Accurately Measuring Cable Pull

I promissed some bike subjects again, and you can't get too much geekier than this one.

I'm (slowly) working on converting my Bike Friday Tikit to use a SRAM i-Motion 9 internal hub.  This bike will have drop bars and I want to convert a Shimano 9sp barend shifter to work with the hub.  To do so I needed a way to measure cable pull very accurately.

I built two devices to do this, but I only took photos of the second and far more successful one:

Cable pull measuring device

And here is another shot of the back of it:

This was really easy to make.  I used a digital scale that came off of the Digital Readout for my mill.  These scales are just like digital calipers and can measure how far the slider in the center moves up and down along the ruler.  I drilled and tapped a hole in a block of aluminum that fit a cable housing stop that I had lying around.  The block is bolted onto one end of the scale.  The center (sliding portion) of the scale has a cable clamp.  The other side of the clamp connects a spring to the far end of the scale.

Twist the shifter (or move the lever of a barend shifter) and the scale moves and reports how far it went.  These digital scales read out in millimeters or inches and let you reset them anywhere, so it is trivial to measure the motion per click.

I did that and generated this spreadsheet (yeah, it is lame to embed this as an image):

A Travel Agent is simply a cable pull amplifier using two concentric pulleys.  It looks like this (image borrowed from Amazon):

The stock pulley amplifies cable pull with a 2:1 ratio.  To make the i9 work with the Shimano shifter I need to make a new pulley with a 1:3.6 ratio.  You can't fit a much larger pulley on the Travel Agent, so that means a 32mm outer pulley to 8.9mm inner pulley.  That isn't a hard item to make on the lathe.

I hope to report back on how the whole system functions soon.  I still need to build the wheel and install it into the bike.

Published 09 October 2008 04:18 PM by AlexWetmore
Filed under: , ,

Comments

# John Speare said on 09 October, 2008 09:42 AM

Tis true, you are king of the bike geeks Alex. Do you think cable pull be in the same ballpark for all SRAM hubs? Reading ibob, kogs, and other bike lists, there seems to be some desire online for bar-end-shifter-internal gear hub shifters. Furhter, it seems like a guy could probably get away with selling these pretty risk free. Just sayin.

# AlexWetmore said on 09 October, 2008 10:01 AM

If the concept works well then I'm open to making the adapters for other hubs.  I'd want to measure shifters directly (and hope to do this for the Nexus 8sp hub soon).

# Jeremy said on 09 October, 2008 10:30 AM

I would buy an adapter to make my i9 work without that horrible shifter in a heartbeat.

Now, make a dual cabled barcon for the rohloff!

# Jimmy Livengood said on 09 October, 2008 10:33 AM

Ahh,  This is truly a bike-geek paradise!

Other than funds, I've hesitated on buying an Alfine hub because I want to use STI shifters with it. I emailed Jtek about it and they said that cable pulls for Shimano internal hubs were consistent per shift, but that STI shifters actually varied the amount of cable pulled depending on which shift you were completing. -(this was about a year ago)

I'm curious if this is true, and if it is if it's enough difference to matter, as I imagine there must be enough tolerance in the system for dirt and cable contamination, too. Plus, to the eye my down tube shifters don't seem to have any eccentricity to the cableway -never measured it, though.

I know Hiawatha Cyclery has been using 8sp bar-end shifters with a travel agent with a Shimano 8sp Alfine hub for a bit now. see:

www.flickr.com/.../2347659679  

I haven't heard anythign about the long-term durability or effect on the hub. You'll certainly have more than a few folks asking for adapters when you get this to work -both iBob types and MTB types.

# AlexWetmore said on 09 October, 2008 10:40 AM

I only measured a lot of variation at the end of the range.  On a derailleur system you use the limit screws on the derailleur there to control indexing, not the clicks in the shifter.  I think the same is true on an internal hub (the shift linkage can't over or undershift the range), but I don't know for sure.

The other place where I see a anomoly is in the second gear, where the shift was only 1.9mm vs around 2.4mm on average.  I will measure another shifter to see if this is consistent or just wear in my shifter.

alex

# AlexWetmore said on 09 October, 2008 10:43 AM

Jeremy -- The Rohloff has huge amounts of cable pull.  I think building a barend that looks decent would be hard, but I might build my own downtube shifter for it at some point.

# Jason said on 09 October, 2008 10:45 AM

Feel free to make a few extras and send them down to Oakland for testing :)

# Tarik said on 09 October, 2008 08:16 PM

Wow, well done alex. I had similar, but  much less elegant thing in mind, but I found a listing of SA cable pulls once that got me off the path. I can't remember what my plan was exactly, nor where I found the listing, so, uh, nevermind. In conclusion. Excellent work!

# AlexWetmore said on 10 October, 2008 06:59 AM

I should take a photo of my first attempt.  It was an ugly thing with sliding pieces of brass and a dial indicator.  Lots more work, less accurate, and much more of a pain to use.

# Rory said on 10 October, 2008 10:24 AM

do you want to measure with a campy-brifter? while I'll say the benefit being using a modern brifter that also has a brake opening thing, but I'm actually more selfish-ly interested for maybe building up another internal geared bike.

# Matt Newlin said on 10 October, 2008 05:34 PM

Pretty cool.  Have you considered a linear travel agent:  a lever in the middle of a long span of cable, like on the down tube?  That way, it would be easy to make the ratio adjustable.  Just make both cable pivots on the same radial line from the lever pivot.  Kinda ugly though.

I suppose you could just turn down one pulley or the other to adjust the ratio too.  Do you know the effective radius of the cable on the pulley?  I'd guess a little larger than the distance to the center of the cable.

# AlexWetmore said on 11 October, 2008 11:56 PM

Rory -- I'm up for measuring any shifters that folks in Seattle want to bring over.  It doesn't take much time and the tool isn't going anywhere (the scale might go back on my mill, but it's easy to reconfigure).

Matt -- I think the effective radius is the depth of the root of the pulley plus the radius of the cable.  I've considered a linear travel agent (like the Sidetrak BPB), but modifying them isn't that much easier.  The pulley's aren't too hard to make on the lathe.

# Justin Miller said on 26 October, 2008 02:01 AM

Alex, if it works it will be a big improvement over the various shifter cludges out there (e.g., hubub's and  drop bars that come apart).  I wonder if the folks at Firebikes did similar research to develop their downtube shifters for Nexus hubs.  www.prestosell.com/.../order.pl

Justin

# Greg said on 26 October, 2008 11:03 AM

I can't tell from looking, but is it necessary to have a full cable housing run going to and exiting the Travel Agent? It would be nice to be able to mount it under the bb shell getting it out of sight and allowing the use of a downtube shifter if you wanted.

# AlexWetmore said on 26 October, 2008 01:52 PM

Greg -- The Travel Agent is designed for housing on both ends.  You might be able to mount it in the final loop of housing that goes to the hub.  The other option would be making a housing stop that fits under the bottom bracket (this would be a fairly easy part for me to make).

Justin -- It's hard for me to tell if their shifter is indexed at all, or just a fancy looking friction shifter.  It says one shifter is for the 3, 4, 7, or 8sp Nexus hubs.  I doubt that the cable pull is identical for all of them (but maybe I'm wrong).

alex

Search

Go

This Blog

Syndication