diff --git a/README.md b/README.md index 79b277d42b163f14b41341a9bfe619a3242363b3..c6ba7d569559ab57aa3984ac316b78d4f9e92045 100644 --- a/README.md +++ b/README.md @@ -8,6 +8,8 @@ Machines described here are *parametric configurations of object-oriented hardwa Machine controllers are *networked collections of input and output devices* that contain bare minimum state - high level planning and interface takes place within *virtual machine controllers* that are similarly easy to assemble, configure, and tune. + + ### How To Play Machine Design / Robotics can be *basically* broken up into three fields: **Hardware** (as in mechanism), **Electronics** (also 'hardware' to some people) and **Control**. @@ -139,24 +141,48 @@ For assembly, it's best to follow along the documentation on the [RCT Gantries]( Filippos has put some documentation together from his experience doing all of this... is it here yet? +Hardware + - most of the plastic parts go together using the 'No.6 Plastic Thread Forming Screws' - to use these, make sure you have a clearance hole (~ 4.5mm OD) and that you pre-drill the other side with a bit in the region of 2.5mm od. + - the Nema 23 Motors are mounted with M5 SHCS, one mount hole can be tapped, the others are slots + - rollers use Shoulder Bolts as axis, these have M6 taps. careful to add the right # of shims in the right places (they're important!) and don't over-tighten, as the M6 threads can pull out of the HDPE quickly + Belts take some time to get right: - tensioning the belt can be a pain - pinch it down on the static end (smaller 'block') first, then get it around the motor, and attached (but not tight) on the other end (l-shaped block) - the motor is mounted on slotted arcs, this is setup so that one mount screw (M5s) can be pinned, and the others rotated, to tension the belt at the motor - the l-shaped block can also add tension: two of the screws whose axis are parallel to the gantry's motion are there to secure the belt in the block. the other two are there to push against the edge of the gantry, thereby pushing the belt block away from the other side, and adding tension - one way to get around this is to use *only one belt* instead of the two meshed belts. This will be 1/2 as stiff, but will cross the threshold of 'working' vs. 'not working'. - - TODO motor stuff / video here ? - ## 4) Electronics - - video ? what to say ? +There are some notes on wiring up the controllers on the [automatakit](https://gitlab.cba.mit.edu/jakeread/automatakit) page. + + + +#### Circuits + +First, I would test each board, and make sure it will load code. Wire up some power to the board (using eye terminals and the 18ga wire provided - I solder these together) and make sure that one green LED lights up - this is connected directly between the +3v3 and GND line. + +- the system is meant to run on 24v, any more voltage than this and the buck regulator might give up + +Once you can confirm each board gets power, you'll want to make sure they can program. Assign someone in the group to be the C-Code / Firmware Wizard, and get them to read up on how to program the boards [here](https://gitlab.cba.mit.edu/jakeread/automatakit/blob/master/reproduction/firmware.md). + +Once you can program things, you'll be interested in soldering RJ45 Jacks onto them. OK. + +#### Boards, meet Motors + +Next, attach motor boards to motors. You can use M3 Socket Head Screws with Spacers for this, or, if they show up in time, fancy standoffs. + +Then you'll want to attach the boards to the motor leads: I strip the motor leads pretty short, and tin them. I also tin the motor pads on the stepper board. I use a multimeter to discover which pairs are coils [see stepper motor doc from adafruit](https://learn.adafruit.com/all-about-stepper-motors/types-of-steppers) - these are BiPolar. One coil should connect to 'A1' and 'A2', and the other, the others. Coil orientation is not imporant, it will only affect direction. + +#### Wire Routing + +Routing cables to everything else will take some finesse. **Making cable run smoothly as the machine moves about is something worth thinking about from the beginning** ! - - circuit assembly - - wiring +My attitude is to lay cables out starting from their terminal position, then winding them back to the first static point. Make [drag chain](https://en.wikipedia.org/wiki/Cable_carrier) type loops, testing for full range of motion without pulling on cables. Make sure that you're never exerting real mechanical load on connectors. Zip ties = friends. ## 5) Controllers - - atkapi hello worlding +To control this thing, we use [RNDMC - RUN DMC - a Reconfigurable Numeric Dataflow Machine Controller](https://gitlab.cba.mit.edu/jakeread/rndmc). # End Effectors diff --git a/images/ee-spindle.jpg b/images/ee-spindle.jpg deleted file mode 100644 index df473cb5270cbf15cbab9986b038ddbf6a4bc6d6..0000000000000000000000000000000000000000 Binary files a/images/ee-spindle.jpg and /dev/null differ diff --git a/video/3min-for-10min-enc.mp4 b/video/3min-for-10min-enc.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..efa38e80592532d33d29c6e7c62cd77ce6eceb31 Binary files /dev/null and b/video/3min-for-10min-enc.mp4 differ diff --git a/video/SEQ-wiring-enc.mp4 b/video/SEQ-wiring-enc.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..cb5de46b31a564d36352346d42f6e88137acbb4b Binary files /dev/null and b/video/SEQ-wiring-enc.mp4 differ