Quilting machine carriage brakes (channel locks)

My wife recently acquired a quilting machine, which is basically a sewing machine head on a carriage which moves freely in two dimensions. The operator can move it around to sew in any imaginable quilting design. But there are times when one needs to "baste," which is stitching in a straight line to hold the quilt together before the fancy stitching starts. In that case it's desirable to restrict the carriage motion to the X dimension by preventing it from moving in the Y dimension. There are no brakes on the carriage, so the company sells what they call "channel locks" that clip over one or more wheels. That sounded like something I could print for far less than the $12 a set that they cost on line.

Here's the finished product, so you can see what I'm talking about.








This project took quite a bit of trial end error, on two fronts.

First, I thought that the "channel locks" worked by acting like chocks, wedging under the leading and trailing edges of the wheels. I had seen a pair in real life, and seen pics on line, so I had an idea of what they should look like but not exactly what the important characteristics were. We knew that "clingy" plastic was supposed to be better than slick plastic. So I figured I'd print them in Nylon, which is more resilient than PLA. As it turns out, they don't work as chocks, but rather as brakes grabbing onto the rims of the wheels, and the ends that rest on the track stop the wheels from turning. So the important aspect is to exert spring pressure inward onto the wheels. Nylon is too flexible and doesn't exert any pressure.

So it was back to PLA, which is more rigid. The inside diameter needs to be just right so it will touch at several points with sufficient pressure. That explains the little "bars" on the inside: drum brakes are hard to get to fit correctly, but "caliper" brakes touching at just a few points are easier to fit. The interplay of inside diameter and total arc are critical: when it's installed, the ends of the arc need to be just long enough to touch the track without any play.

The second aspect of trial and error was making the design fully parametric. This means having every dimension based on a parameter (variable) so you don't need to redraw anything to make changes:

Parametric design sounds easy, but I've found that the sequence of "what is built on what" can make or break such a design. In this case I needed to build the arcs onto each other, and make the arc dependent on the inside diameter which matches the inside points of the bars (the "caliper" touch points) etc. Working with Fusion 360, I found that I needed to add fillets at the 3D body level, not at the 2D sketch level, or any angle adjustment broke them. I don't fully understand that relationship, but it worked. I got in the habit of testing every minor design change by altering the basic inside diameter parameter up and down and checking whether the final diagram behaved correctly.

This process took about 10 cycles of measure-resize-print-test before I got it right, 5 in Nylon and 5 in PLA. That's a lot more than I expected, but at about 20 cents per part, it still cost only about $2 or so. Each iteration took about 20 minutes to print, so I could just do it in and around other stuff on this long weekend. And if they ever break or wear out I can just print more. With two wheels locked, I hope the carriage is held rigidly enough. If the brake is not grabby enough, I might put some Shoe Goo on the caliper touch points.