home..

Week 05 - component demonstration

group page // repo source files // objectives

Contents

meta >

claw >

design changes >

cut new fingers that shouldn’t physically interfere w/ cam or neighboring tiles. However, noted that one of the fingers seems rather tight relative to the other 3; this significantly impacts motion to the point of not closing on the tile properly. wondering if tolerance could be improved; create more clearance for the fingers but maybe use washers to constrain.

since we acquired additional mk4’s, decided to run some test prints on them and impressed w/ the quality. noted that the sirayatech craft ultra-clear is useful for troubleshooting, but appears to have high surface energy which makes the surface too grippy sometimes, possibly creating friction and preventing motion. using the mk4 fdm parts for this new rev of gripper.

made cam path less aggressive (radius travel is half of previous, 3 -> 1.5mm), increased size of compliant tension feature from 1 -> 2mm, and decreased width of travel path by 0.2mm (?).

appears to still get stuck in the center of travel, and one of the fingers are still less mobile than the others.

also adjusted the servo mount point because mk4 appears to have better tolerance and because of that joint is now slipfit.

however, I think the “improvements” didn’t work, so reverting to old design but changing dimensions to be tighter.

the reverted design is a very snug and secure fit. additionally, relaxed some of the cam design.

demo >

o_power pcb eval >

there is some visible variation between slots, but tested an existing h and noted a reasonably good mate. not sure if good enough electrical mate however.

thought I saw some shorts, but upon dmm testing the macro board noted no shorts between power and ground and signal. not sure about between signals, but I believe I left all of them floating and unrouted. fangzheng’s auto probe would be useful for verification, even at the tile verification step. ofc, can have factory test and assemble for us as well.

coincidentally, 8x8 macro tiles are about the size of a voxel, maybe between ben-geometry and miana-geometry.

soldered using lead-free solder + flux; it was easy enough.

need to fab some more h’s to create a demo circuit; the last h pattern used appeared sufficient.

danke >

began assembly of danke; unfortunately messed up orientation a couple of times (upside down pem base, backwards rails, socket instead of flat screws etc.) so took longer than expected.

also had a preload mismatch for linear rails; only 3 of 4 carriages moved easily on mgn9 rails, made do with the 4th one but waiting on a new rail to arrive.

shared interesting machine vendors w jake and mtm discord (eg dfh.fm, west3d.com, etc.).

jake noted a printed part was breaking due to print orientation and issued a hot fix.

nathan noted belt was rubbing on some plastic and CAD is getting corrected.

these are “batch 1” machines, so need to be wary of bugs.

lumenpnp >

tray >

printing additional sla 1x8 trays and fdm 1x8 trays (mk4).

© 2024 Alan Han   •  Powered by Soopr   •  Theme  Moonwalk