Gas sensors-type:MQ-5

Hi all, I'm trying to do a project that will detect the LPG gas leakage and send an SMS to the owner intimating them of the leak(using a GSM module ).I'm using the arduino Uno microcontroller .So,i bought the sensor MQ-5.i really need help on how to get started with calibrating the sensor.First off,do i need to caibrate it?? The only time i've ever performed the calibration process,were on pololu IR sensors for a basic line following bot(my first and only project so far,in robotics ).Basically, just a beginner :-p So,help please!


If you want an accurate sensor, you have to spent hundred(s) of dollars. The MQ-sensors change their resistance if something is detected. With a know concentration of gas, you know what value corresponds with it.

Is the gas detection indoors ? Is the sensor close to a possible leak ?

send an SMS to the owner intimating them

Threatening messages eh. Informing them is so much more friendly.

Calibrating the sensor requires that you have access to a known ammount of the gas so I suspect you will not have the equipment to do that. If you are trying to detect a low level leak then I suspect that this is not possible in your budget.

This was the problem statement(sort of ) that was given -"If the LPG sensor senses any gas leakage from storage the output of the sensor goes low. This low signal is monitored by the microcontroller and it will identify the gas leakage. Now the microcontroller turns on an LED and a Buzzer. After few milliseconds delay, it turns on the exhaust fan for throwing gas out and then sends a message "GAS LEAKAGE? to a mobile no.

This was a final year engineering project done in college. Is it really not feasible indoors? to detect, say, a leak in the kitchen?? i have an old MQ-6 currently.I plan to buy an MQ-5 sensor which i've heard can be bought for under 4 dollars?Will that not suffice for a project such as this? I read the link thanks for that :-) ,but i'm still unclear on the calibration process involved..

and erm, i think "intimating" was probably,accidentally read as "intimidating"? :-p..I want to do the project mainy, to gain some experience working with GSM modules..

If it is just a collage exercise then the exact value you get from a reading to trigger an action is unimportant. Just implement this value as a variable and put a note in the initialisation saying you need a calibrated value here. At its simplest calabration involves putting the probe into the gas concentration you want to measure and seeing what the reading is. You then would have to adjust that reading for humidity and temprature so you need to be able to sense those in your project too. It is not worth doing this in the context of your project.

trapiz_proj: and erm, i think "intimating" was probably,accidentally read as "intimidating"? :-p..I want to do the project mainy, to gain some experience working with GSM modules..

I figured you meant something more like "informing them."

Indoors is better than outdoors. Those sensors work better indoors.

To detect a leak, a precise calibration might not be needed. Perhaps the concentration is high in a corner or at the floor (heavy gasses) or near the ceiling (light gasses) and not so high near the sensor.

You could send the analog value at intervals, to see what normal values are. After a few weeks you would know what the critical level is.

These MQ sensor work best as air quality sensors. The analog value is an indication of the air quality. With an Ethernet Shield, you could send the values to and have a nice graph.


See discussion here for calibration in open air:

Here are my libraries with curves calculated from datasheets :

The MQ series are very cheap, variable, and change their values in time. It is a good entry point but no more after. Oftern the knob on the pre-assemble boards are only sed for the digital signal and its led on the board.