3D Printing a Rollerblind Cog24 Aug 2014
Lately I've been trying to automate the rollerblind in my office. I'd like it to close itself when it gets dark, and reopen in the morning...
I've a bunch of spare steppers lying around from an overly enthusiastic buying spree in Shenzhen's SEG market, so I'll use those for my initial experiments.
First up is creating something to grip the ball-chain-thingy, it looks like this:
The balls on my chain are approximately 4.4mm in diameter, the string between each ball around 7.5mm.
The rollerblind already incorporates a rather elegant design for gripping the chain, so I've started by trying to mimic that.
I used OpenSCAD to design the part because creating it programmatically makes it easy to experiment with the number of teeth, hole sizes, etc.
Here's what I came up with:
It works by trapping the balls between the gaps in the teeth while under tension. The chain is then dragged around as the cog turns, lowering or retracting the blind depending on the direction of rotation.
I split the design into two halves for printing, and added nubs to help align the two halves and aid gluing:
My Mendel90 3D printer is fairly well calibrated but not perfect.
You can't see it, but the hole in the middle of the cog is "D" shaped to mate with the notch on the stepper 5mm shaft. Printing the hole at 5mm proved too tight a fit, so I printed a bunch of test holes with radii ranging from 2.5mm-3mm:
2.7mm was a nice snug fit, giving a 5.4mm diameter (+0.4mm) which is coincidentally the size of aperture in the printer hot-end.
Hot off the printer the new cog was ready to be glued and tested!
I used an EasyDriver, Arduino, and 10K potentiometer as a speed controller to test it out.
The spacing on the teeth isn't quite right yet, so there was some slipping, but the results are promising; it was able to raise the blind unaided. \o/
I'm quite surprised that the stepper could supply enough torque, though I did have to run it at 12V with EasyDriver set to its maximum output current (~750mA).
- Tweak the cog design to correct the tooth spacing.
- Tidy up the SCAD file, it's hacky at the moment.
- Play with gearing to reduce the power required & smooth the motion.
- The driver chip was fucking hot afterwards so I'll probably heatsink it.