Frequently Asked Questions
Q1: Do MQ gas sensors require calibration?
A1: Yes. Out of the box, MQ sensors will produce readings, but to get meaningful values (like ppm) you should calibrate them. Calibration involves determining the sensor’s baseline resistance in clean air (R₀) and adjusting for known reference points of the target gas. For hobby projects, a full calibration might not be feasible (since you need known test gas concentrations). Instead, many users approximate calibration by exposing the sensor to fresh outdoor air (assuming zero or low pollutant level) and maybe a common source (like a cigarette or a lighter) to get a second point. At minimum, allow the sensor to burn-in for 24+ hours, then note the clean-air analog value as your baseline. You can then use the calibrate_linear filter in ESPHome (as shown above) to map voltages to ppm based on datasheet curves or empirical observation. Keep in mind this will be an approximation – MQ sensors are not laboratory-accurate without rigorous calibration, but for relative changes and threshold alerts they work well enough.
Q2: Can I run an MQ sensor on 3.3V: The short answer is not recommended for most. The built-in heater of MQ sensors is designed for 5V supply to reach the proper operating temperature (around 200–400 °C). If you only provide 3.3V, the heater will be cooler, and the sensor’s sensitivity will shift (it may become less sensitive or not operate in the intended range). The MQ-7 is a special case that actually uses a low voltage in part of its cycle, but generally you should provide 5V to the sensor’s VCC. However, the output of the sensor’s voltage divider can be read by a 3.3V ADC pin – that’s fine as long as the output doesn’t exceed 3.3V. Most MQ modules put the sensor in a divider with a load resistor such that in clean air the analog voltage is somewhere around mid-supply (e.g. ~2.5V on 5V supply).
It can go up near 5V in presence of certain gases. So if using with an ESP (3.3V ADC), you have two choices:
1. Power the sensor heater with 5V, but use a voltage divider or level translator on the analog output to scale 0–5V down to 0–3.3V for the MCU. (E.g., two resistors 2:1 ratio). natively, power the sensor with ~4.7V so that its maximum analog output is about 3.3V – but this is tricky and still not guaranteed.
In practice, many have connected MQ sensors directly to ESP8266 A0 which has an internal 220k/100k divider (scaling 0–1V input to 0–3.3V ADC range) – on such boards, providing into A0 and powering it with 5V works because the NodeMCU’s built-in divider protects the ADC. For ESP32, you can set ADC attenuation to handle up to 3.3V, but you must ensure the MQ analog never exceeds that. So, using an external divider (e.g., 2:1) on the analogue line is a safe bet when powering the sensor at 5V.
Q3: Why do I need to preheat MQ series sensors for 24 hours?
A3: Upon first use (or after long storage), the sensor’s oxide surface contains residues and moisture that need to burn off and stabilize. During the first day of powering an MQ sensor, you’ll likely see its readings drift downward gradually as it settles. This burn-in ensures that when you later measure gas, the baseline is stable. For day-to-day usage, you also need a short warm-up each time you power it on – typically 1–5 minutes to reach equilibrium. That’s why in many applications, MQ sensors are run continuously rather than power-cycled, especially for safety monitoring (you don’t want a 2-minute wait dealing a gas leak). The 24-hour initial preheat is mainly for new sensors. If your sensor has been offline for weeks, a few hours of preheat might be warranted again. Tip: During burn-in, place the sensor in a well-ventilated area because it might outgas a little (you might notice a faint “hot metal” smell). After burn-in, do your calibration or baseline reading in clean air, and you’ll get more consistent results long-term.
Q4: How long do MQ series sensors last?
A4: These sensors are rated for a lifespan of around 2 to 5 years of continuous use【54†L177-L184】. Over time, the sensitive material can degrade (for example, sensor pores get contaminated or the heater efficacy changes). In practical terms, they don’t typically fail outright at 5 years, but their calibration may drift – they become less senstive to a higher baseline. High exposure to target gases (especially corrosive gases like H₂S or high concentrations of solvent vapors) can shorten their life or poison the sensor. For instance, an MQ-136 in a sewer might age faster due to constant H₂S. Also, silicone vapors are known to poison MOS sensors – exposure to silicone sealant fumes can coat the sensor and kill its response. So, avoid exposing MQ sensors to things like silicone glue cure vapors or leaded gasoline, etc. If accuracy is crucial, you should recalibrate the sensor every 6 months or so. For hobby/home monitoring, just keep an eye on the baseline over time – if you notice it drifting significantly (e.g., the sensor reads much higher in clean air than it used to), it might be time to replace the sensor. The good news is they’re inexpensive to replace.
Q5: My MQ sensor is giving fluctuating readings – is something wrong?
A5: Some amount of fluctuation is normal. Remember, MQ sensors essentially measure a resistance that can be influenced by temperature and humidity in addition to gas. If your environment’s conditions aren’t constant, the readings will wobble. For example, a breeze or draft can bring fresh air and change the reading slightly, or a rise in humidity might increase the sensor’s base resistance. That’s why using filtering in software (like the moving average or median filters in ESPHome) is recommended to smooth out short-term noise. However, large oscillations might indicate an issue: check your wiring and ADC stability. Ensure the sensor’s analog output is connected properly to a stable ADC reference – noise on the power line can cause ADC noise. It’s also important that the sensor is not oscillating due to a power issue (the heater drawing current shouldn’t sag your 5V rail – use a proper power supply capable of ~200 mA per sensor). Another factor: warm-up drift – if the sensor hasn’t fully warmed up, you’ll see a downward drift. Always compare readings after a consistent warm-up period. In summary, minor fluctuations = normal (you can treat anything under, say, 5% variation as noise), but major swings = investigate power, baseline, or actual gas presence. Often, people map the analog value to qualitative levels (e.g., “Good”, “Moderate”, “High”) rather than rely on an exact ppm at each second.
Q6: Can I use one MQ sensor to detect multiple gases?
A6: An MQ sensor will respond to multiple gases, but it can’t tell them apart. For instance, MQ-9 reacts to both CO and methane – if it alarms, you won’t know which one (it just means something it’s sensitive to is high). Similarly, MQ-135 might spike due to high CO₂ or due to someone spraying perfume – the reading doesn’t label the source. If your goal is general air quality or broad safety (just need to know if there’s a dangerous level of any gas), one sensor can be sufficient as a catch-all (especially MQ-2 or MQ-135 which are broad-spectrum). But if you need to monitor specific gases (like you specifically want to know CO level vs LPG level), you should use the dedicated sensors for each or multiple MQ sensors. In our guide, we provided each MQ’s forte – combining their data can even allow some crude differentiation (e.g., if MQ-7 is high but MQ-5 is normal, it’s likely CO; if both MQ-5 and MQ-7 are high, maybe there’s a fire producing both CO and propane leak). That’s an advanced analysis though. Generally, use the right sensor for the gas of concern for the best results. MQ sensors are cheap, so deploying two or three in different spots (or different types in same spot) is often a good strategy for comprehensive coverage.
Q7: What is the best way to physically mount and protect MQ sensors?
A7: Treat MQ sensors a bit like smoke alarms: place them in the room without obstruction. They come as open devices (with a mesh that must be exposed to air). Do not enclose the sensor in a tight box without vents – you’ll render it ineffective. If you need to put it in an enclosure (for aesthetics or safety), ensure there are plenty of vents or holes near the sensor head. Also consider orientation: the sensors can get warm. Many MQ modules have mounting holes – you can screw them onto a panel or use standoffs. Keep them away from water or excess dust. If using outdoors, shield them from rain and direct wind (wind can cause rapid changes or cool the sensor). A splash-proof, vented cover is ideal for outdoor use. And as mentioned before, give them some distance from high EMI sources or vibrations not overly sensitive to electromagnetic noise, a very noisy environment could affect the analog reading. One more tip: height placement matters for certain gases (CO – head height, LPG – low near floor, smoke – high). Refer back to each sensor’s section where we gave placement advice, and mount accordingly for your application.
Q8: Are MQ sensors suitable for critical safety, like as a carbon monoxide alarm?
A8: Officially, no. For life-critical detection (such as CO alarms required by code, explosive gas detectors in homes, or OSHA-compliant monitors), you should use certified detectors. MQ sensors are great for DIY supplementary non-critical alerts. They are inexpensive and not calibrated or tested to the rigorous standards and certification. For example, a UL-listed CO alarm has to reliably trigger at certain CO concentrations within a time window – an MQ-7 on your ESP might trigger, but it might also miss a low-level CO buildup or false alarm occasionally due to drift. Use it as an extra layer of information – it can even be integrated with Home Assistant to send you notifications or log trends (which many commercial alarms can’t do). But do not solely rely on it for protecting life. In a scenario like a propane leak, an MQ-6 can sense it and maybe shut a valve via your automation – that’s awesome as a proactive measure. Still, have a standalone combustible gas alarm or at least ensure your setup is failsafe (e.g., runs on backup power, tested frequently). Think of MQ sensors as “prosumer” level – very useful but requiring the user to understand their limitations. Many people indeed build their own smoke or CO alarms with MQ sensors for fun, but they themselves will acknowledge it’s not a substitute for a certified unit.
