#robotics Logs

Mar 18 2019

03:02 AM rue_bed2: --
08:34 AM pokmo: hi
08:34 AM pokmo: anyone know why the rocker-bogie mechanism can be used to climb stairs? https://hackaday.com/2017/08/12/simple-step-climbing-robot-climbs-like-its-on-mars/
08:35 AM pokmo: how does the horizontal velocity get converted into vertical?
08:36 AM deshipu: by rotating the rocker, obviously
08:37 AM pokmo: but the rocker doesn't actively rotate
08:37 AM pokmo: only the wheels rotate
08:37 AM deshipu: pushing on the wheels makes them push on the rocker, which then rotates and pushes the wheels up
08:38 AM deshipu: it only works when you have all 6 wheels powered
08:39 AM deshipu: also, in this particular video, the wheels actually climb on the vertical part of the step
08:39 AM pokmo: and i guess, the front wheels must have enough grip to push the rocker up
08:40 AM deshipu: well, they are pushed against it by the other wheels, so the grip doesn't have to be that good
08:40 AM deshipu: as long as there is at least some grip or if it's not perfectly vertical, they will work
08:41 AM pokmo: this mechanism will fail if the stairs have overhangs, right
08:41 AM deshipu: it might
08:42 AM deshipu: depending on how large the overhangs are
08:45 AM pokmo: even a fairly small overhang, it should be enough to jam the rocker
08:47 AM pokmo: but something like this also gets jammed, right https://www.dhresource.com/0x0s/f2-albu-g4-M01-78-B4-rBVaEVmkLaGAQN-2AAIBhK_7tmk572.jpg/190kg-carbon-steel-portable-six-wheeled-stair.jpg
09:15 AM rue_mohr: " They took the extruder off the robot and put it on a gantry they had at Oakridge that had been used for a couple of other projects previously—one was actually 3D printing concrete—and once they did, they were very quickly making pretty good parts.”"
09:16 AM rue_mohr: that robot wildmage was working on, what uni was it at?
09:29 AM deshipu: pokmo: any mechanism can be jammed with a terrain specifically designed to jam it
09:30 AM deshipu: pokmo: the trick is to know what kind of terrain you expect and design the mechanism such that it can handle it
10:31 AM pokmo: deshipu got it
10:55 AM s34n: what are the two red components in the middle of this board? https://images-na.ssl-images-amazon.com/images/I/6123noI7yHL._SL1010_.jpg
11:01 AM deshipu: s34n: if I had to guess, I would say capacitors
11:01 AM deshipu: especially since you can't see any other capacitors on there
11:01 AM s34n: ok. I can identify round capacitors. but I wondered about these
11:02 AM deshipu: would be easier if you had a schematic of the connections
11:02 AM s34n: I've see very similar components in appearance a lot
11:02 AM deshipu: and if you told us what the device is
11:02 AM deshipu: the thing is, any device can be made to look like that
11:02 AM s34n: it's a motor controller with manual potentiometer
11:03 AM deshipu: them yes, they are probably capacitors for filtering the motor's noise
11:04 AM deshipu: why do you need to know?
11:04 AM s34n: I'm trying to learn enough that something this simple makes sense to me
11:05 AM deshipu: draw a schematic of it
11:06 AM s34n: I can't yet. I don't know why it is arranged with the resistor between the "capacitors", for instance
11:07 AM deshipu: why can't you? just look at the connections and draw it
11:08 AM deshipu: it's easy with a single-sided PCB
11:08 AM s34n: I can't see the traces, so I have to guess at the connections
11:09 AM deshipu: why can't you see them?
11:09 AM deshipu: they should be clearly visible, there is nothing at the bottom of the board
11:09 AM s34n: they don't show up in this photo.
11:09 AM s34n: But I just found a photo of the bottom, with the traces
11:10 AM deshipu: they surely show up on the physical device
11:10 AM s34n: I don't have the physical device
11:12 AM deshipu: then why would you insist on reverse-engineering this particular device, instead of looking for a schematic of a motor driver instead, possibly with an explanation already?