I just checked. My zero offset's run from -0.7 to +0.5 °C, based on an hour-long fully-settled ice-bath zero-point calibration, over a sample of about 20 sensors.
On Fri, Jul 15, 2022 at 6:07 PM Russell Senior <[email protected]> wrote: > > On Fri, Jul 15, 2022 at 5:17 PM Tomas Kuchta > <[email protected]> wrote: > > > > On Fri, Jul 15, 2022, 17:04 Russell Senior <[email protected]> > > wrote: > > > > > I have used a bunch of these: > > > > > > https://www.amazon.com/gp/product/B00T0K8NXC/ > > > and > > > https://www.amazon.com/gp/product/B01G7BE9WK/ > > > > > > with https://github.com/merbanan/rtl_433 and an rtlsdr (like this: > > > https://www.amazon.com/gp/product/B00VZ1AWQA/) to decode. > > > > > > For improved accuracy, I calibrated the sensors in an icebath (in > > > several layers of ziplock bags and desiccant) for 0 degrees offset. > > > . > > > > > > These are good sensors with great battery life. I also have ThermPro > > sensors, they work, but the ones Russell lists are better, smaller and much > > longer battery life. > > > > I do not calibrate them. I care about being comfortable and saving power > > rather than worry about whether the temperature reading is 0.2-0.3 degrees > > different from absolutely correct value. > > Fwiw, I didn't calibrate mine for a long time, but mine aren't so much > for automated control, they were for understanding the temperature > environments, and in particular differences and/or gradients in > various microclimates. I noticed that a particular outside location > seemed to get down to freezing before the other sensors. The > significance of small temperature differences increases the closer you > are to freezing, for example. Eventually, I just wanted to understand > whether the difference was due to the sensor or the microenvironment > it happened to be in. Unless vigorously stirred, there can be > significant temperature differences over very short distances, due to > heat sources, stratification, illumination, etc. > > I'd really like to have a lab grade temperature sensor, accurate to > 0.01°C, to actually calibrate against. I encountered sensors when I > worked in Oceanography with that kind of precision, but they were > designed for water temperature and also were several thousand dollars. > I wouldn't like to have one *that* much. Most consumer grade sensors > only claim ±1°C. > > > > > -T
