Freewheeling Design for the Shimano Hollowtec II Crankset

[spinningmagnets]

Stieber CSK35PP sprag clutch with a support bearing
This is STRONG AND STEALTH.
No ENO clicking.

The PAS magnets are a sweet added dessert, but...the steak in this entree is awesome. Well done! Thanks for posting.

 

[bh4801]

Is this a crankset you made yourself? Or purchased somewhere else?

Some parts I purchased, some of them I had made to my specification acc. to 3D design.
The shopping list comprised the Shimano FC-M815-2 crankset, 2 x 38T 104mm BCD 5mm thick DH chainrings, a set of 4 15mm long MORTOP DH chainring bolts, Stieber CSK35PP sprag clutch, stainless 61811.2RS bearing, 2 keyways, 18 x 4mm dia. neodymium magnets, some 7075 aluminum bar stock.
I had the following work done in a mill/lathe shop:
- lathed the spider off the crankset
- lathed a groove for the seal in the crank arm
- milled a groove for the keyway to the crankset
- lathed an adapter to hold together the clutch, bearing, magnets and keyway on the crankset
- CNCd the chainring adapter to hold the 2 chainrings and a keyway
- CNCd the flange between the clutch and the red seal.

 

[bh4801]

The PAS magnets are a sweet added dessert, but...the steak in this entree is awesome. Well done! Thanks for posting.

Thanks spinningmagnets, glad you like my spinning magnets :lol:

 

[simon.s]

Is this a crankset you made yourself? Or purchased somewhere else?

Some parts I purchased, some of them I had made to my specification acc. to 3D design.
The shopping list comprised the Shimano FC-M815-2 crankset, 2 x 38T 104mm BCD 5mm thick DH chainrings, a set of 4 15mm long MORTOP DH chainring bolts, Stieber CSK35PP sprag clutch, stainless 61811.2RS bearing, 2 keyways, 18 x 4mm dia. neodymium magnets, some 7075 aluminum bar stock.
I had the following work done in a mill/lathe shop:
- lathed the spider off the crankset
- lathed a groove for the seal in the crank arm
- milled a groove for the keyway to the crankset
- lathed an adapter to hold together the clutch, bearing, magnets and keyway on the crankset
- CNCd the chainring adapter to hold the 2 chainrings and a keyway
- CNCd the flange between the clutch and the red seal.

That's top bananas, that is. Very, very, nice work. I assume in this picture, the milled cutout is for the key? If so, I think that's the bit I'd be worried about breaking.

 

[bh4801]

That's top bananas, that is. Very, very, nice work. I assume in this picture, the milled cutout is for the key? If so, I think that's the bit I'd be worried about breaking.

Thanks
So far I made 1100km offroad, a handful of DH runs in a bikepark, jumps, shores, drops, the usual stuff
The bike handles just as well as before conversion, the confidence is there for sure, I ride faster, farther and longer.
Regarding your remark, please remember, the crank is PRESSED into the clutch, the beefy inner race of the clutch tightly clamps the crankarm, key and groove.
Only time will tell.

 

[simon.s]

That's top bananas, that is. Very, very, nice work. I assume in this picture, the milled cutout is for the key? If so, I think that's the bit I'd be worried about breaking.

lease remember, the crank is PRESSED into the clutch, the beefy inner race of the clutch tightly clamps the crankarm, key and groove.

Please, don't consider my previous comment in any way a criticism. As you say, only time will tell, and if you've managed 1100km so far on it, it's doing pretty well.

I'm interested in your construction, by the way. My understanding is that you have the crankset with a 35mm OD carrier mounted to it, that carrier having a keyway and carrying the PAS magnet ring, right? The sprag clutch and support bearing are then mounted to that carrier, and the key that holds the sprag clutch runs a dual duty, also keying the carrier to the crankset, via the milled slot in the remaining nub of crankset we see in the photo. Then the new spider and rings mount direct to the outside of the sprag clutch, again held with a key. Is that right?

 

[bh4801]

Please, don't consider my previous comment in any way a criticism. As you say, only time will tell, and if you've managed 1100km so far on it, it's doing pretty well.

Absolutely not , we are exchanging ideas with no personal implication

I'm interested in your construction, by the way. My understanding is that you have the crankset with a 35mm OD carrier mounted to it, that carrier having a keyway and carrying the PAS magnet ring, right? The sprag clutch and support bearing are then mounted to that carrier, and the key that holds the sprag clutch runs a dual duty, also keying the carrier to the crankset, via the milled slot in the remaining nub of crankset we see in the photo. Then the new spider and rings mount direct to the outside of the sprag clutch, again held with a key. Is that right?

Your understanding is absolutely correct

Last, but not least, the red polyurethane v-profile compression seal keeps the elements out - the freewheeling acts without drag and has been maintenance-free since first installed.

 

[simon.s]

Your understanding is absolutely correct

OK. I'd probably have solved the problem differently, the slightest play in the keyway and it *will* get all chewed up by the key, especially if the carrier, like the nub of remaining crankset, is aluminium.

My gut feel would be to lathe the whole of the spider off, then shrink fit or loctite a keywayed steel collar to the crankset, with pretty much everything else the same except bolting the PAS ring to the collar.

That said, the failure mode for that is probably be due to a stress riser at the end of the shrink-fitted collar, i.e. breaking the crankset right half in two, vs a failure to transmit torque with your setup. So probably a tight fit (K7/h6) with loctite 648.

 

[Chalo]

Stieber CSK35PP sprag clutch with a support bearing
This is STRONG AND STEALTH.
No ENO clicking.

The Stieber CSK35PP is rated at 175 Nm nominal, twice that max.
At 170 mm crank length, the crankset is strong enough for a 100 kg / 220 lbs rider STOMPING on his pedals

That's cool. Maybe one day you can make a full scale version for adult riders. :wink:

 

[bh4801]

...the slightest play in the keyway and it *will* get all chewed up by the key...

Unlike cycloid gearboxes, my drive unit is vibration-free (belt primary and planetary secondary), therefore I believe in my construction the the key is sufficiently immobilised in all three axes:
1. inside the steel clutch keyway along its entire length to "left and right" and "up and down" and "rotationally"
2. between the steel clutch keyway and steel crank axle along its "height"
3. between the crank arm and the carrier along its "width", but I do not think this is relevant

That said, the failure mode for that is probably be due to a stress riser at the end of the shrink-fitted collar, i.e. breaking the crankset right half in two, vs a failure to transmit torque with your setup.

Not sure I understand, are you suggesting a choice between right crank arm separating from axle versus losing the ability to drive the bike with pedals?

 

[bh4801]

That's cool. Maybe one day you can make a full scale version for adult riders. :wink:

Good 'ol cranky cheeky Chalo, thanks for chiming in :lol:
One would have though why not use a CSK40 and larger BCD chainrings?
I can see there are 38T 130 BCD ones but... the added width of a CSK40 would consume too much axle real estate and you would not be able to fit the left crank arm.
Therefore no, I do not see it clearing the swingarm/chainstay on a DH MTB unless you additionally mess up your chainline, at least that would be the case on my FSR.
We both know you would not, being famous for proper chainline fetish :wink:

 

[simon.s]

2. between the steel clutch keyway and steel crank axle along its "height"

The crank axle's steel. Of course. I R An Idjit.

Not sure I understand, are you suggesting a choice between right crank arm separating from axle versus losing the ability to drive the bike with pedals?

Nah, I was suggesting that a stress riser failure mode of a shrink fit on a presumed aluminium axle certainly wasn't a good choice compared to a potential failure mode of "chewed up keyway buggers up pedalling". Hence the suggestion of a looser fit and loctite

Like I said, that's some really nice work you've done there.

 

[bh4801]

...Like I said, that's some really nice work you've done there.

I couldn't be happier with your reply, thanks a million! 8)

Now, is there anyone out there to take on Chalo's challenge of fitting a CSK40 to the FC-M815?
That is to boost rider weight rating to some 150+ kgs or DH Elites stomping on their pedals...

A word of caution: the FC-M815 fits an 83 mm BB.
My FSR is 73mm BB and the axle length protruding to the left is just enough to correctly fit the left crank arm.
In my setup, the polished axle races fit nicely with BB bearings on both sides - if anyone dared CSK40 with a support bearing, they would quickly find out, the polished axle races will be misaligned relative to the BB cups, therefore BB bearings would run in unpolished, lathed sections of the axle or there would be too little axle protruding to the left to allow left crank arm to be safety fitted.

Is that a problem?
I never tested.

 

[liveforphysics]

Nice job!

I hope that Sprague clutch treats you well, they can be fickle about locking onto the shaft sometimes or skipping/slipping from debris in them, but are very strong when well supported and clean.

 

[simon.s]

Now, is there anyone out there to take on Chalo's challenge of fitting a CSK40 to the FC-M815?
That is to boost rider weight rating to some 150+ kgs or DH Elites stomping on their pedals...

Given that the CSK series are already bearing supported, you could probably drop the additional support bearing and just run a CSK40 as-is, no?

 

[liveforphysics]

Now, is there anyone out there to take on Chalo's challenge of fitting a CSK40 to the FC-M815?
That is to boost rider weight rating to some 150+ kgs or DH Elites stomping on their pedals...

Given that the CSK series are already bearing supported, you could probably drop the additional support bearing and just run a CSK40 as-is, no?

I don't think it hurts at all to pamper a sprague clutch with bearing support. They are very reliant on the inner and outer cylinder wall geometry staying very flat and round to function.

 

[bh4801]

I don't think it hurts at all to pamper a sprague clutch with bearing support. They are very reliant on the inner and outer cylinder wall geometry staying very flat and round to function.

This has been discussed here: https://endless-sphere.com/forums/viewtopic.php?f=28&t=47618&p=850836&hilit=csk35+failure#p850836

Re: Strong freewheeling crankset
Postby boostjuice » 05 Feb 2014, 03:16
...The CSK40 has a healthy margin of 'headroom' when used in freewheel crank duties...
...It is possible that a CSK35 would be strong enough when supported with an additional parallel bearing (and thus if elation had included this in their design avoided all of the premature failures that plagued their custom cranksets), but for only a 170g weight saving over the CSK40, I didn't want to risk it.

As I said, there was not enough room for a CSK40 in my setup, both on the axle and within 104 mm BCD chainrings, therefore I took the plunge with a CSK35 + support bearing.
The decisive hint came from STIEBER catalog, here 2 x CSK seem to be supporting each another:

 

[ScooterMan101]

So People would have to buy a 83 mm version of a Shimano Hollowtec II Crankset . ?

Issue with the Polished sections on the Spindle is something I have also been wondering about .

With all this work ,
OR
would it be better for a completely new , " Open Source " Freewheel crankset for Mid-Drive E-Bikes to be made ?

This is where I think the three piece design of the Direct Mount Cranksets like Race Face, FSA, Sram, etc. would be a better way to go.
Since the Shimano is two piece with the spindle pressed fit into the drive side crank arm.

With a whole new design , using 3 piece, there would be more room designed in to accept two bearings to support the freewheel.

Off the shelf square taper crank arms could be bought cheaply , then Machined / Milled to accept a spline pattern , that is a little different than what is being sold now,
so as to not infringe on any patents that are being used now, then just have a whole new spindle made, that spindle will have more length on the drive side, thereby allowing the polished parts that are in line with the bearings to be in the right placement and allowing enough room for a two bearing freewheel as well.

24, mm OD , custom spindles for people using shimano bearings

30 mm OD , custom spindles for people using Direct Mount Bearings / BB 30 bearings / PF Bearings

A word of caution: the FC-M815 fits an 83 mm BB.
My FSR is 73mm BB and the axle length protruding to the left is just enough to correctly fit the left crank arm.
In my setup, the polished axle races fit nicely with BB bearings on both sides - if anyone dared CSK40 with a support bearing, they would quickly find out, the polished axle races will be misaligned relative to the BB cups, therefore BB bearings would run in unpolished, lathed sections of the axle or there would be too little axle protruding to the left to allow left crank arm to be safety fitted.

Is that a problem?
I never tested.