6 ms·

I started taking IAQ seriously when I soundproofed my home office. The soundproofing made the room nearly airtight, so I had to think about (at the very least)

I started taking IAQ seriously when I soundproofed my home office. The soundproofing made the room nearly airtight, so I had to think about (at the very least) CO2 buildup within the office. I've hit an interesting problem regarding CO2 monitoring: calibration and validation. Every IAQ CO2 meter I've come across allows self-calibration and/or calibration by exposure to "open air" which is assumed to be about 400 ppm CO2. But none of the meters provide any means to calibrate / validate CO2 levels other than the reference level. For example, suppose I buy two CO2 meters from Amazon. They can both tare to 400 ppm. But once the CO2 in my office builds up, one might read 1200ppm while the other reads 800ppm. According to medical literature, 800 ppm vs. 1200 ppm is a pretty significant difference for human mental performance. If the CO2 level is actually 1200 ppm, I want to crank the office ventilation system, even though it means higher background noise. At just 800 ppm, I may prefer the quiter work environment. Another downside to this difficulty in testing is that it's hard to truly evaluate a particular meter's performance. Which seriously limits side-by-side comparisons of different models, and makes it harder to hold manufacturers / retailers accountable for poor performance.

joekrill 6y ago
Not to derail the conversation, but I'm curious about how you soundproofed your home office. Do you have any information about how you went about that and what was involved? I've been looking into this myself recently.
- DoofusOfDeath 6y ago
  One of the key takeaways is that the approach depends heavily on the space you're trying to soundproof, and just how much sound reduction you need at various frequencies. Feel free to DM me (see my profile) if you'd like to talk specifics.
- turtlebits 6y ago
  Depends on what you're looking for but adding an additional layer of drywall (5/8) with acoustical sealant sandwiched is an inexpensive way reduce sound transmission.
ChuckMcM 6y ago
Agreed, the difference between "real" CO2 meters and consumer CO2 meters is really stark. Given that you can get nitrogen and CO2 in tanks, I expect one could build a calibration setup with a fixed volume tank, two metered gas ports, and some software. Start by purging the tank with nitrogen, then add CO2 for the 400ppm baseline. Calibrate there, then add increments of 100ppm (or 50ppm) CO2 and record the readings up to say 1500ppm total. Then you have a calibration table for that meter that will let you translate from its indicated ppm to the real ppm.
- DoofusOfDeath 6y ago
  That's roughly my understanding about how these devices are actually tested in labs. Probably not something I'd tackle on my own, but it would be awesome if Consumer Reports or a major newspaper hired a lab to do this. With so many people working from home now, I would think it's an increasingly relevant issue.
- nickff 6y ago
  That won't necessarily make the meters match. The sensors can be perfectly calibrated for CO2, but still read differently because most of the sensors are sensitive to a variety of other carbon-containing gasses. Most consumer CO2 detectors use a sensor which measures the IR absorption of its surroundings. These sensors have varying cross-sensitivities to a bunch of other gasses including carbon monoxide, propane, alcohol, etc. Calibrating them for CO2 will not compensate for different responses to CO, which is much more IR-absorbent in many of the same frequency bands.
- GrumpyYoungMan 6y ago
  That's unnecessary since there's pre-calibrated mixtures referred to, unsurprisingly, as "calibration gas" readily available. Googling for "carbon dioxide calibration gas" will find you what you need as well as the appropriate pressure regulators. Note that even the professional meters require calibration on a regular basis; for example the calibration schedule for the Fluke 975 for CO2 is yearly and for carbon monoxide is monthly.
throwaway9d0291 6y ago
> Every IAQ CO2 meter > For example, suppose I buy two CO2 meters from Amazon It sounds like you're looking at retail consumer products, which I think is the problem. The Sensirion SCD30 _sensor_ [0] for example allows forced recalibration [1] to an arbitrary concentration. Also, since you call it an "IAQ" CO2 meter, it might be that these aren't CO2 meters at all. They could be VOC meters that make up a projected CO2 reading based on the presence of other chemicals in the air. You need to make sure the sensor is NDIR-based and actually measures CO2. If you want accuracy, don't buy retail sensors that don't come with spec sheets. I'd find the sensor you want first then find or build a meter based on it. For example the SCD30 can be connected to an ESP32 and connected to WiFi using ESPHome [2], which can be set up with some YAML files. > self-calibration Just by the way, self-calibration is what you want. Real CO2 sensors drift over time and you definitely want them recalibrated at least once a year. Auto-calibration takes away that pain and when implemented in a suitable way, really works. Many calibration routines are not suitable for residential environments as they're based on the assumption that they will see pure fresh air at least once a day. This is a safe assumption in a business environment because most AC systems have an overnight fresh air purge. In a residential environment though, someone might not open the window once a day. Again referencing Sensirion, their auto-calibration cycle looks for two, hour-long valleys in the CO2 readings over the previous two weeks that are within a few percent of each other and uses those as a fresh air reference. In my experience, this works well at home. > Another downside to this difficulty in testing is that it's hard to truly evaluate a particular meter's performance. Which seriously limits side-by-side comparisons of different models, and makes it harder to hold manufacturers / retailers accountable for poor performance. You "just" need a reference to compare to. If you have a look at academic literature, they typically use a lab-grade setup to compare. You're right though, it's difficult for a regular home user to evaluate a sensor like this. [0]: https://www.sensirion.com/en/environmental-sensors/carbon-dioxide-sensors/carbon-dioxide-sensors-co2/ https://www.sensirion.com/en/environmental-sensors/carbon-di... [1]: https://www.sensirion.com/fileadmin/user_upload/customers/sensirion/Dokumente/9.5_CO2/Sensirion_CO2_Sensors_SCD30_Interface_Description.pdf https://www.sensirion.com/fileadmin/user_upload/customers/se... [2]: https://esphome.io/ https://esphome.io/

joekrill 6y ago

Not to derail the conversation, but I'm curious about how you soundproofed your home office. Do you have any information about how you went about that and what was involved? I've been looking into this myself recently.

DoofusOfDeath 6y ago
One of the key takeaways is that the approach depends heavily on the space you're trying to soundproof, and just how much sound reduction you need at various frequencies. Feel free to DM me (see my profile) if you'd like to talk specifics.
turtlebits 6y ago
Depends on what you're looking for but adding an additional layer of drywall (5/8) with acoustical sealant sandwiched is an inexpensive way reduce sound transmission.

ChuckMcM 6y ago

Agreed, the difference between "real" CO2 meters and consumer CO2 meters is really stark. Given that you can get nitrogen and CO2 in tanks, I expect one could build a calibration setup with a fixed volume tank, two metered gas ports, and some software. Start by purging the tank with nitrogen, then add CO2 for the 400ppm baseline. Calibrate there, then add increments of 100ppm (or 50ppm) CO2 and record the readings up to say 1500ppm total. Then you have a calibration table for that meter that will let you translate from its indicated ppm to the real ppm.

DoofusOfDeath 6y ago
That's roughly my understanding about how these devices are actually tested in labs. Probably not something I'd tackle on my own, but it would be awesome if Consumer Reports or a major newspaper hired a lab to do this. With so many people working from home now, I would think it's an increasingly relevant issue.
nickff 6y ago
That won't necessarily make the meters match. The sensors can be perfectly calibrated for CO2, but still read differently because most of the sensors are sensitive to a variety of other carbon-containing gasses. Most consumer CO2 detectors use a sensor which measures the IR absorption of its surroundings. These sensors have varying cross-sensitivities to a bunch of other gasses including carbon monoxide, propane, alcohol, etc. Calibrating them for CO2 will not compensate for different responses to CO, which is much more IR-absorbent in many of the same frequency bands.
GrumpyYoungMan 6y ago
That's unnecessary since there's pre-calibrated mixtures referred to, unsurprisingly, as "calibration gas" readily available. Googling for "carbon dioxide calibration gas" will find you what you need as well as the appropriate pressure regulators. Note that even the professional meters require calibration on a regular basis; for example the calibration schedule for the Fluke 975 for CO2 is yearly and for carbon monoxide is monthly.

throwaway9d0291 6y ago

> Every IAQ CO2 meter > For example, suppose I buy two CO2 meters from Amazon It sounds like you're looking at retail consumer products, which I think is the problem. The Sensirion SCD30 _sensor_ [0] for example allows forced recalibration [1] to an arbitrary concentration. Also, since you call it an "IAQ" CO2 meter, it might be that these aren't CO2 meters at all. They could be VOC meters that make up a projected CO2 reading based on the presence of other chemicals in the air. You need to make sure the sensor is NDIR-based and actually measures CO2. If you want accuracy, don't buy retail sensors that don't come with spec sheets. I'd find the sensor you want first then find or build a meter based on it. For example the SCD30 can be connected to an ESP32 and connected to WiFi using ESPHome [2], which can be set up with some YAML files. > self-calibration Just by the way, self-calibration is what you want. Real CO2 sensors drift over time and you definitely want them recalibrated at least once a year. Auto-calibration takes away that pain and when implemented in a suitable way, really works. Many calibration routines are not suitable for residential environments as they're based on the assumption that they will see pure fresh air at least once a day. This is a safe assumption in a business environment because most AC systems have an overnight fresh air purge. In a residential environment though, someone might not open the window once a day. Again referencing Sensirion, their auto-calibration cycle looks for two, hour-long valleys in the CO2 readings over the previous two weeks that are within a few percent of each other and uses those as a fresh air reference. In my experience, this works well at home. > Another downside to this difficulty in testing is that it's hard to truly evaluate a particular meter's performance. Which seriously limits side-by-side comparisons of different models, and makes it harder to hold manufacturers / retailers accountable for poor performance. You "just" need a reference to compare to. If you have a look at academic literature, they typically use a lab-grade setup to compare. You're right though, it's difficult for a regular home user to evaluate a sensor like this. [0]: https://www.sensirion.com/en/environmental-sensors/carbon-dioxide-sensors/carbon-dioxide-sensors-co2/ https://www.sensirion.com/en/environmental-sensors/carbon-di... [1]: https://www.sensirion.com/fileadmin/user_upload/customers/sensirion/Dokumente/9.5_CO2/Sensirion_CO2_Sensors_SCD30_Interface_Description.pdf https://www.sensirion.com/fileadmin/user_upload/customers/se... [2]: https://esphome.io/ https://esphome.io/