Click to Visit nothinglabs.com           

Wednesday, July 3, 2013

3D Printed Bike Shifter

I'm doing my first big organized bike ride in a couple weeks - the STP (Seattle to Portland). I was thinking it might be fun to do a 3D printed part or two for my bike. I had originally considered a 3D printed belt-drive - but the more I looked at it - the less realistic it seemed.

A more realistic idea was a shifter. Specifically - a friction shifter. The first version took an afternoon to design / print - and surprisingly worked pretty well.

While friction shifters don't "click" into gear like index shifters - they have several advantages.

A slightly bent derailer or worn cables / housings will cause an index shifter to miss shifts (resulting in that click that won't go away). A friction shifter is much more tolerant of minor problems like these. This is why shifters on touring bikes have a friction mode.

My bike is a 1x7 (no front derailer) - so it's been pretty well tested at 7 speeds. This shifter should work for front and/or rear derailers - and pretty much any number of cogs.

The files for this project are published on Thingiverse. See there for printing tips.

The shifter takes about 1.5 hours to print - and only needs a few non-printed parts:

2x #10 Nylon Lock Nuts
2x #10 x 1" Bolts
3x #8 Washers (yup - they fit)
The screw works as a friction adjustment and allows for a large amount of wear before anything needs to be replaced.

I've put about 150 miles on this shifter design (about 80 so far on the last version) - with very few signs of wear.

Expect a few friction adjustments in the first mile or two - then again around mile 50 (probably less frequently after that).

How long will it last? I have no idea - but I suspect a good while - maybe 1000's of miles?

I've done a fair amount of "hardening" to the design and should hopefully stand up to normal use.

Of course - if you're really concerned about reliability - you can always print a backup.

9 comments:

  1. Nice work, but you spent a whole afternoon to produce one piece of something with 3D printer. I think this is just a waste of time and expensive machinery when you can make that same functionality with just warming up a cup of water and use Polymorph or similar hand moldable plastic.

    Check this out:
    https://www.youtube.com/watch?v=oSQaeOod-LQ

    ReplyDelete
    Replies
    1. that's certainly a useful way to make things if you've already 3d printed / milled / somehow otherwise made a mold first. so - the process isn't really comparable.

      Delete
    2. Dude,
      Did you see the forming temperature - 60C. That's like way too hot for fingers. What kinds of toxins are being put on the guy's fingers. Besides, you can't get it consistent unless you make a mold (using 3D CAD techniques). In that case, you might as well use a 3D printer. I wonder how well it holds up to being in the sun.

      The Polymorph is probably best for one offs like Sugru, but not for something that needs tight tolerances and strenght like the shifter. Everything has its place.

      Just saying.

      Mike

      Delete
  2. Excellent! Reblogged to www.cat6racer.com

    ReplyDelete
  3. I'm surprised that for the one aspect of the design that was going to require regular adjustment, you didn't use a method that didn't require another tool, perhaps a thumbscrew in shifter 2.0?

    ReplyDelete
    Replies
    1. it's unclear how much adjustment will be needed in the long-run. need more data / testing.

      I have been working on some ideas to add a thumbscrew (either from the hardware store - or printed).

      I'll definitely update here / thingiverse if/when I make any tweaks.

      -Rich

      Delete
  4. First of all, great work!!!

    >I had originally considered a 3D printed belt-drive - but the more I looked at it - the less realistic it seemed.

    I wanted to make a 3D printed belt drive for my fixie, mind sharing why you abandoned that idea?

    ReplyDelete
    Replies
    1. I didn't get too far - but the main major complication I hit upfront was that a solid belt would require I be able to separate the rear triangle.

      Considered doing a belt with a split - but that hit me as too complicated / prone to compromise strength.

      I will say that I suspect soft pla (see http://makergeeks.com/flsofi.html) might be strong enough to pull off a rubber-band style drive.

      Nylon (same page) might be an option for the pullies.

      -Rich

      Delete