As an Amazon Associate, Akermin earns from qualifying purchases. Product links on this page are affiliate links.
A CO2 monitor works by passing infrared light through a small chamber of room air and measuring how much of a specific wavelength is absorbed. Carbon dioxide molecules absorb infrared at 4.26 micrometres, so the amount of light reaching the detector is a direct measure of how many CO2 molecules are in the sample. The instrument converts that to parts per million and displays it.
Almost every indoor CO2 monitor sold to homeowners uses this method, called non dispersive infrared or NDIR. It is the only low cost technique that is genuinely specific to carbon dioxide. Cheaper electrochemical cells and metal oxide sensors drift and cross react with other gases, which is why a listing that claims CO2 measurement but does not mention NDIR should be treated with suspicion.
The number itself is a ventilation proxy, not a health score. Outdoor air sits near 420 ppm. A room full of people with closed windows will climb past 1000 ppm within an hour or two, and that rise tells you exhaled air is accumulating. The monitor is really a people detector and a fresh air detector wearing a chemistry costume.
The sensor inside: why NDIR is the only method that counts
NDIR stands for non dispersive infrared. A small incandescent lamp or a heated filament emits broadband infrared. That light passes through a tube of sample air and then through an optical filter that lets only a narrow band around 4.26 micrometres reach the detector. Carbon dioxide absorbs strongly at that wavelength and almost nothing else in indoor air does, so the detector signal falls as CO2 rises.
The physics is Beer Lambert absorption: the fraction of light transmitted depends exponentially on the number of absorbing molecules in the path. The instrument does not count molecules one at a time. It measures a ratio of light intensities and solves for concentration using a calibration curve stored in its firmware. That curve is why an NDIR sensor needs occasional recalibration and why its accuracy figure, when published, is stated as a tolerance in ppm.
The alternative technologies are worth knowing because they appear in cheap listings. Electrochemical CO2 cells exist but are rare in consumer devices and have short lives. Metal oxide sensors respond to many gases and are sometimes labelled as CO2 sensors when they are really general reducing gas sensors. A listing that says only high sensitivity sensor without naming NDIR is telling you something by omission. The GOWENIC wall mount monitor is an example of a listing that does name an NDIR sensor, which is the first thing to look for.
Photoacoustic NDIR is a variant used in some small sensors. The infrared source is pulsed, the gas heats and cools, and a microphone hears the pressure wave. It is still infrared absorption, just detected differently. For the buyer, both are legitimate NDIR and both give a real CO2 number.
From absorption to a ppm number you can act on
The raw output of the detector is a voltage. Firmware converts that voltage to absorbance, then to concentration using the calibration curve. Two corrections usually happen before the display updates. The first is temperature and pressure compensation, because the number of molecules in the optical path depends on both. The second is baseline correction, often called ABC or automatic baseline correction, which assumes the sensor sees fresh outdoor air at some point and slowly re zeroes itself to roughly 400 ppm.
That baseline assumption is the most important thing to understand about a consumer CO2 monitor. If the device never sees fresh air, because it lives in a sealed bedroom or a grow tent, the automatic correction will slowly drag a genuinely high reading downward. A monitor that reads 800 ppm in a closed room after a week may be under reading. This is why the advice to air the room and let the monitor see outdoor air is not housekeeping, it is calibration.
The table below sets out how the common ppm bands are usually interpreted. The bands are general guidance from ventilation practice, not a product specification.
| Reading (ppm) | What it usually means | Typical action |
|---|---|---|
| Around 420 | Background outdoor concentration | None; this is the floor the sensor should return to |
| 600 to 800 | Well ventilated occupied room | None |
| 800 to 1000 | Occupied room with moderate ventilation | Consider opening a window or running mechanical ventilation |
| 1000 to 1500 | Poor ventilation, exhaled air accumulating | Ventilate; expect drowsiness and reduced concentration |
| Above 1500 | Very poor ventilation | Ventilate immediately; check whether the room is being used as designed |
These bands describe ventilation, not toxicity. Carbon dioxide at 2000 or 3000 ppm is not poisonous. The reason to act is that the same stale air carries exhaled moisture, body odour, and any airborne particles or pathogens the occupants are producing. The CO2 number is a convenient tracer for all of it.
What the display adds, and what it hides
Most consumer monitors pair the NDIR cell with a temperature and humidity sensor, and some add particulate, formaldehyde or total volatile organic compound channels. The HSCLPM 16 in 1 monitor is an example of a multi channel device that reports CO2 alongside PM1.0, PM2.5, PM10, HCHO and TVOC. That breadth is useful, but it changes what the device is. A multi channel air quality monitor is not a better CO2 monitor; it is a different instrument with a CO2 channel inside it.
The reason to keep them separate in your head is maintenance. The CO2 channel needs its baseline correction. The particulate channel needs a clean optical path and a fan that has not clogged. The VOC channel needs a sensor that may saturate and recover slowly. When one channel drifts, the others do not tell you. A single number on a colour display can look authoritative while one of its inputs is stale.
Alarms are a separate question. The GOWENIC monitor listing states that it triggers audible and visual alarms when levels exceed thresholds. That is useful in a classroom or a meeting room where nobody is watching a screen. But an alarm is not a safety device in the regulatory sense. It does not replace a ventilation design, and it does not detect carbon monoxide, which is a different gas with a different sensor. If your concern is combustion, a CO2 monitor is the wrong instrument.
Portable monitors add battery life as a specification. The Dienmern 8 in 1 and the Dienmern 5 in 1 both publish a 1600 mAh battery and up to 72 hours of use. The 5 in 1 listing also publishes a detection range of 400 to 5000 ppm and an accuracy of plus or minus 50 ppm. Those are the kind of figures to look for; where a listing is silent, the honest answer is that the figure is not stated in the listing bullets.
The common mistake: treating ppm as a personal health score
The most frequent error is to read a CO2 number as a direct measure of how polluted the air is. It is not. A room can read 700 ppm and be full of fine particulate from cooking, or read 1400 ppm and contain nothing but exhaled breath. The CO2 number tells you about ventilation and occupancy. It does not tell you about particles, nitrogen dioxide, carbon monoxide, or mould spores.
The second mistake is to place the monitor badly. CO2 is exhaled at roughly body temperature and is slightly denser than air, but in a normally mixed room it disperses quickly. The real placement issue is proximity to the source and to the sensor’s own baseline. A monitor on a desk beside your face will read higher than one across the room. A monitor inside a closed cupboard will never see fresh air and will drift. The correct position is at breathing height, away from windows, doors and direct sunlight, in the room you actually occupy. The guidance on using a CO2 monitor in a bedroom covers the overnight case, where the door is closed and the baseline correction has the least fresh air to work with.
The third mistake is to assume the number is absolute. Consumer NDIR sensors are usually accurate to tens of ppm, not single ppm. A reading of 812 is not meaningfully different from 798. Look at the trend over ten minutes, not the last digit.
Calibration, drift and the maintenance nobody mentions
NDIR sensors drift. The infrared source ages, the detector responds differently with temperature, and dust settles on the optical surfaces. Manufacturers handle this in two ways. Some run automatic baseline correction in firmware, assuming the lowest reading seen over a period is fresh air. Others expect the user to calibrate against a known reference, usually outdoor air at roughly 400 to 420 ppm, by holding a button in fresh air.
If your monitor has a calibration function, use it. Take the device outside or to a well ventilated space, let it stabilise, and follow the procedure. If it has automatic baseline correction and lives in a room that is never aired, the correction is working against you. The practical fix is to put the monitor outside for an hour occasionally, or to choose a model that lets you disable the automatic correction. The calibration guide walks through the procedure for the common button sequences.
Dust is the other maintenance item. A monitor with a fan, which most multi channel units have, will pull dust into the optical chamber. A CO2 only NDIR cell with no fan is less affected because it relies on diffusion, but it responds more slowly. Neither is maintenance free. If the reading stops moving at all, suspect the sensor rather than the air.
Finally, sensor life. The listing for the Dienmern 5 in 1 names a Swiss made CO2 sensor, which is a useful signal about the sensing element, but no consumer listing publishes a replacement interval. The honest position is that NDIR cells typically last several years and then drift beyond their stated accuracy. When a monitor that used to track your cooking and your meetings stops responding, it is probably the cell, not the room.
Where a CO2 monitor stops being the right instrument
A CO2 monitor is the right tool when the question is how well a space is ventilated for the people in it. It is the wrong tool for almost every other air quality question.
- It does not detect carbon monoxide. CO and CO2 are different molecules with different absorption bands. A CO2 alarm is not a CO alarm, and a CO2 reading will not rise because of a faulty boiler.
- It does not measure particulate matter. If you are concerned about smoke, traffic pollution, cooking fumes or wildfire smoke, you need a particle counter. The PM2.5 monitor guide covers what to look for.
- It does not measure volatile organic compounds or formaldehyde specifically. Multi channel devices that report HCHO and TVOC use separate sensors, and those sensors have their own limitations. The VOC monitor guide explains why a single TVOC number is hard to interpret.
- It is not a confined space gas detector. A portable CO2 monitor is not certified for entry into sewers, tanks or other confined spaces, and it does not measure oxygen or combustible gas.
Where the CO2 monitor is genuinely excellent is in the ordinary rooms where people gather and nobody thinks about ventilation: bedrooms, classrooms, meeting rooms, home offices. In those settings the number is a direct, understandable signal that the room needs air. The meeting room guide shows how to use the trend rather than the absolute value to decide when to break.
If you want the deeper physics of the sensor and its limitations, the technical explanation of CO2 monitors goes further into the optical and firmware details. And if you are choosing between models, the selection guide sets out which published specifications actually matter.
The number is a question, not an answer
A CO2 monitor tells you one thing well: how much of the air in the room has recently been in somebody’s lungs. That is a remarkably useful thing to know, because it stands in for ventilation, occupancy and freshness all at once. It is not a measure of pollution, and it is not a safety device. Read it as a prompt to open a window, and it will serve you well. Read it as a verdict on air quality, and it will mislead you.
Frequently Asked Questions
Does a CO2 monitor detect carbon monoxide?
No. Carbon monoxide and carbon dioxide are different molecules and absorb infrared at different wavelengths. A CO2 monitor's sensor is tuned to the 4.26 micrometre band where CO2 absorbs. It will not respond to CO, and a CO2 alarm is not a substitute for a CO alarm. If you have combustion appliances, you need a separate carbon monoxide alarm.
Why does my monitor read around 400 ppm outside and 1200 ppm in a meeting room?
Outdoor air is roughly 420 ppm CO2. In a closed meeting room, each person exhales CO2 at a rate that raises the concentration steadily. 1200 ppm after an hour is normal for a full room with no ventilation. The rise is the signal to ventilate. It is not a fault in the monitor.
Do I need to calibrate a consumer CO2 monitor?
Many models calibrate themselves automatically by assuming the lowest reading they see is fresh air. That works if the monitor regularly sees outdoor air. If it lives in a sealed room, the automatic correction can drag readings down over time. Some models offer a manual fresh air calibration; use it if it exists. The procedure is usually in the manual and is covered in our calibration guide.
Is a higher ppm reading dangerous?
Carbon dioxide at the levels found in homes and offices, up to a few thousand ppm, is not toxic. The reason to act on a high reading is that it indicates poor ventilation, which also means exhaled moisture, odours and any airborne particles are accumulating. Treat 1000 ppm as a prompt to ventilate, not as a health threshold.
What does NDIR mean and why does it matter?
NDIR stands for non dispersive infrared. It is the technique that measures CO2 specifically by its absorption of infrared light. It is the only low cost method that is genuinely selective for CO2. Listings that claim CO2 measurement without naming NDIR may be using a general gas sensor that cross reacts with other vapours, and the number should be treated with caution.
Can a CO2 monitor tell me if my air is clean?
No. It tells you about ventilation and occupancy. It does not measure particulate matter, volatile organic compounds, carbon monoxide or mould. If you want a picture of overall air quality, you need a multi channel monitor or separate instruments. A CO2 reading of 700 ppm in a room full of cooking smoke is still a well ventilated room with dirty air.