Should I worry about freewheel axle?

[nsew]

“Less arguable than all this is the markedly different measurement of the STX FH-MC32 hub where the flange-to-flange measurement is given on wheelbuilding websites as 60.6, not 55.”

Your rule does show about 60mm. The critical measurement is the flange offset (actually not critical). So for that hub at 135oln, measure from the centre point between lock nuts (67.5) to each flange, giving left and right offset. Which is probably where you are anyway.

[PhilD28]

I've had Sputnik rims from Spa and some have been way over 537, often between 542 and 544. they often come up large, in some cases so large that it's difficult to get tyres to pop out of the well and run true. Even had arguments with them about exactly this. I've built a lot of wheels with Sputniks and they do vary, always measure before ordering spokes.
I've had the same with other Ryde rims too, obviously they have, or have had some quality control problems in the past.

[Bice]

“Less arguable than all this is the markedly different measurement of the STX FH-MC32 hub where the flange-to-flange measurement is given on wheelbuilding websites as 60.6, not 55.”

Your rule does show about 60mm. The critical measurement is the flange offset (actually not critical). So for that hub at 135oln, measure from the centre point between lock nuts (67.5) to each flange, giving left and right offset. Which is probably where you are anyway.

No, photographing this stuff is very difficult unless you put things on a grid background. Look at the angles: the measurement is all over the place in the image, so you are right to question.

I only noticed how bad it was after I had published it. I think I have a grid, because without one there is no reason to believe.

[Bice]

I've had Sputnik rims from Spa and some have been way over 537, often between 542 and 544. they often come up large, in some cases so large that it's difficult to get tyres to pop out of the well and run true. Even had arguments with them about exactly this. I've built a lot of wheels with Sputniks and they do vary, always measure before ordering spokes.
I've had the same with other Ryde rims too, obviously they have, or have had some quality control problems in the past.

Interesting. Checking all this is so much more stressful than assembling the wheel (with the right bits, obviously) ... But I am so glad I stuck with the stern instructions of Roger Musson.

[Brucey]

actually flange spacing does not need to be super-accurately measured; just try inputting values into a spoke calculator and you will find that the flanges need to 'move' by at least 5mm in most builds before it makes a 1mm change in spoke length.

That said, measuring exact is not difficult either; using digital verniers measure one flange width, and 'set zero' there. Then measure across both flanges and you will get your centre to centre width, exact. Very easy.

Re rim ERD; if 542 is correct (so spoke ends finish mid-slot) then using 537 to calculate may seem to 'work' if

a) you are prepared to tolerate the spoke ends being about 1mm underflush of the nipple slot and/or
b) you buy spokes in (lumpen) 2mm length increments and you 'always round up'.

It baffles me that folk bother to measure everything accurately (or say they do, at least) and then 'always round up'. If you round up to the next 2mm length then that can be have the same effect as increasing the ERD by 4mm. Just another way of messing things up IMHO.

The method I use to measure rims is to use standard J-bend spokes and nipples (of the same type you intend to use) about 250mm long (say) and to fit them to the rim with the spoke ends finishing mid-slot depth in the nipples. The spacing of the J bends can then be measured exactly using digital verniers. Repeat two or three times in different places to be sure that the rim isn't egg shaped. Adding 2x the measuring spoke length to the average caliper reading gives you your ERD; simples.

If you measure everything accurately you will know if you are choosing spokes that are slightly shy of the nipple slot or slightly sticking out when you make your length choice (using 2mm increments) and equally if you are using 1mm increments then you can be pretty sure that the spoke ends are going to finish where you expect them to.

FWIW some spoke calculators (including musson's) claim to allow for spoke stretch. The reality is that this allowance is not accurately made and in any event is unlikely to account for an error of more than 0.5mm in spoke length. So in essence if you worry about this you are probably worrying about the wrong sort of thing, there are invariably bigger fish to fry than that.

cheers

[PhilD28]

actually flange spacing does not need to be super-accurately measured; just try inputting values into a spoke calculator and you will find that the flanges need to 'move' by at least 5mm in most builds before it makes a 1mm change in spoke length.

That said, measuring exact is not difficult either; using digital verniers measure one flange width, and 'set zero' there. Then measure across both flanges and you will get your centre to centre width, exact. Very easy.

Re rim ERD; if 542 is correct (so spoke ends finish mid-slot) then using 537 to calculate may seem to 'work' if

a) you are prepared to tolerate the spoke ends being about 1mm underflush of the nipple slot and/or
b) you buy spokes in (lumpen) 2mm length increments and you 'always round up'.

It baffles me that folk bother to measure everything accurately (or say they do, at least) and then 'always round up'. If you round up to the next 2mm length then that can be have the same effect as increasing the ERD by 4mm. Just another way of messing things up IMHO.

The method I use to measure rims is to use standard J-bend spokes and nipples (of the same type you intend to use) about 250mm long (say) and to fit them to the rim with the spoke ends finishing mid-slot depth in the nipples. The spacing of the J bends can then be measured exactly using digital verniers. Repeat two or three times in different places to be sure that the rim isn't egg shaped. Adding 2x the measuring spoke length to the average caliper reading gives you your ERD; simples.

If you measure everything accurately you will know if you are choosing spokes that are slightly shy of the nipple slot or slightly sticking out when you make your length choice (using 2mm increments) and equally if you are using 1mm increments then you can be pretty sure that the spoke ends are going to finish where you expect them to.

FWIW some spoke calculators (including musson's) claim to allow for spoke stretch. The reality is that this allowance is not accurately made and in any event is unlikely to account for an error of more than 0.5mm in spoke length. So in essence if you worry about this you are probably worrying about the wrong sort of thing, there are invariably bigger fish to fry than that.

cheers

Yes, this is the way I measure them and shoot for accurate length spokes that finish where i want them to when tensioned correctly. I know of one wheelbuilding shop where if the spokes come up short, through mismeasurement for whatever reason, they just overtighten the spokes, not exactly an elegent or sound solution.

[Bice]

I’d overlooked STX before but these seem to tick all of the boxes for a reliable touring hub. Another cheat (other than removing the right side spacer) to reduce dish that I’ve used to good effect is to add 2mm of spacer to the left side and reduce the exposed axle thread to 4mm at either end. Creating a 137mm OLN. This has resulted in 110 / 120 kgf spoke tension. If I had the kit to shave off the shoulder of the freehub the wheel would likely be near as dammit symmetrical.

I am going to try this, as it makes sense to me. I have already moved the 1.5mm spacer over to the left. I just want to put a cassette on and see how it all aligns.

But ... this is still going to be a significantly dished wheel. As I have said, the centre-of-flange measurements given on the spokes calculators are not correct at 60.6mm. They are just over 55mm.

With a 137mm OLN adding the 2mm spacer to the left. I have:

Left locknut to centre-of-left-flange at 37mm
Right locknut to centre-of-right-flange at 43/44mm.

[Brucey]

...But ... this is still going to be a significantly dished wheel....

Left locknut to centre-of-left-flange at 37mm
Right locknut to centre-of-right-flange at 43/44mm.

this ought to give around 76% tension balance, which is OK, better than a poke in the eye with a sharp stick...?

Having said that if you can push both flanges just 2mm further to the right, the tension balance goes up to around 89%, much better still.

cheers

[Bice]

I have 5mm of exposed axle on the right and 4mm on the left (of an OLN of 137mm).

I suppose just under dishing the wheel and verifying by putting in the frame is a too approximate process?

[slowster]

Brucey, could the total width of the drive side lock nut unit be reduced further by lapping the inner face of the lock nut and lapping the outer face of the cone?

I presume it would be necessary to file small bevels on the threads of the nuts to replace any thread relief that was removed by lapping. I also presume that the resulting small inboard movement of the seal would not affect the distance of the gap between the seal and the freehub body.

The flats of the cone only need to be just wide enough for whatever cone spanner is used, and after assembly there should be no need to undo the lock nut unit ever again since any adjustments would be made on the non-drive side.

The only problem I can imagine is if this resulted in the chain on the smallest sprocket being too close to the seat stay, since most of us would not have the tools needed to grind the shoulders of the freeehub to move the cassette closer to the spokes.

[nsew]

...But ... this is still going to be a significantly dished wheel....

Left locknut to centre-of-left-flange at 37mm
Right locknut to centre-of-right-flange at 43/44mm.

this ought to give around 76% tension balance, which is OK, better than a poke in the eye with a sharp stick...?

Having said that if you can push both flanges just 2mm further to the right, the tension balance goes up to around 89%, much better still.

cheers

Bice, it’s a bit labour intensive at this point but there was an opportunity after removing the 1.5mm spacer - to set your right side threads at four (4mm) (IIRC Shelton Brown writes 3mm is acceptable). If you’re able to set at 3mm then you can add 3mm. That will get you close to 90/100.

[nsew]

Axle Spacing Adjustment

Typical quick-release axles are 11 or 12 mm longer than the spacing of the hub locknuts. This gives 5.5-6 mm of axle protrusion on each side. You don't actually need nearly this much, so for respacing hubs to wider spacing, if you're not adding more than, say, 5-6 mm of spacers, you don't need a new axle. As long as you have 2 or 3 mm sticking out on each side, that's plenty.
One of my own bikes is set up with the axle cut off flush with the locknuts, and even this is no problem in use, though it is slightly trickier to align the wheel when installing it

https://www.sheldonbrown.com/frame-spacing.html

[Brucey]

Brucey, could the total width of the drive side lock nut unit be reduced further by lapping the inner face of the lock nut and lapping the outer face of the cone?

this depends on the model of freehub that you use. There are 'super short RH cones' which are designed for use in freehubs when the seal is mounted on a separate seal carrier. eg

https://www.sjscycles.co.uk/hub-spares/shimano-deore-fhm525sl-rear-right-lock-nut-unit-y3sp98040/


Also it isn't that difficult to grind a LH cone to be shorter either (tip, mount on a short length of axle in an electric drill, and then grind the cone end whilst it is spinning at high speed), and to add flats to it for the cone spanner. If push comes to shove a seal from a cheap hub can be used in place of whatever seal is usually fitted in the RH end of the freehub body; these seals commonly fit into the freehub body quite well, and have a lip that seals onto a 17mm diameter (which is the commonest diameter for a cone to be). This shimano seal is similar but is also slightly thicker

https://www.sjscycles.co.uk/hub-spares/shimano-fhm495a-rear-right-seal-ring-y3cr08000/

The only problem I can imagine is if this resulted in the chain on the smallest sprocket being too close to the seat stay, since most of us would not have the tools needed to grind the shoulders of the freeehub to move the cassette closer to the spokes.

you can address this by using a dished #1 sprocket from a 10s cassette (with loose sprockets) in a 7/8/9s spaced cluster or a riveted subassembly (carrier type) in a 10s/11s spaced cassette. If using a riveted subassembly the carrier can be shortened to change the offset too.

Another option is to shorten a 8s freehub body and to use that instead. Normally there are enough lockring threads in the end of the body to allow some shortening, but this isn't perfectly consistent. The shoulder dimension on the end of the freehub body is different between 7s and 8s freehubs.

cheers

[slowster]

this depends on the model of freehub that you use. There are 'super short RH cones' which are designed for use in freehubs when the seal is mounted on a separate seal carrier.

I've just had a look at the right hand lock nut unit on an STX hub, and I think it's one of those 'super short RH cones': there is very little spare width of the cone flats that could be removed.

However, it looks like there is a ~2mm wide 'shoulder' at the outboard end of the axle spacer (item 5 on the EV doc which could be reduced. The lock nut itself is ~4.5mm wide, and so could presumably also be reduced: I presume 3mm would be a safe minimum width since it would maintain engagement on 3 threads with a 1mm pitch.

Therefore if, say, ~1.5mm were ground off the axle spacer and 1mm to 1.5mm from the lock nut, that should get the OP close to 90% left hand spoke tension in combination with moving the 1.5mm washer from RH to LH side.

If push comes to shove a seal from a cheap hub can be used in place of whatever seal is usually fitted in the RH end of the freehub body; these seals commonly fit into the freehub body quite well, and have a lip that seals onto a 17mm diameter (which is the commonest diameter for a cone to be). This shimano seal is similar but is also slightly thicker

https://www.sjscycles.co.uk/hub-spares/ ... y3cr08000/

If I understand that correctly, it means removing the seal from the axle and possibly the spacer on which it sits, and fitting a seal inside the freehub body instead, and presumably replacing the former with a narrower spacer or spacers between the cone and lock nut, i.e. whatever width of spacers would be enough to prevent the cassette/chain fouling the drop out/seat stay. I presume that would result in a dishless wheel.

[Brucey]

in most cases if you stick with a 7s freehub body you can retain an original RH seal (although you might need to revise the seal carrier and locknut slightly). However if you want an even shorter freehub body then a revised (thinner) RH seal may be required. BITD UG freehubs could still have bulky RH seals in them, because there wasn't a lockring fighting for space there as well.

cheers