[time-nuts] Temperature (environmental) sensors

Mark Sims holrum at hotmail.com
Tue Jan 3 13:03:05 EST 2017

I did a LOT of testing environmental sensors when I built my ultrasonic anemometer weather station that is part of a rocket launch control system.  The best humidity sensor I found was the DHT21/SHT11/AM2301/SHT15.  They run around $3 and can also provide temperature to 0.1C res / 0.5C accuracy.  I tested lots of MUCH more expensive humidity sensors and that one agreed the best to my lab grade Vasilia.  That said,  all humidity sensors are rather sucky.

For pressure the MS5611 sensor was, by far, the best.  It can measure the pressure change equivalent to less than a foot of altitude.   It can also provide a high-res temperature reading (of the sensor, not the environment).  It uses 24 bit A/D converters.  I use a 10 DOF IMU board from Ebay ($10) or you can get the sensor only on a board for around $7.

The weather station board uses a ATMEGA32 to read all the sensors and outputs an RS-232 stream that looks like NMEA data.  It uses four $1 SR04 ultrasonic  sensors to measure wind velocity/direction/air temperature.  It also supports a UV intensity sensor.  It uses the magnetometer/accelerometers on the IMU board as a digital compass so you don't have to orient the anemometer to north.  I've used the board without the ultrasonics as an environmental sensor.

I use sensor noise from the IMUs and Atmel ADCs to implement a true random number generator that feeds the crypto systems that protect the launch controllers from neer-do-wells.  The TRNG output passes all tests of randomness with flying colors.

More information about the time-nuts mailing list