AZDelivery MQ-135 Air Quality Gas Smoke Detection Sensor Module compatible with Arduino and Raspberry Pi including E-Book!

Only used the analogue output but output to arduino and read the value.Simple 5v supply but seem to work ok on 3.3v tooCalibration is still a work in progress

The soldering was a tad bit loose & needed constant adjusting to get any output from the sensor itself. But worked fine after a bit of fiddling around

Can print pretty graphs using this sensor as an input.

After testing and leaving these devices to 'settle down', the readings between the devices whilst located next to each other vary so much that I just can't rely on them, they also fluctuate quite a bit which was a surprise too.I've literally got two of these side by side and one is nearly (!) double the value (analogue) to the other (swapped ports to confirm it wasn't my MCU).I can calibrate these which I may do (or get a different supplier in the hope they are more aligned), but not a great initial impression. I did test with some chemicals nearby and they did react very quickly, hence the 3 stars.

It’s a bit vague what this actually detects (although the datasheet lists all the different gasses/vapours), it’s not just organic molecules, it’s not just acidic gases or alkaline gasses, it’s just kind of a weird mix of unrelated things.I would say this is a great cheap sensor for VOCs (e.g. in a workshop) but the fact it picks up atmospheric CO2 and NOx too makes it kind of pointless. It just tells you there’s some amount of atmosphere around.If you used this in only a controlled environment where no more than one detectable gas could be present, then it’s great…but no one buying a cheap sensor has that use-case.So, it’s fun to play with and see that numbers go up when you use alcohol/paint/ammonia/breathe nearby, but for the home-gamer that’s about it.

I bought this sensor to pair with an Arduino Uno to make a CO2 gas detector with the goal of getting CO2 concentration in ppm. The sensor picks up various gases, the datasheet quotes ammonia, various nitrogen-oxygen molecules, CO2, and a few others, so initially I was sceptical about it actually differentiating what it was picking up (and it cannot do this). However, given that CO2 is far more abundant in the atmosphere than the other things this sensor can detect, I thought that it would be worth a go (there are a few online articles covering how to do this that are worth reading if this is your aim).Out of the box, it's useless for CO2 quantification. The datasheet (easily available if you Google MQ-135 datasheet) says that the load resistor (for reference, that is the central one of the three SMD resistors you can see on the back of the module) should be between 10 and 47k ohms. The one on this module as shipped is 1k. Therefore, if you want this to work as it should, you ought to check the datasheet and will probably need to replace the load resistor. I bought a set of cheap SMD resistors off Amazon (0805) and mounted a 22k ohm resistor (required desoldering the factory one and attaching the new one). The result seems fairly convincing, I get 420ish ppm CO2 outside, which is approx what you'd expect.I will test this further at work against lab instruments over a range of CO2 concentrations just to see how well it holds up and will update this if I remember. Breathing out on the sensor causes a fairly modest spike in CO2 that is considerably lower than I'd expect (a quick Google tells me that CO2 in exhaled breath is ~38,000 ppm), but this could be a) because the sensor is calibrated over a much lower concentration or b) because the sensor isn't designed for CO2 concentrations that high (datasheet doesn't actually say as far as I can see).UPDATE: tried this today in a meeting held with 10 people in a seminar room vs a reference CO2 monitor (this thing: Carbon Dioxide Detector Temperature RH Humidity Indoor Air Quality Wall Mountable CO2 Monitor). The reference monitor was fairly consistent at ~500 ppm over the course of an hour (guess there's decent air circulation in there). Initially the MQ-135 was pretty close, and spent ~5 mins within 5 ppm of the reference. However, the MQ-135 measured ppm dropped and stabilized at around ~440 ppm after approx 10 mins - probably coinciding with the sensor heating up. This value is about 20 ppm higher than it typically reads in a well ventilated room but plotting the ref values against the MQ-135 shows no relationship, the MQ-135 doesn't track the small variations the reference detector showed effectively, even if calibrated and with a replacement load resistor. Bluntly put, it's a crude detector that isn't designed to be quantitative. You can get it to read ~420 ppm CO2, but that's only because you're telling it that that's the atmospheric CO2 concentration, it will not effectively track changes. If you want an accurate reading for CO2 concentration, try an NDIR sensor. Was fun to mess around with though.