An LED clock and using GitHub

Fresh from Ponoko, it all fits!

I have completed an LED clock and have a repository for it at GitHub. As I learn more about git the more I love it. My repository at GitHub has code, notes, svg files for the laser cut parts and even a blender file for the 3D layout. Open Source rules...

I wanted to make something practical, colourful and easily shown off to anyone coming through my door and portable enough to take places. I settled on using 12 RGB LEDs to tell the time with colour and brightness.

Getting good pixs of LEDs is hard

How does my clock tell time? The current hour is Magenta. All other hours are a background Green, except those that tell current Minutes and Seconds. Minutes move through green-yellow-RED-yellow-green. The bright RED occurs at each exact 5 minute interval. Seconds move from green -> cyan -> BLUE -> cyan ->green. The bright BLUE occurs at each 5 second interval. Fairly easy to get the hang of it. Not so easy to describe. Here is an example showing the time as 9:05:24.

 

Blender render of the partial clock assembly.

I learned to use Blender to create 3D renders of my clock assembly. I imported the same Inkscape .svg files that I created for the laser cut parts and modified them for 3D. A bit of a learning curve but possible for guy who turns 70 next week.

• 
  

I also have a repository for my updated Solar Monitor Project which is being done in collaboration with Anthony.

The current version of the Solar Monitor uses an 1.8" 18-bit color TFT LCD display with microSD card breakout and a DS1307 Real Time Clock breakout board kit, both sourced from Adafruit.

1.8' TFT  display

The TFT LCD, a sample screenshot shown here, displays (top to bottom):

• (text): battery voltage and current
• (graph): battery current for the last 100 seconds (positive and negative)
• (graph): integrated current for the last few hours (positive and negative)
• (text): date and time
• (bar graph): battery "percent charged" for the last few days

The monitor also logs the battery current and voltage, integrated for one second, on a micro SD card. Over time a power profile will develop which can used to determine patterns of appliance usage and specific appliance power consumption.  All this from one current sensor with a timestamp I hope. A Bayes network looms.