It's wheel day again (and so on).
A customer brought in both front and rear wheels—apparently a Gokiso complete wheelset.
This is a customer who had a Nomu Lab Wheel #5 built with a Gokiso hub years back,
but this new one doesn't roll nearly as well, so they're asking for a rebuild. Gokiso-lution, if you will (not a typo, intentional pun).
"Apparently a complete wheelset" is because this wheel isn't the original owner's,
so I don't have all the details.
The rear wheel is the performance problem, but since this carbon tubular rim doesn't have particularly great height-to-weight ratio,
the customer wants to rebuild it with what you'd call Chinese carbon lightweight tubular rims—
so I'm rebuilding the front wheel too.
The hub is a Gokiso Climber hub, and
the rim is Gokiso as well.
I'm not certain, but even though Gokiso sells hubs individually,
they don't seem to sell rims separately—
the only way to get one appears to be buying a complete wheelset.
I checked my 2016 catalog too, and it only listed rim replacement costs for complete wheels.
No separate rim sales there either.
The rim is an Equinox WH175, same as on Nomu Lab Wheel #2 and others.
It has internal nipples and low rim height,
so the end face of the inverted nipple on the outer side sits
dangerously close to the rim tape surface.
Use something like Panaracer Poly Lite rim tape and you'll get pinch flats within a few months.
Gokiso does sell what you'd call veloplug for their wheels, but
unless you're doing something weird like double-wrapping tubeless tape on this non-tubeless rim,
veloplugs might be the better move.
By the way, when it came in, there was Schwalbe stretch-band rim tape on it—
not applied by the current owner.
The rim holes on the inner circumference have obvious radial runout,
so you can't build it rim-reversed.
The outer holes are centered, but many had nipples whose cores were offset from center
with no discernible pattern.
On a rear wheel, you'd expect alternating free-side centered and non-free-side tilted nipples,
but this is a front wheel with no octo.
Could spoke plain sections get bent by leverage when applying a slight tilting force with the internal nipple tool?
Sure. But I think I've figured out why this happened so often.
Here's the thing though: this doesn't happen from years of use.
If it's like this, it was like this from the factory.
The spokes have a capital H stamp on the head, so they're Hoshi Starbrite,
and the magnetic test confirmed they're genuine.
Not counterfeit like what flooded out from Hoshi during that rough period.
That's not luck—there's a reason.
Traditional Japanese spoke nipples have 3.4mm hex flats on the grip section,
but internal nipple tools only come in 3.2mm, so
this wheel was built with DT brass nipples.
Internal "proprietary" nipples don't have a slot on the face that contacts the outer circumference.
If there's a slot, shaved carbon rim debris collects there—
the tool subtly scrapes the rim.
That's not inherently a problem; plenty of wheels use inverted standard nipples as internals.
But internal proprietary nipples sometimes have the thread start point earlier than inverted types,
and if you missize the spokes on the long side,
you run out of thread and bottom-out.
Here's the face of this wheel's inverted-type nipple where it contacted the rim,
and there's a round mark on the end face.
I wrote about this before (→here),
it's the identifying mark for a #15 nipple.
Of the aero Starbrights that Hoshi lost the specs for
(if you say they didn't, then do a full reproduction and I'll buy them),
only Type III passes through round holes.
And that Aero SB3 has square-aero barreling.
Also, aero SB in general tends to have oddly long plain sections
(round-profile parts) under the head,
and that proportion isn't consistent relative to total spoke length.
These H-stamped aero spokes we have here have elliptic-aero barreling,
very short plain sections under the head,
and are #15-based.
These are custom orders: manufactured from the same process as NJS track racing spokes at 305mm,
just with different length and color specs.
Once I figured that out, everything clicked.
First, the magnetism. To my knowledge, fake Starbrights don't exist for track racing aero SB.
And probably never will.
The spokes got bent in the rim because they're thinner than #14 and bend more easily.
A 0.2mm difference makes an enormous difference in this regard.
And if this wasn't Gokiso in-house assembly but rather built somewhere else from rim and hub,
they'd need to have spokes with those exact special specs and color
in the precise length for this wheel—
which is basically impossible.
Also on the rear wheel note: doing anti-freewheel radial lacing on left-right equal spoke count rears
(confirmed garbage wheels since the 20th century)
well into the 21st century has that delightfully Gokiso flavor to it.
Built.
Though it's labeled 35mm height, measured 35.5mm,
but nominal dimension and all that.
Shimano's R9100 C40 wheel has the WO rim version as the exact same rim (cosmetics differ)
as the 9000 C35 at 35mm height, yet it calls itself C40,
and the tubular version is 37mm high on a redesigned rim, also calling itself C40—
so compared to that, this is small potatoes.
Gokiso Climber Hub, 20H
Black CX-RAY outpoke radial laced.
The hub flange connects to the hub body like an upright pine root,
so even with inpoke radial lacing, spoke tension shouldn't create axial play on the hub shaft—
but the flange holes are set pretty far inboard relative to the apparent flange outer diameter,
so inpoke radial lacing causes spokes to contact the hub flange.
Looking at the flange interior on the bare hub, there are marks from outpoke,
so there's definitely contact to varying degrees.
This rim, despite being called Chinese carbon, comes through distributors,
so accountability is clear.
The claimed weight is 270g, but
actual weight came in over 400g.
Oh sorry, that was the pre-rebuild rim.
This is the one we used this time.
The customer has tubular experience already,
so tire handling isn't an issue.
A customer brought in both front and rear wheels—apparently a Gokiso complete wheelset.
This is a customer who had a Nomu Lab Wheel #5 built with a Gokiso hub years back,
but this new one doesn't roll nearly as well, so they're asking for a rebuild. Gokiso-lution, if you will (not a typo, intentional pun).
"Apparently a complete wheelset" is because this wheel isn't the original owner's,
so I don't have all the details.
The rear wheel is the performance problem, but since this carbon tubular rim doesn't have particularly great height-to-weight ratio,
the customer wants to rebuild it with what you'd call Chinese carbon lightweight tubular rims—
so I'm rebuilding the front wheel too.
The hub is a Gokiso Climber hub, and
the rim is Gokiso as well.
I'm not certain, but even though Gokiso sells hubs individually,
they don't seem to sell rims separately—
the only way to get one appears to be buying a complete wheelset.
I checked my 2016 catalog too, and it only listed rim replacement costs for complete wheels.
No separate rim sales there either.
The rim is an Equinox WH175, same as on Nomu Lab Wheel #2 and others.
It has internal nipples and low rim height,
so the end face of the inverted nipple on the outer side sits
dangerously close to the rim tape surface.
Use something like Panaracer Poly Lite rim tape and you'll get pinch flats within a few months.
Gokiso does sell what you'd call veloplug for their wheels, but
unless you're doing something weird like double-wrapping tubeless tape on this non-tubeless rim,
veloplugs might be the better move.
By the way, when it came in, there was Schwalbe stretch-band rim tape on it—
not applied by the current owner.
The rim holes on the inner circumference have obvious radial runout,
so you can't build it rim-reversed.
The outer holes are centered, but many had nipples whose cores were offset from center
with no discernible pattern.
On a rear wheel, you'd expect alternating free-side centered and non-free-side tilted nipples,
but this is a front wheel with no octo.
Could spoke plain sections get bent by leverage when applying a slight tilting force with the internal nipple tool?
Sure. But I think I've figured out why this happened so often.
Here's the thing though: this doesn't happen from years of use.
If it's like this, it was like this from the factory.
The spokes have a capital H stamp on the head, so they're Hoshi Starbrite,
and the magnetic test confirmed they're genuine.
Not counterfeit like what flooded out from Hoshi during that rough period.
That's not luck—there's a reason.
Traditional Japanese spoke nipples have 3.4mm hex flats on the grip section,
but internal nipple tools only come in 3.2mm, so
this wheel was built with DT brass nipples.
Internal "proprietary" nipples don't have a slot on the face that contacts the outer circumference.
If there's a slot, shaved carbon rim debris collects there—
the tool subtly scrapes the rim.
That's not inherently a problem; plenty of wheels use inverted standard nipples as internals.
But internal proprietary nipples sometimes have the thread start point earlier than inverted types,
and if you missize the spokes on the long side,
you run out of thread and bottom-out.
Here's the face of this wheel's inverted-type nipple where it contacted the rim,
and there's a round mark on the end face.
I wrote about this before (→here),
it's the identifying mark for a #15 nipple.
Of the aero Starbrights that Hoshi lost the specs for
(if you say they didn't, then do a full reproduction and I'll buy them),
only Type III passes through round holes.
And that Aero SB3 has square-aero barreling.
Also, aero SB in general tends to have oddly long plain sections
(round-profile parts) under the head,
and that proportion isn't consistent relative to total spoke length.
These H-stamped aero spokes we have here have elliptic-aero barreling,
very short plain sections under the head,
and are #15-based.
These are custom orders: manufactured from the same process as NJS track racing spokes at 305mm,
just with different length and color specs.
Once I figured that out, everything clicked.
First, the magnetism. To my knowledge, fake Starbrights don't exist for track racing aero SB.
And probably never will.
The spokes got bent in the rim because they're thinner than #14 and bend more easily.
A 0.2mm difference makes an enormous difference in this regard.
And if this wasn't Gokiso in-house assembly but rather built somewhere else from rim and hub,
they'd need to have spokes with those exact special specs and color
in the precise length for this wheel—
which is basically impossible.
Also on the rear wheel note: doing anti-freewheel radial lacing on left-right equal spoke count rears
(confirmed garbage wheels since the 20th century)
well into the 21st century has that delightfully Gokiso flavor to it.
Built.
Though it's labeled 35mm height, measured 35.5mm,
but nominal dimension and all that.
Shimano's R9100 C40 wheel has the WO rim version as the exact same rim (cosmetics differ)
as the 9000 C35 at 35mm height, yet it calls itself C40,
and the tubular version is 37mm high on a redesigned rim, also calling itself C40—
so compared to that, this is small potatoes.
Gokiso Climber Hub, 20H
Black CX-RAY outpoke radial laced.
The hub flange connects to the hub body like an upright pine root,
so even with inpoke radial lacing, spoke tension shouldn't create axial play on the hub shaft—
but the flange holes are set pretty far inboard relative to the apparent flange outer diameter,
so inpoke radial lacing causes spokes to contact the hub flange.
Looking at the flange interior on the bare hub, there are marks from outpoke,
so there's definitely contact to varying degrees.
This rim, despite being called Chinese carbon, comes through distributors,
so accountability is clear.
The claimed weight is 270g, but
actual weight came in over 400g.
Oh sorry, that was the pre-rebuild rim.
This is the one we used this time.
The customer has tubular experience already,
so tire handling isn't an issue.