MQ5 Gas Senor module - How to Callibrate

Question

I have just bought MQ5 Gas Sensor module for Arduino, (http://www.dhgate.com/product/100-pcs-lot-gas-sensor-module-mq-5-lpg-natural/133345721.html)

According to the Data sheet,

"Resistance value of MQ-5 is difference to various kinds and various concentration gases. So, When using this components, sensitivity adjustment is very necessary. we recommend that you calibrate the detector for 1000ppm H2 or LPG concentration in air and use value of Load resistance ( RL) about 20 KΩ(10KΩ to 47KΩ)."

It says to calibrate the detector to 1000ppm. What is meant by ppm here ? is it the resistance ?

Also, There is a knob through which i can very the resistance, but how am i supposed to know what is the current value of the resistance when i am changing it ?

Datasheet link: http://www.dfrobot.com/image/data/SEN0130/MQ-5.pdf

Answer 1

The ppm is parts per million so it is recommending calibrating using a gas that contains 1000 parts per million of the gas the sensor is detecting. There are products called "calibration gas" made for this specific purpose. For example a Google search for "hydrogen 1000 ppm calibration gas" yielded the following result:

Calibration Gas Cylinder, 1000ppm Hydrogen (H2) / Air, (6D) 103 ltr, CZF7E290152

You might be able to find some cheaper / smaller altenatives but most I've seen tend to be sold in reasonably large cylinders and aren't all that cheap for hobbyist purposes.

While the datasheet isn't that clear figure 3 shows that it seems to have a fairly linear response depending on the ppm concentration. According to the Atmosphere of Earth aricle on Wikipedia the concentration of hydrogen in the atmosphere is 0.55 ppm by volume. So for practical purposes you should be able to treat exposure to air as close to 0 ppm and calculate and extrapolate the slope between that and 1000ppm to calculate other values.

Answer 2

The load resistor changes the response characteristics and sensitivity of the sensor. The value you set it to depends on the unique characteristics of that particular sensor (the sensor resistor value under the target analyte concentration and your heater value). They suggest that value (as well as the 1000PPM calibration value) as a rule of thumb. Depending on your requirements, you may want to calibrate at different concentrations. To properly calibrate that sensor and derive a function that gives you a PPM value over a range, you need to expose the sensor to varying concentrations of the analyte and measure the output at each interval then develop a best fit formula (via regression or your method of choice). However, if you change the resistor value, you have to repeat the whole process over. Unfortunately, that sensor is NOT linear. It's roughly log/linear but it's not that simple since the heater pulse width, voltage and various other factors impact the response characteristics of the sensor. You also have to be aware of integration time, since it takes some time for the sensor to adjust to changing analyte concentrations. In our experimental results, the time integration function is non-linear as well.

PPM in that context is a measurement of analyte (gas) concentration (parts per million).

H2 calibration gas cylinders can be fairly cheap ($40) but it depends on purity, cylinder size, concentration, etc. Your local university (environmental sciences folks) might have access to an H2 generator or calibration gas.