Metrology

It is incredibly cheap and easy these days to hook up a sensor to an arduino and start measuring something, but how do you know how accurate your results are? Do you just trust the sensor's data sheet and assume that you configured it correctly? That sentiment is more eloquently expressed in the first paragraph of this Cave Pearl blog post. The web pages here are about my effort to answer those questions.

From one perspective these pages could be considered performance tests of sensors in question. Whilst that is certainly one aspect, I am starting from a presumption that many commercial sensors are likely to outperform the test criteria I can easily construct. Where there is a discrepancy between a sensor and my home made reference source, we must consider the possibilty that it could be my calibrator that is wrong. I hope to reveal a lot about how well these sensors work, but I expect to learn at least as much about experimental design and the limits of my ability to set up reference calibrators. The experiments presented are not intended as a practical method for routine testing or calibration. Primarily I focus on whether it is even possible for me to achieve a true absolute calibration with my simple home-made equipment. Secondarily, once we have shown it should be possible, I test the absolute accuracy of various devices.

• Hygrometers including Bosch Sensortec BME280, Measurement Specialties HTU21D, Silicon Labs Si7021, Sensirion SHT71, Aosong DHT11, DHT22, AM2302, AM2320, AM2321
• Thermometers including Bosch Sensortec BME280 and Maxim Integrated DS18B20.
• Barometers. Using a Bosch Sensortec BMP085 far outside its specified operating range.

This project has grown way out of hand. As my friend Bob said, I either need to get a good editor to sort out all these web pages or else get a life and stop writing them. The links above go to comprehensive write-ups.

2017-04-02 7:44 PM