd have needed to remove the ratchet ring.
The bearing's rubber seal snaps into the outer race groove. Between it and the inner race there's a fold/ridge, and it's constantly rubbing—the rotation feels a bit heavy, but sealing is excellent. That's contact-type. Non-contact has a slight gap between seal and inner race. With contact-type there's an understanding that the seal should be brown or orange—but both hub shell bearings here had orange contact-type seals on the outward-facing side.
I flipped the bearing over. The right side of the image is the right bearing. The right bearing has a non-contact black seal, while the left bearing has a contact-type gray seal with visible folds. The inner-facing side—toward the hub shell or freebody interior—sometimes uses a bearing with reduced seal performance on just one side, prioritizing rotation smoothness.
Since single-side contact or dual-contact 15267 bearings aren't readily available, I replaced both with dual non-contact 15267 steel ball bearings.
For the freebody bearings: both have rough, grinding rotation, but the inner one shows rust and is in worse shape. The size is 15245, aka 6802.
I removed the bearings. The severely rusted one on the left is from the freebody's inner position, and visible in the upper right is the 15267 from before.
Flipped them. Both appear to have gray contact-type seals.
Opposite to which one actually failed, generally the outer bearing gets water damage more easily, so I replaced the outer with dual contact-type and the inner with dual non-contact.
Most freebodies have both inner and outer bearings installed from the outside. But this freebody's inner bearing installs from the inside.
With this bearing arrangement, some freebodies skip the spacer that holds both inner races—a petty weight-saving measure. This hub kept its spacer. The failure to include one on Tactic Racing hubs caused freebody bearing wear at an abnormally accelerated rate (→here), which is why they later released a spacer-equipped freebody version as V2 (Version Two) for Newmen (→here).
I reinstalled the hub and checked the center with a centering gauge. Dead on. I did some minor truing, but no center shift appeared after work.
I applied a generous amount of Scope Cycling's red grease—the non-silent type (→here)—to the ratchet teeth, but the ratchet noise is still prominent.
Next: the front wheel. But first:
With a 12mm through-axle, if the hub shaft's inner diameter is also 12mm, the shaft's outer diameter is typically 15mm. That means shaft thickness is 1.5mm.
Since the hub shaft outer diameter is 15mm, you can use bearings with 15mm inner diameter—15267, 15287 (6902), 15245 (6802), and so on.
With a 15mm through-axle, the hub shaft outer diameter often goes to 17mm, with shaft thickness of 1mm. You can use 17307 bearings, aka 6903.
However, with a 15mm inner-diameter hub shaft you have: the black drawing (top) shows a press-on end ultimately 100mm wide with 12mm inner diameter; the red shows a 15×100mm press-on end, ultimately 100mm wide and flush with the shaft inner diameter; and the blue shows a BOOST standard 15×110mm end with 10mm added width total. By offering multiple press-on ends, a single hub shaft can serve three different standards.
Separately, the 17mm outer-diameter hub shaft gets shaft thickness down to just 1mm! The cycling industry wanted that thickness to be 1.5mm instead, so they created a bearing for exactly that need:
The 18307 (right in the image above).
↑ 17307 aka 6903
↑ 18307
Stacked together you can see the 0.5mm radius difference clearly.
Actually, no standard bearing with an 18mm inner diameter exists in any outer diameter or width combination. The 18307 is used not just for 15mm through-axle front hubs, but many manufacturers like Mavic and DT also use it in 12mm through-axle rear hubs.
Now the front wheel. Equal numbers 20H tangent-laced both sides, just like the rear.
I've already replaced it with new non-contact 6903 bearings, but the front hub shaft was inner diameter 15mm / outer diameter 17mm.
The 6903 (17307) has a simple millimeter diameter difference of 13mm between outer and inner, which isn't smaller than the 15267's 11mm or the 15287 (6902)'s 13mm. So I don't think there's a "bearing inner diameter got bigger so the internal balls got smaller, meaning shorter life" issue. But I do suspect that 15mm inner-diameter shafts with 12mm-spec press-on ends—where the through-axle is only supported at the press-on end—suffer bearing damage more easily under axle compression.
The customer hadn't mentioned it, but the front hub has what feels like a safe-dial clicking sensation—different from the sensation of rust or ball wear—which makes me think the cause is over-tightening of the through-axle.
If you took this bearing-fresh front wheel and cranked down Specialized's X-12 through-axle without hesitation—actually trying to break it on purpose—you could destroy the bearing pretty quickly, I bet.
As the customer said, there was lateral runout, but the interim center showed the rim shifted left. You'd have to turn all the nipples on at least one side, and given the nipple structure that's extremely tedious. It does fix with a low-tension-side tightening bias though, so it's manageable. There was also radial runout. Not enough to notice while riding, but it showed clearly on the truing stand.
This wheel has a nipple structure where a hack can't touch it or will leave tool marks, yet there's no sign of tampering. If it had received proper maintenance in the past, the interim center wouldn't have drifted this far, and no one competent would miss a single sharp vertical hop in the runout. It's clear this wheel has gone untouched since it was bought.
The customer isn't the original owner, but they got it from a friend—not an auction from a stranger.
It took a long time, but I chased down the lateral and radial runout with a tightening bias and got the center true. For the previous owner's sake I'll stay quiet, but the front wheel became "a proper wheel" from this moment onward.