So many different types of sensors have become available in the market. For years we have seen and used Single (mainly) and Dual-Beam NDIR sensors for CO2 detection, but now another technology has surfaced, the Photoacoustic.
I decided to do a quick comparison of the different technologies in order to determine which one is the best and what are the differences if there is one. Price always affects the Bill of Material BOM, so we need to make a wise choice depending on the application of the sensor (commercial, real estate, industrial, scientific, etc).
Single and Dual Beam NDIR
CO2 is a gas with an asymmetric molecular structure that has strong absorption of infrared. This is the reason we use a Non-Dispersive Infrared NDIR sensor which is based on tunable diode laser spectroscopy.
The key components of the NDIR sensor are an infrared (IR) source, gas chamber, and an infrared detector (photodiode). When the light wave passes through measured gas (CO2), the intensity of the light wave will be significantly weakened. The intensity attenuation is related to the concentration of measured gas. Generally, there are two main categories: Single and Dual Beam (sometimes also called single or dual-channel) according to the number of the beam of infrared detectors.
Dual-channel CO2 sensors have an additional internal reference beam within the sensor itself, which is used for self-calibrating. The IR detector is integrated with two photosensitive devices and two narrowband filters.
The reference beam is only activated during recalibration, and the reading of the other constantly used beam is adjusted to match it. Single-channel CO2 sensors use a technique called Automatic Baseline Calibration ABC. It uses a single beam and logs readings over time, and the lowest CO2 reading is assumed to be the ambient air ~400 ppm. The ABC logic is used by sensors mostly at night.
This technology is actually a miniaturized spectroscopic CO2 sensor that is based on the photoacoustic effect.
Low-cost photoacoustic-based sensor comprised of a microphone in a hermetically sealed chamber filled with CO2. Photoacoustic spectroscopy is the measurement of the effect of absorbed electromagnetic energy (light) on the matter by means of acoustic detection. The absorbed energy from the light causes local heating, generating a thermal expansion which creates a sound or pressure wave.
The sensors I have used for the experiment are the Sensirion SCD30 (Dual Beam NDIR), Senseair S8 (Single Beam NDIR), and Sensirion SCD41 (Photoacoustic). I took into consideration all the available calibration methods to ensure fairness among the results. The sensors were placed in my 20m3/706ft3 office where I work during the day. There is natural ventilation 24h as I leave a window slightly open. Below you can see the specifications for each sensor and the 48h measurements graph.
The correlation results among all the sensors are very encouraging. However, we can see an average difference between the Single Beam and Dual Beam sensor of 97 ppm, which is quite high. The average difference between the Photoacoustic and the Dual Beam is 66 ppm. At first, the photoacoustic sensor follows the dual-beam, but then it follows the single beam sensor.
Interestingly, the Dual Beam sensor never reached the atmospheric CO2 background conditions during both nights of around 411 ppm (the window was slightly open in the office and next to the sensors). Finally, I have to include the common cliché, further research is needed (FRIN).
The Photoacoustic sensor (~30€) is more economic than the Dual Beam NDIR (~39€) and equal to the Single Beam NDIR (~31€) which makes it very competitive and although the market is fixed to the NDIR technology at the moment, I think we will see a shift as the benefits (size and price) we get from the photoacoustic technology are important. There are other CO2 photoacoustic sensors available in the market, like the XENSIVTM PAS CO2 by Infineon, which on specs offers better accuracy than the Sensirion SCD4x, but I haven’t had the chance to test it.