Scioly.org

MoMoney$$$;)0) wrote: ↑February 16th, 2020, 3:58 pm So I was wondering how exact you guys are getting your detector to be since mine is still around ± 0.5 Degrees and sometimes more accurate, but how exact have you guys gotten you detector to get to place. Thanks!

lifeof3.141592 wrote: ↑February 17th, 2020, 8:21 am

MoMoney$$$;)0) wrote: ↑February 16th, 2020, 3:58 pm So I was wondering how exact you guys are getting your detector to be since mine is still around ± 0.5 Degrees and sometimes more accurate, but how exact have you guys gotten you detector to get to place. Thanks!

in total or 0.5 degrees per test for a total of 2?

jinhusong wrote: ↑February 17th, 2020, 3:44 pm Depend on what sensor you use.

For TMP36, You should not use its linear formula, come up with your own high level polynomial (or LUT).
More important, its output noice extremely depends on the source. Better use standalone battery pack (like 2AA or 3AA). Still, read like 1000 time and take average. With all these, you may still not be able to get 0.1C resolution, but it should be way better than 0.5C.

For thermistor, normal setup of 10K resistor and 10K thermistor with 1000 average reading should easily get you 0.2C with Arduino UNO, it should get you 0.1C with Arduino DUE.

Student here cuts Steinhart formula into 4 sections with 5 datapoint, got 0.0, 0.0, 0.2, 0.1C.
3 dataPoint direct got like 0.1, 0.2, 0.1, 0.2C.

Hope these help.

MTV<=>Operator wrote: ↑February 17th, 2020, 5:56 pm

jinhusong wrote: ↑February 17th, 2020, 3:44 pm Depend on what sensor you use.

For TMP36, You should not use its linear formula, come up with your own high level polynomial (or LUT).
More important, its output noice extremely depends on the source. Better use standalone battery pack (like 2AA or 3AA). Still, read like 1000 time and take average. With all these, you may still not be able to get 0.1C resolution, but it should be way better than 0.5C.

For thermistor, normal setup of 10K resistor and 10K thermistor with 1000 average reading should easily get you 0.2C with Arduino UNO, it should get you 0.1C with Arduino DUE.

Student here cuts Steinhart formula into 4 sections with 5 datapoint, got 0.0, 0.0, 0.2, 0.1C.
3 dataPoint direct got like 0.1, 0.2, 0.1, 0.2C.

Hope these help.

When you say normal setup, what type of equation are you using? Also how exactly do you cut the formula into 4 sections?

MoMoney$$$;)0) wrote: ↑February 17th, 2020, 7:13 pm

MTV<=>Operator wrote: ↑February 17th, 2020, 5:56 pm

jinhusong wrote: ↑February 17th, 2020, 3:44 pm Depend on what sensor you use.

For TMP36, You should not use its linear formula, come up with your own high level polynomial (or LUT).
More important, its output noice extremely depends on the source. Better use standalone battery pack (like 2AA or 3AA). Still, read like 1000 time and take average. With all these, you may still not be able to get 0.1C resolution, but it should be way better than 0.5C.

For thermistor, normal setup of 10K resistor and 10K thermistor with 1000 average reading should easily get you 0.2C with Arduino UNO, it should get you 0.1C with Arduino DUE.

Student here cuts Steinhart formula into 4 sections with 5 datapoint, got 0.0, 0.0, 0.2, 0.1C.
3 dataPoint direct got like 0.1, 0.2, 0.1, 0.2C.

Hope these help.

When you say normal setup, what type of equation are you using? Also how exactly do you cut the formula into 4 sections?

It's quite easily actually once you know what you're doing, and if you're acquainted with the language you're working with. I'm pretty sure he's talking about using the Stienhart-Hart equation given with a thermistor, and then changing it with some testing.

MTV<=>Operator wrote: ↑February 18th, 2020, 5:45 pm

MoMoney$$$;)0) wrote: ↑February 17th, 2020, 7:13 pm

MTV<=>Operator wrote: ↑February 17th, 2020, 5:56 pm

When you say normal setup, what type of equation are you using? Also how exactly do you cut the formula into 4 sections?

It's quite easily actually once you know what you're doing, and if you're acquainted with the language you're working with. I'm pretty sure he's talking about using the Stienhart-Hart equation given with a thermistor, and then changing it with some testing.

Changing it as in adding a constant? I don't see how one could find new coefficients in the 20 minutes of allotted calibration time.