For the device – HDC3020, or HDC3020-Q1 for the automotive version – the company is claiming “<0.21% relative humidity accuracy drift per year and less than 5% relative humidity drift from temperature and humidity stress, tested up to 85% relative humidity and 85°C”.
Again emphasising accuracy, the company said that the parts are “verified with a procedure traceable to the National Institute of Standards and Technology across the supply voltage of 1.62 to 5.5V and the widest temperature and humidity range at ±1.5% relative humidity”.
Taking a look at the data sheet, which is new (18 Jun 2021) and covered in ‘TBD’ entries, accuracy is ±1.5% typical and ±2% max over 10 – 90% RH over 1.62V to 5.5V, but only at 25°C. There there are no accuracy figures given across -40 to +125°C.
There are multiple low-power operating modes, although the data sheet has a few blanks in their parameters. Active power consumption is typically 125μA or 100μA in modes 0 and 3 respectively, then sleep typically drops this to 400 to 550nA depending on mode.
As the device cycles though sampling and sleeping, averaged figures are typically between 700nA and 2μA.
A 30mA heater is available to remove condensation.
A couple of other devices are mentioned. While HDC3020 has no protection over the air inlet to the sensor element:
HDC3021 comes with factory-installed polyimide tape covering the opening against PCB soldering, PCB wash and conformal coating.
HDC3022 has a permanent IP67-rated hydrophobic PTFE filter cover over the opening to the humidity sensor “designed to adhere to the package over lifetime operation while
maintaining the same response time as a sensor without the membrane”, said TI. “The cover has a filtration efficiency of 99.99% down to a particle size of 100nm.”
Pre-production versions of the HDC3020 and HDC3020-Q1 are available
Applications are foreseen in cold-chain data loggers, wireless environmental sensors and IP network cameras.
The HDC3020 data sheet is here, and the HDC3020-Q1 doesn’t have one yet.