Prototype chain catcher

[danl1]

Inspired by the 'Dropping the chain on a compact downshift' thread and a problem with my wife's bike, I cooked up this little chain watcher. My wife's bike was an inspiration because it's a carbon Trek 5000, and the glob o' epoxy they call a BB is so huge that none of the commercial chain keepers will fit it.

Ok, so on a properly maintained and conservatively shifted bike, chain keepers are almost never necessary. Still, stuff happens, bikes fall down, you get caught in a bad break and need a stupid shift... Rather than grinding down expensive frame parts or missing a key shift in a race, a chain catcher is cheap insurance.

I made it out of a 5" piece of aluminum arrowshaft. After cutting it with a tubing cutter, I bent it to an approx. 10" radius. This radius allows it to clear the front of the FD cage, swoop down to the side of the ring to do it's job, then swoop back inward, so that should the chain somehow get past it, it won't keep it from being shifted back on (a fairly common complaint with the Third Eye and Deda units.) The curve also allows it to be aside the little ring just where the chain is starting to engage the teeth of the ring - where it can do the most good. Depending on fit, the fang style units sometimes are rotated further forward than ideal.

A note on tubing and bending: If you happen not to know an archer, you can get an aluminum practice arrow at the local sporting goods big-box for $2. I had it lying around, but it's still a cheap source of raw material. Thin-wall tubes would rather kink than bend, but if you first fill them with sand or table salt and tape the ends shut, they'll behave much better. A mandrel to bend them around is better still, but that's too much like work for this endeavor.

After the tube was bent, I held it in position to determine the angle to flatten to go under the FD fixing bolt. To make the curve do all the right things, it's not on the same axis as the rings - it's normal axis is offset to the rear, so that the bend goes inward and down as it nears the bottom bracket. Flatten the appropriate angle with some vice-grips, drill a hole for the FD clamp bolt to pass through, and it's done. Time: Less than to describe it. Weight: 3 grams.

Here's a lousy picture, that should be more than worth the above words:



[OK, so it's not my wife's bike. I'll build her one some day - this is the prototype. And yes, the bike's a mess.]

I purposely detuned the FD and shifted as badly as I could. There are a few grease marks on it, which I didn't get on the workstand, so it appears to be doing the job.

[scubad]

Please take a picture from headon to see the bends and how you finished the end.

Thanks.

ScubaD

[cydswipe]

You gotta get better pics. It's too dark to see everything.

[roadfix]

I think one TDF team mechanic fabricated something similar a couple of years back for his team's fleet....

[danl1]

Originally Posted by roadfix

I think one TDF team mechanic fabricated something similar a couple of years back for his team's fleet....

Yep. What set me thinking was this version on David Millar's TT bike this year:

[danl1]

Originally Posted by cydswipe

You gotta get better pics. It's too dark to see everything.

No kidding. My little digi has a crappy flash, and I can't get enough light into it's puny lens to go natural. I could wait for a sunny day and head outside, or burn and scan some real film through the SLR, but I'm not feeling all artsy right now.

[danl1]

Originally Posted by scubad

Please take a picture from headon to see the bends and how you finished the end.

Thanks.

ScubaD

Here's what I have - another lousy pic, I'm afraid. In this picture, the chain is on the small ring. The large ring is tough to see as it's a dead-on shot. This thing is surprisingly difficult to find a good angle to shoot.

The end is unfinished - just cut off with a tubing cutter and deburred. Despite how it appears from any given angle, there is only one smooth, single plane bend in the tubing. It is the angle of the flat that makes it appear to curve both front-to-back and out-to-in as you go from the mount down.

[jamesau]

Originally Posted by danl1

To make the curve do all the right things, it's not on the same axis as the rings - it's normal axis is offset to the rear, so that the bend goes inward and down as it nears the bottom bracket.

Can't quite picture what you're trying to say.

The one difficulty that seems evident with your prototype: the whole arrangement will tend to pivot around the fixing bolt unless the bottom of the deflector is braced, say at the bottom of the seat tube. Notice Millar's arrangement is braced against the bracket to keep the deflector from shifting once force is applied to it.

Let us know how it works and if you make any improvements. Thanks for sharing. (I use a jumpstop when I can but, increasingly, modern frames do not permit it.)

[danl1]

Originally Posted by jamesau

Can't quite picture what you're trying to say.

The one difficulty that seems evident with your prototype: the whole arrangement will tend to pivot around the fixing bolt unless the bottom of the deflector is braced, say at the bottom of the seat tube. Notice Millar's arrangement is braced against the bracket to keep the deflector from shifting once force is applied to it.

Let us know how it works and if you make any improvements. Thanks for sharing. (I use a jumpstop when I can but, increasingly, modern frames do not permit it.)

It is tough to explain, so let me try this way: if the bend were on the same axis as the rings, it would appear flat when viewed from the front. If it were on an axis perpendicular to the rings, it would appear flat when viewed from the side, but curve over in front of the seat tube. But since it's off-axis, it appears flat when viewed from somewhere right-front of the bike, and the curve (if extended) would carry it behind the seat tube and into the back wheel. That way, the curve can both move left-right so that it properly aligns closely to the chainring, and front-back such that it can align on the proper spot on the ring.

You are right to worry about lateral movement - I had the same concern. Mine does essentially the same thing as Millar's to some extent - the flat portion is hard up against the hanger, and it won't move a bit. Bending may be another matter, but this thing is surprisingly stiff. Some sort of modification would want to be made for a braze-on hanger - something to shim between the gadget and the seat tube, or continuing the bottom bend so that it braced against the BB or seat tube.

Another factor is that the Millar design isn't ideal - by moving straight down along the seat tube, it's interaction only occurs after the chain has engaged (or missed) the chainring. If the chain derails, you have three pieces of metal fighting for the same space, and something has to bend. That design hopes that it's the chainring that bends, so creating a ramp and realignment spring effect that will pop the chain upward, or that the chain will still be riding high on the tooth tips, and so will take a nudge.

In contrast, mine intersects the chain below the FD cage, right at the point where the teeth are starting to engage the chain. The bend is such that a misaligned chain will naturally slide down the tube onto the ring. Instead of needing a clear gap to ensure that things don't jam up, mine is only a mm or so from the chain in the small/big combination. Because of both this lateral and fore-aft placement, a misbehaving chain has far less force on it as compared to the further forward placement that occurs with the Millar treatment or the 'Fang' style watchers (depending on clearance). It's the same set of reasons the Jumpstop is so good - the angled part of the plate helps nudge the chain on, and the design moves the effective contact point back to where the chain first engages the ring.

Sadly, I won't be getting rich off of this. Its design makes it a one-off sort of thing. While this one works perfectly on my bike, it won't fit at all on my wife's, for example. But it works great - I've been riding with my FD seriously detuned for a week, and have intentionally been making simultaneous and crosschain shifts. I have a fair bit of grease streaking on the catcher, indicating that it's doing it's job. I've never felt a skip, grind, or jam as was occasionally noticed when the Third Eye was on the job, even though it's results were nearly perfect for me, too.

To my eye, it seems to blend in with the mech well, not appearing obvious as a band around the seat tube can. That might be an 'eye of the beholder' thing, tho. I'm going to keep using it, and will report back if I see any problems.

[cmg]

sounds like a great solution. if i could see it. didn't think of using a arrow shaft, thanks for that.

[jamesau]

Originally Posted by danl1

It is tough to explain, so let me try this way...

I get it. Cool. Thanks for explaining. The rest of this post is just a bit of a ramble...

Possible derailment concerns me when the clearance between the small ring and chainstay cannot easily accommodate the chain; the chain can wedge itself in there and damage the stay rather than just dropping harmlessly in the slot. On most bikes this is probably most evident with a triple crank. My Stumpjumer suffered a gouged stay and bent chain links; that was my hard-learned lesson. Since then, I'm more concerned with proper FD alignment/adjustment, not shifting the RD and FD simultaneously, and use of chainwatchers.

My Look 481 can't accommodate available chainwatchers, so I wrap the vulnerable portion of the chainstay with a section of tough protective tape. It hasn't been challenged yet (knock on wood).

My LS Ultimate has plenty of clearance so I don't worry about it; there have been a few occasions when the chain has derailed and dropped harmlessly to the bottom bracket.

All other worthy bikes that can accommodate a Jumpstop, have one.

[SkiRacer55]

Originally Posted by danl1

Yep. What set me thinking was this version on David Millar's TT bike this year:

...I'll be happy to buy them. I guess I really don't understand why Deda and all those other people can't figure out that people like us with fat, ovalized carbon down tubes can't get with a Dog Fang or something like that...

[Cocolo]

Here is the one I just finished for my SuperSix. Is made from an aluminum Specialized bottle cage that I had lying around. It weighs all of 4.9 grams and seems to do the job just fine. I used my Dremel tool to cut the cage and I shaped it with a grinding bit. Picture quality (mobile phone) is not the greatest but you get the idea.

[danl1]

Originally Posted by Cocolo

Here is the one I just finished for my SuperSix. Is made from an aluminum Specialized bottle cage that I had lying around. It weighs all of 4.9 grams and seems to do the job just fine. I used my Dremel tool to cut the cage and I shaped it with a grinding bit. Picture quality (mobile phone) is not the greatest but you get the idea.

Great idea.

[DrSmile]

Hrm, I just slap on a Deda DogFang and call it a day. True it's plastic, but it sure works.

[danl1]

Originally Posted by DrSmile

Hrm, I just slap on a Deda DogFang and call it a day. True it's plastic, but it sure works.

I've used them here and there, too, or rather Third Eye's version. They won't fit on a lot of the new carbon wondersteeds, tho.