Connecting Arduino microcontroller with MQ sensors is quite straightforward, but then many challenges come. Namely, you need to (1) make a real-time recording system that stores the sensor output, (2) deal with somewhat slow recovery of the sensor, (3) manage sensor cross-sensitivity and (4) compensate for humidity and temperature shift of the sensor sensitivity (5) and have means of data analysis.
In my first setup, I managed to make an operating system that has a sample injection port (A), the sensor (B), micro SD card module (C) and Arduino Uno (D).
I implemented micro SD card module for Arduino, and with some coding, I manage to record data with a time resolution of 1 second (could be set up much lower).
My Arduino code for the system is available here!
I also made a script for peak integration that makes the data analysis robust. I am happy to share this as well.
The most significant improvement of my system, which made sub-ambient methane measurement possible, is that the sensor was enclosed and exposed to the stream of nitrogen (see my previous blogs for more on the sensitivity and detection limit). This simple approach effectively ruled out some of the challenges mentioned earlier, and this is how. Fist, the recovery time is much better if the sensor is not exposed for a long time to high concentrations. In my system, I load just 2 mL of air, and recovery is around 1 min for ambient level, 4 min for the concentration of 5 ppm, and approximately 20 min for the concentration of 10 ppm. So, you may play with dilutions and keep your system running with < 2 min resolution. Also, loading 1-2 mL in the stream of 50-100 mL/min of nitrogen enables somewhat stable temperature and humidity conditions, so the sensitivity shift that sensors suffer should not affect the measurement. However, for more advanced system, I recommend adding temperature and humidity sensor and calibrate properly.
A tin-oxide sensor such as MQ-4 is also sensitive to H2, CO, alcohols and other alkanes. Although this should be further tested, the ambient concentration of these are in ppb range and not ppm so the effect should not be significant, at least no for rural and remote areas, far away from the traffic and gas-lines leaks.
To conclude, as far as I tested, this is a good system that could be used for scientific projects in quasi-stable temperature conditions of a laboratory or a tent. It is a low budget solution and could be combined and developed much more: autosampler, GPS, wi-fi, remote control, cheaper zero air, battery power, etc.
Hope this will help projects which have a limited budget or enough courage to try something new (e.g. dozens of these systems, simultaneously measure at a remote site, sending data to a base).