Thread Sensor Node

This idea is about creating a fixed Thread sensors node, that acts as a router (Full Thread Device) and has a full set of high power consumption sensors as it will be connected to a permanent USB power source :

  • Thread Router or Gateway
  • Air quality
  • Light colors
  • Temp hum press
  • mic

image

We could use the dongle as a basis or use an nRF52840 module
I already have many of the mentioned sensors and was already planning to start working on a firmware with Zephyr

  • Light colors can also have proximity user motions APDS9960

  • The mic can signal events like sound of door banging, or simple command to switch light on or off directly from the uC

  • I think using the same sensor for Temp hum press would be practical MS8607

Maybe simply a “hat” for the nRF52840-usb-dongle with ready sensor modules can help with quick prototyping.

the resolution of the ADPS9960 is pretty bad for light reading. im testing these sensors TSL2561, VEML7700, and BH1750. I hav decided to use the BH1750 for my sensor nodes.

Can you specify what “bad” means in terms of accuracy, resolution, number of bits and integration time ? Did you really use it properly with its maximum spec limit or could it be a limitation of the library you used ?
What are your selection criteria ?

I wanted additional info about ambient light color and not only a single brightness value.

ADPS9960 seems to be a good choice compared to the others uni mentioned.
TSL2561 - unavailable
VEML7700 - is the 90° version of the VEML6030
BH1750 - is not recommended for new designs

1 Like

A very interesting candidate for CO² is this one: https://www.mouser.de/datasheet/2/196/Infineon_EVAL_PASCO2_SENSOR_DataSheet_v01_00_EN-2399950.pdf

PAS CO2 from Infineon - it uses a new technique to sens the CO² based on lasers inside the enclosure which makes the sensor much smaller then usual CO2 sensors. But there might be cheaper ones, just wanted to mention that one.

I don’t mind an expensive sensor, for CO² I would go with the following selection criteria

  • How accurate are the measures (at least realistic)
  • Does the sensor needs to be in an aerated spot or be part of a machine that has air flow (maybe we have to add a small fan)
  • Does the accuracy fail over time after 1 or 2 years ? How long can it be used
  • If it is really > 10 € or even > 20 € then it would go on a design on its own as air quality station and not part of the minimal Sensor node

at lower light levels, the ADPS9960 reads zeros while for the BH1750, there are still quite a bit of range left before it reads zeros. so as an ambient light sensor, the BH can reads at lower light. However, it certainly doesnt do RGB readings.

for CO2 sensor, the SCD30 seems like a good quality one. it’s quite expensive but consider that it also has humidity and temperature sensors built-in it maybe a good sensor to use. I may get one later to check if the humidity sensor on this one is any good. if not, the CO2 sensor has to be exceptionally good for me to be willing to pay for it at that price but it’s hard to test a CO2 sensor.

I’ll test it and compare with the MH-Z14 and MH-Z19. for CO2 sensors especially, i think it needs to have an onboard air chamber.

I get what you mean, that probably has to do with integration time, as it can increase the resolution, I’ll check that out.
For CO2, I’m not sure if they can be reliable enough to be used as an alert, otherwise just for the fun of watching a curve :wink:

co2 hav great effect on the body over all health, especially with the quality of sleep. most American houses are very well insulated, as the result the air cycles are pretty much enclosed, so the co2 get congested inside the house over time. i always open the windows when possible to let the fresh air in. Most ppl would never open their windows. so they hav to breath the bad air day in day out. especially in the winter bc they stay in the house most of the time.

I think it’s ok to calibrate CO2 sensor just by using the outside air. and use it as reference too see how bad the indoor air is. another thing i think i can test is to put some liquor at certain distance away from the test sensors and see which one still able to pickup some deviation to judge their sensitivity.

That’s true, some people worry about air quality and often open the windows, others don’t. Guess who’s going to need the sensors vs who’s going to get them.
Office meeting rooms are also important although the behavior is different during the pandemic.
Calibration with the outside air is a good idea.
I was even thinking about motorized window openers, at least from the top, to automate the reaction to aerate the room.