> Can you make them small and cheap enough to put in a phone, so I'd get a notification if the CO2 levels are rising?
Yes! It's possible. But the air circulation inside a smartphone is limited and newer phones are completely sealed for waterproofing, so there are still some doubts.
> How are CO2 sensors built?
Based on my very limited experience on playing with electronics, I would say...
1. Previously, many CO2 sensors use an electrochemical process, the chemical cell in the sensor reacts with air, and outputs an analog voltage proportional to the gas concentration. They are large, have a distinct cylindrical shape similar to a microphone module, needs manual calibration, and common in the industry. For example, have a look at the datasheet of MG811 sensor, which mentioned how the chemistry works: https://sandboxelectronics.com/files/SEN-000007/MG811.pdf
2. NDIR sensors is another traditional option. Inside a large rectangular blackbox, it has an infrared LED and some electronics. Basically, you shoot a beam of infrared light though the air and detect it on the other end, and correlate your measurement with the gas concentration. NDIR is reasonably accuracy, very popular in the industry, and can detect particulate air pollutants, and many types of gas. As an example, take a look at the MH-Z16 datasheet: https://sandboxelectronics.com/wp-content/uploads/2018/08/Z1...
3. It's now possible to build an all-in-one digital gas sensor on a piece of semiconductor chip! Examples include:
Curious where you are getting those figures from. Looking at the usual suspects (newark, digikey, sparkfun etc...) gives figures like $5-20 a piece (even at bulk).
Ive run into this before with some hobbyist electronics (high fov IR LEDs) where the prices were much higher than one would expect.
> Curious where you are getting those figures from. Looking at the usual suspects (newark, digikey, sparkfun etc...) gives figures like $5-20 a piece (even at bulk).
edit: I meant to say the cost is 5.76 USD each per 1000 pieces, I should've been clear. In the industry, it's common to quote the price as "1k price", which is the 1 pcs price per a bunch of 1000 pcs.
> I've run into this before with some hobbyist electronics (high fov IR LEDs) where the prices were much higher than one would expect.
Hobbyist electronics are more expensive than the raw price, sometimes an order of magnitude, mainly because they're not mass-produced. If the part is expensive, this effect is there. But usually the cheaper the part, the stronger the effect.
In electronics mass-production, when you already have a PCB and you add a new chip, the cost is often close to the chip itself.
On the other hand, a breakout board or a devboard usually only has few chips on the board, sometimes nothing more than the soldered chip, with a voltage regulator and some pull-up resistors for "plug-and-play" experiments. Also, they are produced in small quantity so the cost cannot be low, and finally vendors like Sparkfun usually sell them to the hobbyists at a premium. For hobbyists, the point is to have something ready to work with, plus it's increasing impractical to solder some chips by yourself due to the increasingly small footprints, and anyway, even if you are gonna run a production, you need to see how it works first. So these hobbyists electronics are more expensive compared to the cost in the industry.
Interesting links, thanks! So it's mostly an IR lamp, an IR sensor with an optical filter and the air to test in between. I don't understand from the article how long the waveguide actually is. Is this just a cylinder that is a few mm long bent around a corner, or is this some hollow tube much longer wound around the center of the sensor many times?
