The Glowboard Plotter was built with a piece of 120cm x 60cm Glow-In-The-Dark sheet, a stepper, 64 UV Leds controlled by a MAX7219 attached to a raspberry pi. The Leds are attached to a wooden bar which moves over the sheet. A little webserver is used to receive images and controls the Leds and the stepper so that images are plotted.
The biggest problem is that the glowboard is quite noisy. This is caused by the very simple python program, doing the calculation of the images and controlling of the steppers. So, every time the LEDs are updated, the stepper movement is stopped for a short period of time. You can hear the stuttering sound in the video above.
To fix this, the movement of the stepper has to be controlled by a separate process, which is capable of realtime processing. The easiest way to do this would be to let an arduino control the movements of the stepper, and give our python program running on the raspberry pi feedback about its position.
Another issue is that the openrails use steel wheels rolling over an aluminium profile. By using steel shafts with some plastic linear bearings we could further decrease the noise.
Finally, by placing the stepper motor behind the board and adding some dampening material we hope that we can reach a sound level which not annoying to people hanging around.
All-in-all, these changes mean that we have to build an all new version 2 of the glowboard. Maybe we can increase the (physical) vertical resolution from 64 to 128, which would mean that we have to decrease the distance of the UV LEDs from the board, so that we then can have an "interlaced" vertical resolution of 256px!
Stll way to go for a full HD glowboard, but hey. :-)
The Raspberry Pi 3 attached to the Glowboard Plotter runs a minimal webserver written in python, which can be used to send images and set number of passes. HTML/CSS was done by @sesshotv, the youngest hacker currently hanging around in our hackspace.
By applying multiple passes, every pass containing fewer LEDs (lightest pixels are contained in all passes, darkest pixels only in the first, and so on) the glowboard can now plot grayscale images. We limited the number of passes to 10, which gives very nice results. After a while, the contrast decreases (because brightness of the glow-in-the-dark material decreases exponentially), and the image looks smooth for a few minutes.