Progress is being made on the hardware. The motors are mounted, the EggBot controller board is mounted, new spools are printed and installed.
Here is the overall mounting scheme so far
I ended up printing these small spools from thingiverse.
I used a screw specifically made for screwing into plastic. It has deep coarse threads that bite well into the soft ABS plastic. We’ll see if it holds long term but it seems to have a good grip right now.
The EggBot board has an issue where the upper and lower copper planes are one power and one ground. The problem is the mounting holes do not have copper pull backs so if you put a metal screw or stand off through the mounting holes the power plane and ground planes are shorted together. So that’s no good. Mounting the EggBot board requires using non conducting mounting hardware. I decided to print a mounting platform and hold downs.
I printed reliefs in the mounting platform for the pins that protrude from the bottom of the EggBot board. The four pins in the corners go through the EggBot board mounting holes to align the board and keep it in place when held down.
I’ve done some simple tests communicating with the EggBot board and moving the motors, making sure everything is still in working order and nothing got crimped or crushed.
Next steps are to string the gondola and get some code working.
I’ve combed through enough of Wilba’s code to understand the fundamentals. I’ve done some cleanup and added comments and a bit of debugging functionality. I’ve uploaded the latest updates to github.
Searching the web for drawbot stuff I came across a post by Wilba on XKCD’s forum with some pictures and comments describing his drawbot geometry. This was a big help when trying to understand the code.
Wilba’s drawbot code supports using a larger spool and calculating the tangent point of the cord from the spool when calculating the pen position. The drawbot I’m making is much simpler. I’m using fishing line through an eyelet which fixes the cord origin regardless of the spool diameter or angle to the pen. Using a string around a spool does have the issue of the spool diameter changing due to wrap up of the string around the spool, but I’m using fishing line on a reasonably sized spool so the diameter change should be small.
At this point I’m going to fork the code from DrawbotWilba to DrawbotOne. DrawbotOne will be simpler to start with since it will only support the cord entering the drawing space from a fixed point at the eyelet.
I’ve taken Wilba6582’s eggbot drawbot code and cleaned it up enough to build under Processing 2.0. It requires grabbing the latest controlP5 library and I’ve removed the gamepad control.
I haven’t tried running it against an EiBotBoard yet but that’s the next step. The code builds and runs, it rasterizes an image, and brings up a gui with the ability to drag the pen around.
I’m in the process of organizing and cleaning up the code and I’ll continue to push changes to github.
We’ll see where this goes but it gives a nice starting point for jumping off, a big thanks to Wilba6582 for doing the initial work.
Here’s the interface up and running from the barely modified code.
I mentioned in an earlier post I had found a video on youtube of someone else using an EggBotBoard to drive a drawbot. That was pretty exciting because most of the drawbots out there (and maybe we’ll find later for good reason) are based on the Arduino / Adafruit motor shield combo. Since I decided to use the EggBotBoard it was good to find a kindred spirit that had already blazed a trail.
I left a comment on the video asking about what software was used and Wilba6582 responded and sent me a bit of code. I’ve just started to look at the code but from Wilba’s comments in the email he sent the code is targeted for the Processing environment and as shown in the video actually parses a file and drives the drawbot.
I’m posting the code here along with Wilba’s release. As I dig more into the code I’ll decide how much can be used as is, how much needs to be updated, and how much needs to be re-written wholesale. Part of my problem is that I’m not fully familiar with the Processing environment so I’ll need to take a moment or two to figure that out.
In the meantime I’m working on mounting the motors and electronics on a portable rigid surface (think piece of wood) so I can place the whole package above the surface I want to draw on, plug it in, and draw away.
As mentioned in an earlier post I have decided to use the Eggbot control board (EiBotBoard) to drive my drawbot. I’ve gotten the electronics up and running, mostly just plug it in, screw down the stepper motor wires, and plug in the pen lift servo motor. Here are some videos of the EiBotBoard in action.
Some commentary on why I am using the EiBotBoard. The EiBotBoard is a nicely integrated dual stepper motor, multi servo motor driver that accepts high level motion commands over USB from a serial console.
Description of the motors selected for the drawbot. A couple NEMA 17 stepper motors I had laying around in my parts box. They are 0.8 amp motors, well within the 1.25 amps the EiBotBoard can supply.
Motion demo of the EiBotBoard driving two steppers and a pen lift servo. The script that is running is posted below. I am using a cygwin command window to echo the commands to the USB virtual comm port. Keep in mind Windows enumerates the COM ports starting at 1 and cygwin enumerates /dev/ttySxx starting at 0. In the script below I am echoing to ttyS18 because Windows enumerated the COM port at COM19.
Finally more commentary on why I like using the EiBotBoard.
Send the version command to the board, establishes communication with the board.
echo “v” > /dev/ttyS18
Move both steppers at 1/16 size steps (the default) for 500 steps each and take 1/2 second to move (500 msecs)
echo “sm,500,500,500” > /dev/ttyS18
Set the pen lift servo min and max values, 6000 and 26000, determined from experimenting with this servo
echo “sc,4,6000” > /dev/ttyS18
echo “sc,5,26000” > /dev/ttyS18
Each TP command toggles the pen position from up or down
echo “tp” > /dev/ttyS18
Set the stepper fraction to 1/4 sized steps.
echo “em,3,3” > /dev/ttyS18
Step the motors for sm, duration in msecs, number of steps motor1, number of steps motor2
echo “sm,1000,2000,2000” > /dev/ttyS18
echo “sm,1000,-2000,2000” > /dev/ttyS18
echo “sm,1000,2000,-2000” > /dev/ttyS18
echo “sm,1000,-2000,2000” > /dev/ttyS18
echo “sm,1000,2000,-2000” > /dev/ttyS18
echo “sm,1000,-2000,2000” > /dev/ttyS18
Disable the motor drivers to avoid over heating the motors.
echo “em,0,0” > /dev/ttyS18
Not a huge amount of progress on my drawbot but I’ve stumbled across a couple interesting ideas I want to capture as part of this build stream.
I have printed a gondola and a couple spools. I printed this gondola and it looks good.
It holds a sharpie marker great, tight with just the right amount of tip poking out. It has good mount points for the cords and a place to put some weights on the bottom. The problem is, and I didn’t think about it before printing it, there isn’t a way to mount a pen lifter servo. I could probably rig something up that would work fine, but I’d rather have a designed in solution. So I’ll likely be printing a new gondola soon. This gondola will probably work fine for a TSP single line art but I want a pen lifter so I can do pointillist style images.
I also printed out some spools, the ones I mentioned in my earlier post, the ones makerblock recommended I shouldn’t use. They printed fine and I think I can glue them together to avoid the coming apart and unspooling problem. But they aren’t ideal, for a lot of the reasons makerblock pointed out. They press fit onto the motor shaft, no set screw, they are kinda narrow so you get cord build up changing the spool diameter. They looked good on paper but after printing and holding them in my hand I can see some of the issues makerblock pointed out. So I’ll be printing new spools.
On to the interesting part. I found a link on youtube to a guy who made a drawbot using the Eggbot hardware. So far I haven’t found any more details but at least someone tried, and succeeded, at this before.
Second, and I think more cool. I stumbled across a video for building a drawbot on the cheap. This guy uses some super cheap stepper motors, less than $5 USD each with motor drivers, simple 3D printed spools, and cup hooks for cord guides. No motor mounts, he just screws through the motor mounting tab into the wood drawing surface. Doing a quick BOM using the cheap motors and drivers, a super simple bare bones arduino clone, 3D printed spools, cup hooks or eyelets for guides, a cheap wall wart power supply, and a hand made or 3D printed gondola I think you could make a drawbot for less than $50 UDS, maybe less than $35 with some searching.
To me this is a super exciting way to introduce robotics to newbies. The drawbot itself is cool, people just seem to like watching the drawing appear physically. The cost is pretty cheap, it seems within the range of a learning or science project, and it’s super simple to build. I think a high school or middle schooler could easily assemble a simple drawbot in a day. Build up some kits and sponsor a build a drawbot day at your local school.
Drawing board, melamine coated shelf board from HD
Stepper motors with drivers
A cheap arduino clone, preferably already assembled to avoid soldering
Cables and wiring harness, preferably already assembled for plug and play
3D printed spools and gondola
Cup hooks and screws
Some fishing line
A sharpie marker
The software would need to be developed to drive it but that problem is being worked. I think this is a cool idea that could be pretty interesting for young wanna-be nerdlings.
Search DealExtreme for stepper motors, I’ve found some motors with drivers for less than $4 USD each.
I’ve decided to use an Eggbot control board, the EiBotBoard, to drive my polar plotter.
Not for any particularly wonderful reason. The board costs right at $50, which is the cost of an Arduino uno R3 ($30) plus an Adafruit Motor Shield ($20).
It seems the software to drive a polar plotter already exists for the Arduino / motor shield combo. Eggbot board, not so much.
So I’m looking at the same cost and more work to use the Eggbot board. But hey, if it was too easy it wouldn’t be fun :). I like the form factor of the EiBotBoard and if I can get it working as a polar plotter I can add a little back to the community.
Ever since I saw a post about a drawbot on Hackaday I’ve been wanting to build my own.
A quick search of drawbot or plotterbot or polargraphic plotter will lead you to plenty of youtube videos and websites describing and demoing polargaphic style plotters.
I don’t expect to break any significantly new ground here. I’m planning on building a pretty standard two stepper motor, servo pen lift, monofilament on pulleys, weighted gondola polar plotter.
One thing I am doing that is possibly different, at least I haven’t seen it’s like yet, is I’m planning on using an EggbotEiBotBoard for motion control.
If you don’t know what an Eggbot is google it, watch a couple youtube videos, visit a few websites. It is a cool toy, and fundamentally nearly identical to a polar plotter. Basically an Eggbot is a spherical plotter. It should be a short step to a polar plotter. Certainly the hardware should be a drop in. The EiBotBoard has built in two axis coordinated stepper motor control, a pen lift servo control, and open source software for control and drawing conversion.
I plan on 3D printing the motor mounts and the pen gondola. Thingiverse has multiple models for motor mounts of various sizes and multiple different gondola styles. No real rocket science there. Download the models, update as needed for my motors and pen, and print them.
The interesting part is likely to the modification of the path generation to interface to the EiBotBoard. The Eggbot community has written a plugin for inkscape to wrap a flat vector drawing to a sphere and output the motion commands to the Eggbot EiBotBoard. It seems like a reasonable launching point to start with that plugin and modify it to convert line art to polar graphics and then drive the EiBotBoard on a polar plotter.
Often my blog posts are generated after the fact, after I’ve already completed the project, or done enough to figure out what I want to say or how the project is going to go down. In this case I’m blogging as I go. I haven’t started my polargraphic plotter project yet, other than a vague idea of what hardware I want to use and some of the mechanics that I think will work.
Next steps include picking parts and getting them ordered, picking some models and getting them downloaded and printed, downloading the Eggbot interface software and figuring out what’s going on there, then hook it all up and print, easy-peasy-pie.
