Your web page information for the USB TC-08 quotes temperature accuracy; " Sum of ±0.2% of reading and ±0.5 °C "
1) Does this include the cold junction compensation error ? (is this the ±0.5°C?)
2) what is the definition here of the zero & span "of reading" ?
eg if the reading is 300°C on a k-type thermocouple, is the accuracy;
based on the zero of the k-type range, i.e. 0.2% * (-270 to 300°C) + 0.5°C = 1.64°C
or is it based on a 0°C reference junction 0.2% * (0 to 300°C) + 0.5°C = 1.1°C
or is it based on actual mV input ? (i.e. still zeroed on 0°C reference junction, but not linear with temperature.)
3) are the channels isolated from each other, or would I need to use isolated thermocouples to avoid ground loops ?
4) Update rate is listed as 100ms per channel. Is this a characteristic of the TC-08, or the bus/Picolog software? I need one reading with rapid update, some others which update time doesn't matter. So if I put the single reading into one TC-08, and the others into a second TC-08, both into a single PC/Picolog session, will I get 100ms sample rate on the first reading?