Another wheel day (and so on).
Continuing from the other day.
I built a rear wheel with a Super Champ wooden rim.
Campagnolo Earth Mark Record
120mm width freewheel large-flange hub, 36H
NJS track 15-gauge Aerostar Bright
88-Italian laced, no tie-and-solder.
The over-locknut dimension was
measured at 120.5mm width,
so with a nominally 120mm hub,
the freewheel installed would be either 5S or 6S—
fortunately that saved me here.
↑Hub before assembly
The distance from the end of the freewheel threads
to the right end is clearly very short.
The flange width (outer to outer) measured 57.3mm by my measurement.
Here, regarding the flange offset of this hub,
I'll compare it with various Shimano hubs.
The hubs I'm comparing are
this Record Large-Flange Quick-Release (LFQR) at 120mm width,
the single-threaded fixed-gear Dura-Ace track hub HB-7600-R-S,
the disc front hub HB-RS770
(I've flipped left and right to compare offset),
and finally the last single-purchase Dura-Ace-grade hub
FH-9000 with 11S freebody—four hubs total.
I'm using Shimano hubs as reference points both because
their published dimensional specs are accurate,
and because their interpretation of flange width matches mine:
outer to outer, the measurement you get pinching the flanges with calipers
(many manufacturers publish flange width as C-C dimension instead).
In other words, the spoke length I calculate
is for inpoke configuration, whereas
outpoke spokes advance further into the nipple threads—
but I absorb that difference in the redundancy of the nipple thread length.
Though front wheels are usually built with radial outpoke lacing,
I calculate spoke length using inpoke values,
so if I'm cutting spokes, I sometimes cut them
0.5mm shorter than the calculated length.
If the calculated length is 284.52mm →
rounded to 285mm, and I have a 285mm shelf spoke,
I'll use it as-is. But if I'm cutting from a 288mm spoke,
I'd cut it to 284.5mm instead.
I've added my measured left and right flange widths for the Record hub,
and the published flange width and offset for Shimano hubs.
With Shimano's published specs, you get left and right flange widths
from "flange width ÷ 2 ± offset amount"—
I keep saying they should just publish the actual left and right widths instead of making us calculate,
but getting this perspective on Shimano's "offset amount" concept might be valuable.
Since the information types differ, I'll add
the Shimano-style values for the Record hub
and the left and right flange widths for the Shimano hubs.
This reveals characteristics like:
"The Record 120mm has less offset than the single-threaded fixed-gear hub"
and "The Record 120mm right flange width matches
the disc front hub's rotor-side flange width"
and so on.
For Dura-Ace track hubs HB-7600 and HB-7710 in large and small flange versions,
front, double-threaded rear, and single-threaded rear—six hub variants total—
if you build them with eight-cross lacing using the current one-spec
Araya 16B Gold 36H track rim, you'd get
eight different spoke lengths: front left/right, double-threaded left/right,
single-threaded right, and single-threaded left.
By my spoke length calculations, the shortest would be 304.72mm
and the longest 306.52mm—
rounded to 1mm increments, that's 305–307mm.
But currently, track spokes only come in
305mm length in Hoshi 15-gauge round or aero-butted profile.
What I'm getting at is:
this Record 120mm hub has low offset comparable to fixed-gear hubs,
so there's no need to use mismatched left/right diameters
or different spoke counts on each side.
I've added "right flange width as a percentage of total flange width."
If this were 50% and the flange diameters matched,
the theoretical spoke tension would be equal on both sides (front hub condition).
From this angle too, the Record 120mm shows numbers comparable
to a single-threaded fixed-gear hub.
It can't be helped, but the FH-9000's right flange width
is so narrow—under 19mm—that it creates severe offset.
The 9-speed Dura-Ace hub FH-7700 had flanges at
right 21.1mm, left 36.9mm, which in current Shimano terms
is 58.0mm width with 7.9mm offset,
giving a right-flange percentage of 36.3%.
Both FH-7700 and FH-9000 are high-low flange designs
with a 1mm diameter difference left to right, but
that's not a difference pronounced enough to significantly affect wheel character.
If this wooden rim were built with an FH-9000 hub,
even if I cranked the freewheel-side tension to the rim's limit,
the non-freewheel side would immediately go slack at very low tension.
In that case I'd have to use tricks like mismatched left/right diameter and spoke count,
but even then it'd be touch-and-go.
So I'm really glad this hub is the Record 120mm.
As I've already written, this wheel needs enough tension
to avoid breaking the wooden rim's holes, yet high enough
that later the spokes won't loosen.
Spoke tension variation gets tighter—smaller spread—
as overall tension increases.
The three blue lines in the diagram above represent
the highest spoke tension on the high-tension-side flange
when the wheel is built at each tension level,
and their corresponding minimum tension is
↑like this in terms of scatter.
Staying under the upper line is an absolute requirement,
but if I'm too timid and build loosely, I'll drop below the lower line.
With a hub showing low offset—what Shimano calls low offset amount,
or high percentage of right-flange width in total width—
I can compress this scatter range vertically by reducing
how far the minimum tension drops from the maximum.
Usually, when I build front and rear wheels with the same spoke count and rim,
I aim for lightness in the driven wheel's outer zone,
so I use the lighter rim for the rear wheel.
Sometimes the weight difference is dramatic enough that
the aluminum nipple weight is 0g on the rear only, or
the TPU tube weight is 0g on the rear only, when compared to the front.
But this time, expecting buckling resistance, I used the heavier rim for the rear.
Wooden rims vary widely, so I can't call this rim's weight typical for wood,
the same as with aluminum or carbon—
but I have no intention of telling you the actual measured weight.
↑wow this guy's personality really sucks
Sorry for the wait!
Please enjoy these images!
↑Noooo stooooop!
Continuing from the other day.
I built a rear wheel with a Super Champ wooden rim.
Campagnolo Earth Mark Record
120mm width freewheel large-flange hub, 36H
NJS track 15-gauge Aerostar Bright
88-Italian laced, no tie-and-solder.
The over-locknut dimension was
measured at 120.5mm width,
so with a nominally 120mm hub,
the freewheel installed would be either 5S or 6S—
fortunately that saved me here.
↑Hub before assembly
The distance from the end of the freewheel threads
to the right end is clearly very short.
The flange width (outer to outer) measured 57.3mm by my measurement.
Here, regarding the flange offset of this hub,
I'll compare it with various Shimano hubs.
The hubs I'm comparing are
this Record Large-Flange Quick-Release (LFQR) at 120mm width,
the single-threaded fixed-gear Dura-Ace track hub HB-7600-R-S,
the disc front hub HB-RS770
(I've flipped left and right to compare offset),
and finally the last single-purchase Dura-Ace-grade hub
FH-9000 with 11S freebody—four hubs total.
I'm using Shimano hubs as reference points both because
their published dimensional specs are accurate,
and because their interpretation of flange width matches mine:
outer to outer, the measurement you get pinching the flanges with calipers
(many manufacturers publish flange width as C-C dimension instead).
In other words, the spoke length I calculate
is for inpoke configuration, whereas
outpoke spokes advance further into the nipple threads—
but I absorb that difference in the redundancy of the nipple thread length.
Though front wheels are usually built with radial outpoke lacing,
I calculate spoke length using inpoke values,
so if I'm cutting spokes, I sometimes cut them
0.5mm shorter than the calculated length.
If the calculated length is 284.52mm →
rounded to 285mm, and I have a 285mm shelf spoke,
I'll use it as-is. But if I'm cutting from a 288mm spoke,
I'd cut it to 284.5mm instead.
I've added my measured left and right flange widths for the Record hub,
and the published flange width and offset for Shimano hubs.
With Shimano's published specs, you get left and right flange widths
from "flange width ÷ 2 ± offset amount"—
I keep saying they should just publish the actual left and right widths instead of making us calculate,
but getting this perspective on Shimano's "offset amount" concept might be valuable.
Since the information types differ, I'll add
the Shimano-style values for the Record hub
and the left and right flange widths for the Shimano hubs.
This reveals characteristics like:
"The Record 120mm has less offset than the single-threaded fixed-gear hub"
and "The Record 120mm right flange width matches
the disc front hub's rotor-side flange width"
and so on.
For Dura-Ace track hubs HB-7600 and HB-7710 in large and small flange versions,
front, double-threaded rear, and single-threaded rear—six hub variants total—
if you build them with eight-cross lacing using the current one-spec
Araya 16B Gold 36H track rim, you'd get
eight different spoke lengths: front left/right, double-threaded left/right,
single-threaded right, and single-threaded left.
By my spoke length calculations, the shortest would be 304.72mm
and the longest 306.52mm—
rounded to 1mm increments, that's 305–307mm.
But currently, track spokes only come in
305mm length in Hoshi 15-gauge round or aero-butted profile.
What I'm getting at is:
this Record 120mm hub has low offset comparable to fixed-gear hubs,
so there's no need to use mismatched left/right diameters
or different spoke counts on each side.
I've added "right flange width as a percentage of total flange width."
If this were 50% and the flange diameters matched,
the theoretical spoke tension would be equal on both sides (front hub condition).
From this angle too, the Record 120mm shows numbers comparable
to a single-threaded fixed-gear hub.
It can't be helped, but the FH-9000's right flange width
is so narrow—under 19mm—that it creates severe offset.
The 9-speed Dura-Ace hub FH-7700 had flanges at
right 21.1mm, left 36.9mm, which in current Shimano terms
is 58.0mm width with 7.9mm offset,
giving a right-flange percentage of 36.3%.
Both FH-7700 and FH-9000 are high-low flange designs
with a 1mm diameter difference left to right, but
that's not a difference pronounced enough to significantly affect wheel character.
If this wooden rim were built with an FH-9000 hub,
even if I cranked the freewheel-side tension to the rim's limit,
the non-freewheel side would immediately go slack at very low tension.
In that case I'd have to use tricks like mismatched left/right diameter and spoke count,
but even then it'd be touch-and-go.
So I'm really glad this hub is the Record 120mm.
As I've already written, this wheel needs enough tension
to avoid breaking the wooden rim's holes, yet high enough
that later the spokes won't loosen.
Spoke tension variation gets tighter—smaller spread—
as overall tension increases.
The three blue lines in the diagram above represent
the highest spoke tension on the high-tension-side flange
when the wheel is built at each tension level,
and their corresponding minimum tension is
↑like this in terms of scatter.
Staying under the upper line is an absolute requirement,
but if I'm too timid and build loosely, I'll drop below the lower line.
With a hub showing low offset—what Shimano calls low offset amount,
or high percentage of right-flange width in total width—
I can compress this scatter range vertically by reducing
how far the minimum tension drops from the maximum.
Usually, when I build front and rear wheels with the same spoke count and rim,
I aim for lightness in the driven wheel's outer zone,
so I use the lighter rim for the rear wheel.
Sometimes the weight difference is dramatic enough that
the aluminum nipple weight is 0g on the rear only, or
the TPU tube weight is 0g on the rear only, when compared to the front.
But this time, expecting buckling resistance, I used the heavier rim for the rear.
Wooden rims vary widely, so I can't call this rim's weight typical for wood,
the same as with aluminum or carbon—
but I have no intention of telling you the actual measured weight.
↑wow this guy's personality really sucks
Sorry for the wait!
Please enjoy these images!
↑Noooo stooooop!