Skip to main content
7 events
when toggle format what by license comment
Sep 6, 2023 at 23:13 comment added anrieff ... this of course assumes you're sampling the ADC rarely enough, and won't work if you're sampling it thousands of times per second. For a temperature sensor, it probably does not make sense to sample it more than a few times per second.
Sep 6, 2023 at 23:10 comment added anrieff ADCs cannot measure any voltage input. If that input is too weak - the ADC itself will pull it in some direction, distorting the reading. For those ADC internal operation quirks, read here if interested. 100k+100k is a weak signal. Placing a capacitor to ground will make it a bit beefier, as the 100nF capacitor is orders of magnitude larger than the 10-20pF capacitor in the ADC, so when the ADC samples the signal, the voltage in the 100nF capacitor will only decrease a tiny bit...
Sep 6, 2023 at 22:50 vote accept Hazardous Voltage
Sep 6, 2023 at 22:50
Sep 6, 2023 at 22:50 comment added Hazardous Voltage interesting, but why is using a capacitor would make the accuracy better?
Sep 6, 2023 at 10:56 comment added anrieff I'd suggest that you use a 10k NTC and 10k fixed resistor. This would get by the ADC requirements I believe. If you only have a 100k NTC and want to use it: then your divider is a 100kNTC+100k fixed resistor and the impedance is too much. Place a small capacitor - 10nF..100nF - from the ADC pin to GND. The capacitor will be parallel to your NTC or the fixed resistor (it doesn't really matter much whether the NTC will be on the "top" or on the "bottom" of the divider, the conversion code will be different, but the accuracy will be the same).
Sep 5, 2023 at 16:14 comment added Hazardous Voltage What will happen if I used a ntc 100k with a voltage divider? Would it affect the ADC?
Sep 5, 2023 at 13:02 history answered anrieff CC BY-SA 4.0