First, after doing some analysis, it appeared that the 3mm pitch 6mm wide belt was unlikely to be able to carry the full torque from the motors. So I’ve switched to a 5mm pitch 15mm wide belt, which while still unable to carry the full torque indefinitely is only a factor of 2 or 3 off instead of a factor of 20 off. Secondly, I added a bearing opposite the upper pulley so that it is supported from both sides. The recommended belt tension for this belt works out to something like 120lb, which is a fair amount of cantilevering, even over the 16mm wide pulley. The updated CAD looks like:
And the newly added bearing can be seen in this section view:
I did a first test print of all these parts and put them together. While there were a few tweaks necessary for the second revision, it looks like this leg is probably usable.
Another of the failure modes observed during the 2019 Maker Faire was in my quickly slapped together leg design. The shoulder joint was required to squeeze two motors together against a strongly tensioned belt, using nothing but a relatively thin section of printed plastic. This caused it to deform, leading to belt tooth skipping, and then eventually to fail, leading to delamination of the shoulder joint.
My plan to resolve this is to switch to a leg design where the upper and lower leg are in series rather than opposing one another. This is more like the Mini-Cheetah design from Ben Katz. This has the benefit of getting the leg out to the side, so the upper leg is free to rotate 360 degrees, only limited by cable harnessing. As seems to be my pattern, I’ll try making something out of 3d printed PETG first, optimize it some, and if I fail there, switch to metal. Here’s a render of the current CAD:
Eric from CireRobotics helpfully pointed out that I’m way over the design limit for the 6mm Gates belt I was using, so I’ll also be trying to bump up to a beefier belt in this iteration.