Best freehub engagement design

Back by popular demand, the general all-things Road forum!

Moderator: robbosmans

User avatar
CCCP
Posts: 56
Joined: Wed Nov 23, 2005 5:20 am

by CCCP

I am currently in the process of designing a rear hub, and am investigating the two design options of freehub engagement mechanism...

(Better explained) I am looking at the advantages/disadvantages of the 'standard' 3 pawl/ratchet design, compared to the clutch plate/star ratchet design.

In case my explanation is not thorough enough, the '3 pawl/ratchet' design I am referring to looks like this;-
Image

And for the 'clutch plate/star ratchet' type;-
Image
Image

Could anyone comment on the above?

by Weenie


Visit starbike.com Online Retailer for HighEnd cycling components
Great Prices ✓    Broad Selection ✓    Worldwide Delivery ✓

www.starbike.com



ProTech
Posts: 286
Joined: Sat Nov 26, 2005 1:53 pm
Location: Montreal

by ProTech

DT Star ratchet is very effective,a bit loud and sometimes a skips if the mechanic that services the hubs puts regular grease in there instead of the DT grease.

DT patents are going to stand in your way if you want to go commercial venture with this.

The 3 pawls design is good but the Am Classic 6 pawl tangential system is better by far and requires no spring just a guide plate/toothed ring for the pawls to engage,but yet again patents probably are owned by Am Classic,but I've seen Taiwan manufacturers that have hubs that are similar to it.

Karbona Hubs

User avatar
Boonen
Posts: 2407
Joined: Thu Aug 05, 2004 7:33 am
Location: Europe

by Boonen

Personally I like the DT-style freehub mechanism best. It is very easy to take apart and service, even for people who aren't that experienced with bikemechanics. The engaging mechanism is also very direct, the freehub will not spin much before it engages so there isn't that additional stress to the freehub you get when it engages. (no big 'bang' you sometimes hear coming out of the rear wheel of a mavic wheel)
The three pawl design is pretty straight forward as well, but it takes a bit more work to take it apart and because of the three pawls it engages a tad slower. If you go with the three-pawl design I think the design like the one you pictured with a groove in the pawls that are held by a spring is far better and easier to work with than the old standard with little springs pushing against each pawl. A problem you can get with the three-pawl design that I never had or saw with a DT hub is water that gets inside the free-hub and freezes. I know it isn't supposed to happen but somehow you still see this quite often here in the winter when people ride in the forest and clean their bikes with a powercleaner every weekend.

If I had to choose between the two I'd go with something like the DT-design because of it's servicability and reliability. I would say this also has the best potential to save weight in without compromising strength or durability, but maybe you can come up with something smart for the 3-pawl design as well. I think the simpeler you can make the design the better. Less parts that can break and easier servicability.

User avatar
collideous
Posts: 476
Joined: Tue Nov 29, 2005 5:53 am

by collideous

The DT/Hügi design is fantastic. This is what I used when I built my own rear hub. I bought a Hügi hub at the time, tore it apart, and integrated most of the Hügi star ratchet into my custom hub (Pulstar like straight-pull spokes crossed on one side, straight-pull radially laced spokes on the other). DT provided the spokes as well. Based on the explosion drawing you attached it looks like the DT design has not changed much in the last 12 years. That's when I made my hub.

columbusSLX
Posts: 63
Joined: Tue Oct 12, 2004 5:52 am

by columbusSLX

What about the Shimano system? or the Chris King? or the Phil Wood?

User avatar
CCCP
Posts: 56
Joined: Wed Nov 23, 2005 5:20 am

by CCCP

ProTech wrote:DT Star ratchet is very effective,a bit loud and sometimes a skips if the mechanic that services the hubs puts regular grease in there instead of the DT grease.

Yes, that is one disadvantage of the Hügi; where the friction/resistance between the clutch plates caused by a high-viscosity grease is greater than the spring's tension.. But that is easily remedied - You just have to use the right grease.

ProTech wrote:DT patents are going to stand in your way if you want to go commercial venture with this.

Hub, especially for bicycles and the like
US Patent: 6,588,564

There is the patent assigned to DT Swiss, relative to the Hügi engagement mechanism design which would prevent me from reproducing an alike mechanism in my design. For which reason, I contacted DT Swiss, who have offered to provide all of the Hügi internals for a hub. But, I do not wish to produce something that isn't my own...

ProTech wrote:The 3 pawls design is good but the Am Classic 6 pawl tangential system is better by far and requires no spring just a guide plate/toothed ring for the pawls to engage,but yet again patents probably are owned by Am Classic,but I've seen Taiwan manufacturers that have hubs that are similar to it.

To what I know, there are no patents attached to this design for the reason that it is a universal design.

I am currently considering the option of a 'take' upon this design, further refining it to decrease "engagement time"; the period it takes between a force of the freehub body to engage with the hub body.

ProTech wrote:Karbona Hubs

I certainly hope this was not a recommendation.

Boonen wrote:Personally I like the DT-style freehub mechanism best. It is very easy to take apart and service, even for people who aren't that experienced with bikemechanics. The engaging mechanism is also very direct, the freehub will not spin much before it engages so there isn't that additional stress to the freehub you get when it engages. (no big 'bang' you sometimes hear coming out of the rear wheel of a mavic wheel)
The three pawl design is pretty straight forward as well, but it takes a bit more work to take it apart and because of the three pawls it engages a tad slower. If you go with the three-pawl design I think the design like the one you pictured with a groove in the pawls that are held by a spring is far better and easier to work with than the old standard with little springs pushing against each pawl. A problem you can get with the three-pawl design that I never had or saw with a DT hub is water that gets inside the free-hub and freezes. I know it isn't supposed to happen but somehow you still see this quite often here in the winter when people ride in the forest and clean their bikes with a powercleaner every weekend.


I think as previously spoken by ProTech, a six-pawl design can easily reduce the "bang" issue with many of today's hubs, also there is increased surface area between the pawls and the ratchet (Same Force, More Area, Decreased Pressure).

One issue that I think has been left undiscussed is equal-spacing between the pawls on the 'standard' design. If they are all spaced out evenly, wouldn't this mean that the relative position of each pawl is identical to the tooth of the ratchet throughout? So if they are all fully disengaged, wouldn't this result in a larger required movement in order to engage? I have a thought on how to improve this, but it is too technical...

maxxevv
Posts: 2012
Joined: Wed Feb 09, 2005 4:51 am

by maxxevv

As an engineer myself, here's some food for thought:

Does and engagement mechanism has to have pawls or engagement "teeth" to work ??

User avatar
collideous
Posts: 476
Joined: Tue Nov 29, 2005 5:53 am

by collideous

maxxevv wrote:Does and engagement mechanism has to have pawls or engagement "teeth" to work ??


One way roller bearing or a sprag clutch are other possibilities.

tcr
Posts: 1143
Joined: Wed Mar 09, 2005 2:28 pm
Location: Australia

by tcr

Surely you of all people wouldn't copy someone elses design CCCP?? :shock: :lol:

User avatar
CCCP
Posts: 56
Joined: Wed Nov 23, 2005 5:20 am

by CCCP

tcr wrote:Surely you of all people wouldn't copy someone elses design CCCP?? :shock: :lol:


What did I just say?! :roll:

CCCP wrote:There is the patent assigned to DT Swiss, relative to the Hügi engagement mechanism design which would prevent me from reproducing an alike mechanism in my design. For which reason, I contacted DT Swiss, who have offered to provide all of the Hügi internals for a hub. But, I do not wish to produce something that isn't my own...

User avatar
CCCP
Posts: 56
Joined: Wed Nov 23, 2005 5:20 am

by CCCP

collideous wrote:
maxxevv wrote:Does and engagement mechanism has to have pawls or engagement "teeth" to work ??


One way roller bearing or a sprag clutch are other possibilities.


I'm having troubles figuring out how you could apply a Sprag Clutch to a freehub engagement mechanism. Could you elaborate on this?

ie. How would you fix the sprag to the two parts (Hub Body + Freehub Body)?

http://www.renold.com wrote:A Sprag Clutch is a free-wheel device having an inner race, and an outer race either of which can be the input or output member. The input member can be arranged to drive the output member in a chosen direction and permit the output member to over-run in the same direction.

In general, Sprag Clutches are able to transmit greater torques, within given overall dimensions, than other types of free-wheel device.

There are 3 basic applications for a sprag clutch: overrunning; indexing; backstopping.


What's the difference between a sprag clutch and an overrunning clutch?

User avatar
TunedCannondaleR700
Posts: 338
Joined: Wed May 18, 2005 10:37 am
Location: California

by TunedCannondaleR700

So when would we be able to buy the design you are working on now.
Cannondale is quite simply the best

User avatar
Boonen
Posts: 2407
Joined: Thu Aug 05, 2004 7:33 am
Location: Europe

by Boonen

CCCP wrote:One issue that I think has been left undiscussed is equal-spacing between the pawls on the 'standard' design. If they are all spaced out evenly, wouldn't this mean that the relative position of each pawl is identical to the tooth of the ratchet throughout? So if they are all fully disengaged, wouldn't this result in a larger required movement in order to engage? I have a thought on how to improve this, but it is too technical...


If you take the standard design and make the pawls unevenly spaced this will either make no difference at all or make the 'bang' problem (sorry, don't know a better description) even worse. The pawls have to engage in the tooth inside the hub. If you space the pawls unevenly with the idea of making the gap to the next engage smaller not all the pawls will be able to engage at the same time, making the mechanism much weaker and reducing the risk of the pawls skipping over the teeth. If you want to do something like this you would have to make multiple rows of pawls, but the only advantage would be faster engagement, the skipping of the pawls would be the same if you'd only use three pawls. (you could ofcourse also use 4 pawls like Tune does) Ofcourse multiple rows of pawls would only make the system heavier and more complicated.
I would be interested to hear about your ideas of improving this, but if it's a good idea you want to use you might not want to tell anybody before you get an octrooi or even a patent. (If you tell people I believe this would rule out the octrooi option.)

ProTech
Posts: 286
Joined: Sat Nov 26, 2005 1:53 pm
Location: Montreal

by ProTech

ProTech wrote:Karbona Hubs

CCCP wrote:I certainly hope this was not a recommendation.


My link to Karbona was only to show the Am Classic style hub are widely available,hence the possibility of manufacturing one with that style of pawl engagement.

CCCP wrote:I think as previously spoken by ProTech, a six-pawl design can easily reduce the "bang" issue with many of today's hubs, also there is increased surface area between the pawls and the ratchet (Same Force, More Area, Decreased Pressure).


Yes it does feel like the 6 pawl system makes the engament is a bit less harsh than a regular 2/3 pawl unit and feels more "positive" at the same time.

CCCP wrote:One issue that I think has been left undiscussed is equal-spacing between the pawls on the 'standard' design. If they are all spaced out evenly, wouldn't this mean that the relative position of each pawl is identical to the tooth of the ratchet throughout? So if they are all fully disengaged, wouldn't this result in a larger required movement in order to engage? I have a thought on how to improve this, but it is too technical...


This is something I have noticed with the Am Classic hubs it does take a relatively long cassette shell travel from freewheeling before the pawls engage,it's similar to a regular pawl system, but it should be faster because of the 6 pawl tangential design.
My guess is that they have not made the angle of the guide plate channels that engage the pawls shallow enough,the pawl travel must be at upwards angle a bit too much.

by Weenie


Visit starbike.com Online Retailer for HighEnd cycling components
Great Prices ✓    Broad Selection ✓    Worldwide Delivery ✓

www.starbike.com



ProTech
Posts: 286
Joined: Sat Nov 26, 2005 1:53 pm
Location: Montreal

by ProTech

Boonen wrote:If you take the standard design and make the pawls unevenly spaced this will either make no difference at all or make the 'bang' problem (sorry, don't know a better description) even worse. The pawls have to engage in the tooth inside the hub. If you space the pawls unevenly with the idea of making the gap to the next engage smaller not all the pawls will be able to engage at the same time, making the mechanism much weaker


This is the exact problem shimano hub bodies have; two pawls on a "V" like sequential set-up near each other nothing on the other end of the shell.

This system sometimes skips because if load is being applied to the other end of the cassette shell after feewheeling,the play *normal* then makes the first pawl skip,thankfully the second one engages but not before a relatively long cassette body tarvel. This has been the design flaw of all Shimano hub shell,it only gets worse or sometimes starts to appear as the hub gets older.
The same can issue be found on some 3 pawl "Y" system.

A 4 pawl unit *Tune style* arranged in a " "X" type has much less chance of skipping.

Post Reply