Commit by micah :: r12207 tedrx/scripts/ (update-graphs.sh tedrx-daemon.py):

Move graph output to a separate directory (on tmpfs), temporarily disable voltage graphs.

Commit by micah :: r11970 tedrx/scripts/update-graphs.sh:

Add a copy of my script to start the daemons, update an RSS feed, and generate
graphs.

Commit by micah :: r11969 tedrx/scripts/tedrx-daemon.py:

Increase the XFF for the first two RRAs, to make this script more
resilient to dropped packets. (That is to say, more willing to make up
data when we aren't getting any from the transmitter.)

Commit by micah :: r11968 tedrx/ (scripts scripts/tedrx-daemon.py):

Add a sample daemon which logs power data to an RRD.

Commit by micah :: r11894 tedrx/hardware/kwclock-v1-schematic.png:

Mime type

Commit by micah :: r11893 tedrx/firmware/common/filter.c:

Tune the digital bandpass filter: Use a more accurate center frequency, increase the selectivity, and decrease the hysteresis. In my ultra-scientific tests (blinking LED) this seems to be working better.

Commit by micah :: r11892 tedrx/hardware/ (4 files):

Add a schematic for the Kilowatt Clock

Commit by micah :: r11891 tedrx/firmware/ (19 files in 4 dirs):

Firmware split: The actual packet decoding and software demodulation code has been moved into a common library, and there are now two top-level apps: A receiver with serial output, and firmware for the Kilowatt Clock.

Commit by micah :: r11889 tedrx/hardware/ (2 files):

Set mime types

Commit by micah :: r11888 tedrx/ (3 files in 2 dirs):

Hardware revision 2: Single power supply, and only one op-amp.

Commit by micah :: r11887 tedrx/ (3 files in 2 dirs):

Checking in a schematic of my receiver prototype.

Commit by micah :: r11886 tedrx/ (5 files in 3 dirs):

This could use a bit of polish, but it works. My homebrew receiver is now decoding actual power and voltage readings from the TED transmitter. Also checking in the test harness and spreadsheet I used to determine the linear relationship between the raw data and actual watts/volts.

Commit by micah :: r11885 tedrx/firmware/main.c:

Added some notes with what I know so far about the meaning of the packet contents. Looks like the MTU is actually reporting power/voltage readings with much more precision than the stock RDU displays.

Commit by micah :: r11884 tedrx/firmware/ (tedrx.h main.c filter.c):

New bit/byte framing code in the serial receive engine: Each bit now has weighted majority-detect, and bytes are synchronized to the edge of each start bit. Now I'm reliably receiving a valid-looking 11 byte packet every second. Woo.

Commit by micah :: r11883 tedrx/firmware/ (6 files):

Two failed ideas later, I'm finally at a demodulation and receive scheme that seem to be working really well. The data is indeed 1200 baud, using an encoding almost identical to RS-232 8-N-2. Receiver isn't quite done, but it's late...

Commit by micah :: r11882 tedrx/firmware/ (6 files):

Adding a PLL which locks onto the mains frequency, and generates a real-time estimate of the AC phase angle. Performance is disappointing so far- phase shift is really unstable, and the drawbacks are outweighing the benefits. I'll probably switch to a dumb counter which resets on every zero crossing...

Commit by micah :: r11881 tedrx/firmware/ (tedrx.h serial.c main.c filter.c):

Indent and untabify. This is what I get for writing in C...