Gas Sensors for Air Quality and Leak Detection
Product Specifications
Product Overview
An air quality reading is only useful when the number, the unit and the conditions behind it are all stated. Three sensing principles cover most of the products we quote, and they differ in what they actually measure. A photoacoustic NDIR sensor measures the CO2 molecule itself: an infrared source is pulsed, the gas absorbs part of the light, the absorption heats the gas and the pressure wave is picked up by a microphone inside the package. A metal oxide sensor measures a change in the resistance of a heated tin dioxide or tungsten oxide layer, which responds to many reducing gases at once and therefore reports a sum, not a species. An electrochemical or index output sensor sits between the two, converting a raw signal into a calibrated index or a parts per million figure with an algorithm.
The distinction matters because the numbers are not interchangeable. The SCD41 reports 400 to 5,000 ppm with an accuracy of plus or minus 40 ppm plus 5 percent of the reading, and that figure is tied to a specified range. The SGP41 does not report ppm at all: it reports a VOC index and a NOx index on a 1 to 500 scale, where 100 is the average of the last 24 hours, values below 100 mean the air improved and values above mean it got worse. A metal oxide cell such as the MQ-2 reports a resistance ratio that has to be converted with a calibration curve for a named gas, and its sensitivity curve is drawn for one concentration of one gas in a standard test condition.
Power is the second decision. A photoacoustic CO2 sensor draws 15 mA average in periodic mode and 205 mA peak at 3.3 V, so the supply has to hold its voltage through the measurement burst and the regulator has to be chosen for the peak, not the average. An index sensor draws 34 microamps idle and heats only during a measurement, which suit a battery node that wakes once a minute. A metal oxide cell is the opposite: its heater runs at 5 V and about 150 mA continuously, and the MQ-7 adds a second complication, because it is measured at a low heater voltage and cleaned at a high one on a 60 second and 90 second cycle.
Preheat and drift are the terms that decide whether a reading is believable in the field. A metal oxide cell needs 24 to 48 hours of preheat before its resistance is stable, and its sensitivity moves with humidity and temperature, which is why the data sheet curves are drawn at 20 C and 65 percent relative humidity. A photoacoustic sensor is calibrated in production and uses an automatic self calibration that assumes the space sees about 400 ppm fresh air once a week; a sealed room that never ventilates will drift. An index sensor needs its own conditioning command and a switch on time before the algorithm is valid.
The table below covers the three groups we quote most often: the optical CO2 parts for rooms and ducts, the index sensors for demand controlled ventilation, and the heated cells for leak and alarm products. Range, accuracy, supply, current, interface and package come from the maker data sheet for each order code.
Key Specifications
| Parameter | Photoacoustic CO2, SCD41 | VOC and NOx index, SGP41 | Heated metal oxide, MQ series |
|---|---|---|---|
| What is measured | CO2 concentration by infrared absorption | VOC index and NOx index, algorithm output | Resistance change of a heated tin dioxide layer |
| Range | 400 to 5000 ppm specified, 0 to 40000 ppm output | 1 to 500 index points | 20 to 2000 ppm CO on the MQ-7, 300 to 10000 ppm combustible gas on the MQ-2 |
| Accuracy or sensitivity | plus or minus 40 ppm plus 5 percent of reading | Index scale, average air maps to 100 | Resistance ratio per gas from the sensitivity curve |
| Supply | 2.4 V to 5.5 V | 1.7 V to 3.6 V | 5.0 V plus or minus 0.1 V for the heater |
| Current | 15 mA average, 205 mA peak at 3.3 V | 34 microamps idle, heater pulsed during measurement | About 150 mA at 5 V, 350 mW to 900 mW heater power |
| Interface | I2C | I2C | Analog, through an external load resistor |
| Response time | 60 s at 63 percent | Under 10 s for VOC, under 250 s for NOx | Set by the heater cycle and the load resistor |
| Warm up | Calibrated in production, self calibration in use | Conditioning command, then 60 s switch on | 24 to 48 hours preheat before the first stable reading |
| Package | 10.1 x 10.1 x 6.5 mm, reflowable | 2.44 x 2.44 x 0.9 mm DFN | Plastic cap, 6 pin, through hole |
Included Models and Order Codes
The table lists the order codes we quote most often for indoor air quality, ventilation control and gas leak alarms. Range, accuracy or sensitivity, supply, current, interface and package come from the maker data sheet for that code. Where a part reports an index rather than a concentration, the table says so, because an index cannot be converted to ppm by arithmetic.
| Part Number | Principle | Target | Range | Supply and current | Interface | Package |
|---|---|---|---|---|---|---|
| SCD41 | Photoacoustic NDIR | CO2 | 400 to 5000 ppm specified | 2.4 to 5.5 V, 15 mA average | I2C | 10.1 x 10.1 x 6.5 mm |
| SCD40 | Photoacoustic NDIR | CO2 | 400 to 2000 ppm specified | 2.4 to 5.5 V, 15 mA average | I2C | 10.1 x 10.1 x 6.5 mm |
| SGP41 | MOX with algorithm | VOC and NOx index | 1 to 500 index points | 1.7 to 3.6 V, 34 uA idle | I2C | 2.44 x 2.44 x 0.9 mm DFN |
| SGP40 | MOX with algorithm | VOC index | 1 to 500 index points | 1.7 to 3.6 V, 34 uA idle | I2C | 2.44 x 2.44 x 0.9 mm DFN |
| CCS811 | MOX with MCU | eCO2 and TVOC | 400 to 8192 ppm eCO2, 0 to 1187 ppb TVOC | 1.8 to 3.3 V, 26 mA measuring | I2C | 2.7 x 4.0 x 1.1 mm LGA |
| MQ-2 | Heated SnO2 | Combustible gas and smoke | 300 to 10000 ppm | 5 V heater, about 150 mA | Analog | Plastic cap, 6 pin |
| MQ-4 | Heated SnO2 | Methane and natural gas | 300 to 10000 ppm | 5 V heater, 900 mW max | Analog | Plastic cap, 6 pin |
| MQ-7 | Heated SnO2, cycled | Carbon monoxide | 20 to 2000 ppm | 5 V 60 s then 1.4 V 90 s, 350 mW | Analog | Plastic cap, 6 pin |
| MQ-8 | Heated SnO2 | Hydrogen | 100 to 10000 ppm | 5 V heater, 800 mW class | Analog | Plastic cap, 6 pin |
Key Features
CO2 measured, not inferred – the SCD41 absorbs infrared light at the CO2 band, so the reading does not drift when other gases are present, which is what an equivalent CO2 estimate cannot promise.
Accuracy stated against the reading – plus or minus 40 ppm plus 5 percent of reading means the error is small at 800 ppm and grows with concentration, and the quotation states it that way instead of quoting one number.
Index output needs no gas bottle – the SGP41 maps the raw signal to a 1 to 500 index with an algorithm that runs on the host, so there is no field calibration gas and no per unit adjustment.
Two gases from one package – the SGP41 reports VOC and NOx from the same heater and the same I2C address, which covers cooking odour and traffic exhaust on one board.
Heater cycle written down – the MQ-7 is measured at 1.4 V and cleaned at 5.0 V on a 60 s and 90 s cycle, and the quotation carries both phases because a constant 5 V heater gives a reading that is not the one the curve describes.
Reflowable optical package – the SCD4x is a reflowable SMD part with an integrated humidity and temperature sensor, so the reading is compensated without a second part on the board.
Applications
These parts are quoted for products that measure air quality or watch for a gas leak, typically where:
Room and duct CO2 for ventilation – a CO2 sensor inside a room unit or a return duct drives a damper or a fan speed, and the 400 to 5000 ppm range covers both a full meeting room and a fresh air reference.
Demand controlled ventilation and air purifiers – a VOC index sensor moves the fan when the odour rises, and the index scale means the control threshold does not have to change with the gas mixture.
Gas leak alarms and portable detectors – a heated metal oxide cell watches for combustible gas or carbon monoxide at a cost that suits a single point alarm, with a buzzer and a relay on the same board.
The sensor sits next to the regulator and the I2C pull ups, so it is quoted with the SUPERB PCB and the SUPERB PCBA build for that assembly.
Sourcing and Cross-Reference
Selection starts with the question the product has to answer. If the answer is a concentration in ppm, the part has to be calibrated against that gas, which points to the optical CO2 sensors or to a cell with a published curve. If the answer is more or less than usual, an index sensor is the cheaper and more honest choice, because it does not claim a ppm figure it cannot support.
The second check is the enclosure. A photoacoustic sensor needs a gas path to the room and a supply that holds its voltage through a 205 mA peak, and a metal oxide cell needs a mesh or a sintered cap that lets gas in but keeps dust and splashes out. Humidity above 95 percent relative and condensation on the sensing layer will change a metal oxide reading, and a conformal coat over the mesh will stop it working altogether.
Where an order code is short of supply or has been marked not recommended for new builds, the replacement is checked on principle, range, supply, current, interface and package footprint at once. Two CO2 sensors with the same range can differ by a factor of two in peak current, and two index sensors can differ in whether the algorithm runs on the host or inside the part. The route for that comparison is on our source obsolete components page, and the pin level check is on pin to pin cross reference.
For volume builds, SUPERB Assembly support covers the gas path in the housing, the mesh or membrane over the sensing element and the thermal isolation of the heater from the rest of the board.
Q: Can you tell me which of these reports a real ppm value and which reports an index, and quote the supply parts for the 205 mA peak of the CO2 sensor?
A: Yes. Send the target gas, the range you have to report and the supply budget, and the SUPERB Tech team will say which principle fits and quote the matching code from the maker data sheet.
Why Buyers Source Gas Sensors Through SUPERB
- Principle stated on the quotation: an index part and a ppm part are never quoted as if they were the same thing, which is the most common cause of a product that has to be restarted.
- Peak current carried into the power budget: the 205 mA CO2 burst is quoted next to the 15 mA average, so the regulator and the decoupling are chosen once.
- Heater cycle documented: the MQ-7 timing, the preheat time and the humidity limits are given with the part, because those are what a field failure usually traces back to.
- Cross reference checked on six parameters: principle, range, supply, current, interface and footprint are compared before a substitute is offered, not just the package.
Related Product Families
This page is part of the component range we supply for industrial, telecom, medical and defence builds. Continue with: Sensors and detectors, Humidity and temperature sensors, Ambient light and proximity sensors, Varistors and gas discharge tubes, Low noise LDO regulators, Sourcing obsolete components, BOM shortage recovery.