Pivoting
The elevator has two stub axle MAXSplines that hold it in the pivot. The axles rotate with the elevator so that the CANCoder attachment on the outside can be rotated.
Rigging
The rigging is a belt from the base stage to the middle stage, and a Dyneema loop from the middle stage to the inner stage. The Dyneema is tensioned on a spool held by a WCP ratchet plate. We can use a ½” wrench on the hex on the outside to easily tighten the spool.
The elevator is counterbalanced by two 6 lb constant-force springs, which help offset the weight of the elevator in near-vertical positions, but overpower motor brake mode at lower angles. Once the full mass at the end of the elevator is on, we will have to re-evaluate the strength of the springs and whether we will need two.
Wiring
The elevator is driven by two Kraken X60s, using the ReFire adapter boards and custom-length CAN Molex SL connections.
We use the ReFire boards for clean, rigidly attached wiring, the convenience of Molex and PowerPoles, and the protection of the CAN wires next to the motor, as opposed to the official PowerPole Adapter Board. In most of the places we use these boards, servicing the screw terminals would require removing the motor from the mechanism or disassembling the mechanism.
All the wires from the elevator and attached mechanisms go through the cable chain shown here, down to the brainpan, and into the underside of the robot.
A side effect of the CAN cable protector is that it’s hard to see the CAN wire, but it is noticeable enough for maintenance up close. On one side of the base stage, wires and two CAN connections go halfway up to our climb motor:
On the other side, power and CAN wires for the hand and wrist motors go through the base stage, the cable chain, and the length of the inner stage.
With ten power wires plus CAN going to various places on the elevator, it was important to do this wiring with care, and especially important to do it before mounting the elevator on the robot. We chose not to label the wires before running them through the cable chain. Instead, we used a multimeter to check continuity after the wiring harness was assembled so that we knew which wires coming out of the bottom cable chain should go to the same PDH slot.
Second Iteration
After assembling the elevator with the sprockets right up against the side plates, we discovered that the width of the pivot it had to fit into was a little tight, especially with chain master links which were wider than the sprocket. Also, we didn’t really need two chain runs. So in a second quick iteration only shown in full robot pictures, we took off one sprocket and spaced the other off the side plate slightly. This shifts our elevator laterally by 3/16”, which is small enough to not affect our understanding of the robot as symmetric.