I’m an art student back with a new project for the new year.
Last year I made a drawing clock (Drawing Clock), and this year I plan to make a wall drawing machine, controllable by both myself in situ, and by stepper motors, programmatically, in my absence. The idea is that the machine will always be drawing, whether I am there or not. As an artist, I’m exploring what happens when we hand over our own agency to external sources, whether that be other humans, machines, or AI.
The wall drawing machine will be a Y-plotter, similar to the likes of http://www.polargraph.co.uk/
However, unlike the polargraph, I want my machine to be controllable by hand and machine, and record my own movements for playback and subversion. Kind of like a giant human etch-a-sketch with record, playback, and interesting interventions.
Here is a picture of the non-electronic predecessor of my project where I pull two strings attached to pulleys to control the XY position of a piece of charcoal attached to a gondola:
To electronicise this drawing machine, I was looking at replacing the pulleys with stepper motors with encoders, but wow they’re not cheap! Can you think of another solution that might work? If I have to shell out at the end of the day, I will, but had hoped to ask the hive mind here first
Awesome project idea! With a feature wishlist like that, I couldn’t think of anything off-the-shelf that would fit the bill. Closest I found was a similar concept, but in 3D:
You’re definitely after encoders for that recording function, and as for whether you go with brushless motors or stepper motors as your driving motors is up to you.
I’m a fan of the ODrive project, but that might be overkill for the power that you need.
Does it need to be quiet? That will help us pick out some hardware.
Keen to see where this goes!
Again, a very interesting project. The first thing that comes to mind is to switch out steppers for servos depending on the precision of the control that you need. They’re often used for various RC and robotics projects and with the appropriate gearing and power supply can kick out a pretty significant amount of torque while maintaining rotational accuracy quite nicely.
Not sure exactly what kind of specs you’d need for your project, but I’d imagine that it won’t be very much weight, and you’ll need continuous rotation to essentially reel up or down the lines. If you do decide to go down the Servo route, just be sure you’ve got continuous rotation and you should be fine (not to mention that you won’t need any expensive stepper motor drivers to be able to use the motors depending on the microcontroller that you use with it)
This motor is purely an example of what I’m describing and likely won’t be the best option off the bat so to speak
Thanks for the feedback and encouragement James and Bryce!
Woah some of the examples done with oDrive look like they could fly my machine to the moon - could definitely be overkill! It doesn’t need to be quiet in the slightest.
Just wondering, with servos, what sort of precision might I expect? I have only worked with micro-servos before to do something far less intricate (drawing straight lines). The more “human” I can get the machine to be, the better. I imagine that a rapid series of high-torque movements are going to be necessary, just judging by how the gondola needs to move suddenly against the surface to leave a good looking charcoal mark. There’s a lot of sharp backwardsie-forwardsie jigging the lines around, rather than smooth line marking so to speak. Would a servo serve me best then?
The gondola currently weighs 300g - it needs the weight in order to make a good mark on the wall.
As for a microcontroller, I currently have a few rasperry pi’s around, but open to exploring other options. Clear tutorials are my best friend - I have an understanding of programming and a curious mind, but I don’t have a maths brain so happy to pay extra for noob-friendliness! I really appreciate all of the resources you guys have available here, the support is second to none!
Awesome project, I came across a similar project here: but on a Whiteboard(Made by Hoppuman
He published the code here: TikTokHacks/Whiteboard at master · hoppuman/TikTokHacks · GitHub
If you need to do some movements with lots of jerk you’ll need some juiced up motors, steppers would be your best bet I think - what was the fastest acceleration do you think you’ll need? From those numbers everyone could start sizing up a motor.
To get a better mark can you apply more force into the wall with a mass moving back and forth(on another servo maybe?)
Cheers Liam! I’ve dug up video of the original machine in action because I really don’t know how to describe it in terms that would translate to a hardware requirement, as I am a complete novice in this area.
It’s a really strange feeling to try and find hardware of a similar spec to my bodily capabilities
Oooh interesting - It doesnt look like a massive amount of force would be required, I’d check out some of the higher powered stepper motors, just a note, you can over-volt motors AS LONG as your motor driver has current limiting capabilites.
Also some different sized pulleys will let you change speed and torque - a 3D printer is handy for this!
Something like this might work, I havent done any maths tho: https://core-electronics.com.au/stepper-motor-unipolar-bipolar-200-steps-rev-42-48mm-4v-1-2-a-phase.html
Yeah anything where you are trying to emulate a human is super interesting.
About a year ago there was a post which may interest you
It is about a device called DrawBo.
Click on the Forum search mag glass icon (not the products search) and enter “Drawing”. When I just did that it was top of the list.
Cheers Bob! Ah, what a shame that project got cancelled. There are some really awesome ones out there. It seems like the commercial ones kind of get a bit of support but never really take off. I once saw one that used a spray paint can, which was my personal favourite! I’m personally interested in charcoal and watercolour (with a paintbrush), to step a little further from the digital, back in to the oldschool art material world. I figure if I spec this out for charcoal marks, then anything else should be a cakewalk.
What a great idea. I like using unusual ways to produce music so this resonates with me.
If you google “DrawBot” it still gets a fair bit of attention.