[Note -- a follow-up discussion around this technology actually works -- with some great commentary from the Public Lab community on ways of perhaps improving upon it -- is here]

What I want to do

Make a simple, cheap interface between an inexpensive particulate matter sensors described here and here -- and a laptop, PC, or Raspberry Pi. This is based on Jeff Warren's suggestion in a comment thread that we might using the DigiSpark to make such a connection. Such a design would not be as nicely independent as the DustDuino or a Speck designs; but if one had a laptop (or Raspberry Pi) available to make the measurement, it might be a very inexpensive solution.

The idea is the device measures the signals coming out of the particulate matter sensor, and the laptop / Raspberry Pi can measure out the particle levels via the USB port:

A breakdown of the sensor innards of the type of device under consideration is depicted here (image is from Willie's great research note):

In the same comment thread referred to above, I wrote some words about how I think the sensor works. From what I can gather, the sensor readout provides a 'high' or 'low' voltage depending on whether a particle floating in the air chamber reflects some of an internal IR beam's light into a photodiode inside the device. Tracking the percentage of 'high' vs. 'low' signal, and using additional information (assumptions about the particle sizes necessary to reflect a certain amount of IR light into the photodiode, maybe?) the particulate matter concentration in the air can (apparently) be assessed.

My attempt and results

Here's a design for a simple microcontroller interface between one of these simple, cheap particulate matter sensors, and a laptop / RPi / PC. I tried to make the board size as small as possible:

... the optimal arrangement of pins and components to make the PCB as tiny as possible ending up resulting in a 'P' shape -- where 'P' stands for 'Particulate matter' or 'Public', say!

I chose the 32u4 processor over the DigiSpark's ATTiny, because:

• The 32u4 allows for a regular serial interface, so that we can easily suck the values into a laptop or Raspberry Pi
• The 32u4 has more memory, so that we can add additional functionality down the line
• The 32u4 has on-board USB, so that we can program the mini-meter board easily (with the appropriate bootloader installed, it should show up as a 'Leonardo' in the Arduino IDE).

[I've also added some fuses and diodes -- inspired by the Adafruit Flora design, which uses the same processor -- so that powering the board improperly doesn't 'result in tears', as Lady Ada writes in her Flora description.]

The board itself is designed with a 'PCB USB plug' (Jeff's nice suggestion) -- the board shape and traces on the board allow for insertion into a USB port directly. This means, I think, that we need to ask for printed PCBs of a certain thickness -- 1.6 mm, I think. I've also placed a microUSB header on the bottom, in case folks would prefer to attach via a cable.

Other than a 'power' LED on the board, there are no other frills.

The device has D6 and D8 broken out for the two lines on the PM sensor; we also have A0, A1, and SDA + SCL, for additional measurements (e.g., I'm imagining we might want to measure temperature, or air velocity).

Questions and next steps

• Could such a device be used easily with a Raspberry Pi via USB? I believe so ... will check.