Ya I’ll try and a bang out a 3d model for the thing
Maybe with hackspace banner or something
3 Likes
We also decided that it would be a lot easier and better aesthetically to not flip them upside down, but to have a closeable hole in the top for mounting on the ceiling and a hidden hole in the bottom for puppet holding and mounting at events.
2 Likes
Ok so I’m having trouble getting a copy of rhino to work on my computer
But I have a trial version at work so I mocked up the chassis we were talking about where to top can screw in and the bottom could be a puppet
2 Likes
And here’s a shot of a more complete model
I’m still working on the brain bit but we’re getting the idea right
This is all scaled and it’s a tad larger than the. Globe space so we may want to drink it
Love this project! Has anyone been looking into the tentacle mechanisms lately? 
I have a few spare 9g servos I can donate to the cause, but all my spare time is devoted to Project Hackenburg right now.
1 Like
This is @flashstick 's baby
But I’m pretty sure we’ll take all the gear and help we can get!!
1 Like
looks hilarious. luvin it. you could always just pull the eyes in and press the tenticles up against the sphere . scale of the body compared to the sphere is nice.
Agreed bit if it is gonna move those tentacle the contact with the walls.may be a problem
This is about the proportions I was thinking, everything is pretty close to actual scale.
3 Likes
Works for me
Any thoughts on how to start that bad boy?
Hey I did some tweaks to aim at what you got there
What do you think about this then ??
Obvi I haven’t done the bottom stuff
What are the dimensions on your model? I have everything measured out on my Maya version, I can post them whenever my computer decided to finish updating.
I’m concerned about the space inside the brain, because that needs to fit all the servos and wiring and lighting connections and stuff, but I think we won’t know how much room all that will take until we actually get supplies to start building.
I think the tentacles are going to be the first step for building so we can get the space needed for the servos figured out.
1 Like
I’m trying to work out controls for the various elements.
(I’m abbreviating Codename: Squiggles to CN:S for ease of typing. )
I want everything to be able to be controlled via bluetooth (I think) from a wall panel for different modes.
- Passive Mode: when nobody is directly interacting but CN:S is still moving around of their own accord
- Active Mode: when people are around and CN:S is interacting with them based on input from the wall panel for moods and lighting and such
- Controlled Mode: when someone is using a controller (like an xbox one) to move the eyes and maybe control tentacle speed and direction
- Sentry Mode: low power mode that will save energy and not move around as much, but also switch back to passive mode if movement is sensed
Within the main body of CN:S we need control for
- at least 12 servos for the tentacles (2 per tentacle)
- lights in the tentacles
- lights in the main body
- motion for the jellyfish tendrils
- the eyes, however those are going to be done (I’m still unsure about this)
I think it would be cool to have the whole body able to rotate to focus on movement in the room on Sentry Mode, but idk how feasible that is.
Am I right in thinking a Raspberry Pi or something would be better than a microcontroller for all of this? I’ve only ever worked with Arduino-type things but I feel like this is more complex than an Arduino can handle.
2 Likes
We would love your servos! How many do you have?
We need 12 of them, but if you have less than that we could still use them for prototyping!
2 Likes
This can totally be done with a slightly more powerful micro-controller like an esp32 (which gives you wifi), or a teensy 3.2 or similar. As long as it has a hardware i2c port to handle a bunch of expanders. I don’t like pi’s because you start adding a whole other software stack of things to go wrong, but I am not a good coder.
Servo control: 16-CHANNEL 12BIT PWM SERVO DRIVER I2C PCA9685 - drives 16 servos, controlled off of i2c so you can daisy chain them if you want. Quick to update over high speed i2c. Work great.
Sentry mode: several PIR sensors, pin goes low when it senses a change in the local heat: SENSOR, PIR MOTION, HP-208N or similar. Connect one in each direction you want to sense. If you start running out of pins on your micro-controller switch to a teensy 3.5 or 4.0 or go for a port expander chip like the mcp23017 (which also talks over i2c) MCP23017 - i2c 16 input/output port expander : ID 732 : $6.95 : Adafruit Industries, Unique & fun DIY electronics and kits and has an easy library to use it with.
You can also use a relay to apply or kill power to the servos when they are moving/not. Just basically wire it in to the power input to the above board. This will also minimize heating of the servos and make them last longer. Get a powerful relay (30 amp+ dc rated) as the inrush current to servos is substantial.
Lights are easy with something like the fastled library and apa102s or neopixels etc and a fairly fast microcontroller GitHub - FastLED/FastLED: The FastLED library for colored LED animation on Arduino. Please direct questions/requests for help to the FastLED Reddit community: http://fastled.io/r We'd like to use github "issues" just for tracking library bugs / enhancements.
Eyes could be something like led matrix panels: LED MATRIX - SERIAL INTERFACE - RED/GREEN which would let you program different expressions. They make them in single color or rgb versions: NEOPIXEL NEOMATRIX 8X8 64 RGB LED
Could also just do neopixel rings: https://leeselectronic.com/en/product/15215.html
As for the code, a good way to handle something like this is to have conditional flags for various modes, and a loop that just goes through with if than else statements to do different things depending on which mode it currently is in. With that you can avoid delay statements, I can sit down and run you though the idea sometime if you like.
code for my beaver, there are just various functions that get called to check inputs, update the motors, update the displays etc. All on timers.
Rotating the body can be done, you have issues with the electrical connections, solveable if you don’t want it to rotate more than 360 degrees, otherwise you need a slip ring and that gets fun when you have a lot of power running through the thing (needed for the servos). Basic idea is to mount the whole thing on a ring and find someway to rotate it. I have done it in the past with a chain sprocket from a bicycle and a small stepper motor.
1 Like
I have one of these (or at least a knock off of a knock off) I will happily donate to the project…
3 Likes
I do feel like a RasPi with Python will get you to a workable project faster. Microcontroller development has a lot of (performance-related) benefits, but it is way slower to develop for.
I’ve had good luck with the RasPi’s servo driver library for Python, buuut, you can totally use I2C chips like this for the Pi and that’s probably more robust.
Also long-term stability with the Pi is an issue I’ve had, so David is totally right, but stability improvements seem like stretch goal over getting V1 working fast.
All of these are very much my own biases
2 Likes