Proto #2 The Intensimeter
As the Color Cube could prove that the electromagnetic field can be detected around people, it needed a new way to display not just the direction of change of the field but also its intensity. The Intensimeter still has the LED just like the Color Cube, but it also has a power meter to show the absolute power of the field around its antenna. It also got a external antenna port, so I could test all the different antenna/sensor designs.
The heart of the Proto #2 is Atmel ATmega168. Just like the previous version, it was assembled on a breadboard and it was programmed by an USBASP ISP programmer. The CR2032 battery power the microcontroller the LEDs and it can be charged through the ISP port.
New Materials: $0
Net cost per unit
$_3.00 - Atmel ATmega168
$_0.35 - 16Mhz crystal
$_1.10 - reset button
$_0.80 - capacitors
$_0.60 - CR2032 battery
$_0.25 - 2x5 IDC socket for ISP
$_0.50 - RGB LED
$_1.00 - indicator leds
$_0.50 - resistance
$_1.80 - panel
$_1.20 - test antennas
$11.10 - TOTAL
The only software it needed was the Atmel bytecode which was uploaded to the ATmega168 by the USBASP. The code was compiled with Atmel Studio's C compiler.
The software averages the measured values and compares it to the currently read one. It also has a simple algorithm to calculate the RGB components of the color to set for the LED. It also drives the intensity indicator LEDs and the heartbeat LED.
22.1h - development (between 2016-03-09 and 2016-03-23)
_3.5h - assembly
25.6h - TOTAL
The software reads the field strength by the built-in ADC continuously and calculates and sets the color of the LED, and turns on the number of power indicator LEDs needed to be turned on.
If the field strength remains roughly the same, the RGB LED keeps its green color. If the field strengthens (gets more positive) it changes to red, if it weakens it becomes blue.
The antenna source can be changed and also the reference value of the ADC can be selected on the fly by changing the jumper.
To charge the battery both JP1 and JP2 must be closed and the charging voltage must be on the pin 2 (+3.3V) and pin 10 (GND) of the ISP port. To be able to have 3.3V input voltage you should use a USBASP which has the option to select the ISP voltage of 5V and 3.3V.
So, to charge the battery, close both JP1 and JP2, connect the USBASP and the board with a 2x5 IDC cable and connect the USBASP to your computers USB port. Charge it at least to 1.8Vs, but no more than 3.0Vs. This will be both the VCC of the microcontroller and the reference input of the ADC. The lower the battery the more sensitive the measurements will be, but as soon as it drops below 1.8Vs the microcontroller will not start.
The results were again promising: the LED changed its color if someone got close to the device, and the intensity indicator showed correctly if the field got weaker or stronger. The drained 3V battery was a good ADC reference voltage, better than the fairly stable USB 5V.
After testing different antennas, still the original, long blanked cable was to most accurate. Unfortunately that's not feasible for a flat-detector-panel design.
The prototype worked as expected, so the further development is justifiable. One of the challenges ahead will be to find a good antenna for the ADC input, which is also cheap, easy to manufacture and small enough to place many of them very close to each other.