Hello everyone,

I have a datalogger that uses a NCU18WB473D6SRB - R1(47Kohm) in a voltage divider with a R19 - 62Kohm resistor a R4 - 4.7kohm resistor and
there is also a LM4040DIM3X-2.0 2.048V voltage reference IC. This circuit is part of a datalogger powered by a CR2032 battery with extended temperature range. The circuit is measuring with a good error until it gets to -10 and below where errors start to be around 1C to 4C at -20C and the measurements oscillate a lot at low temperatures. And this type of behaviour does not seem like just errors due to tolerances and temperature drifts.

Has anyone found problems like this and if so what can be a possible solution?




Thank you for your time!

 

Hi Cr,
Welcome to AAC.
The datasheet for the NTC shows it should work OK down to approx -40C.
What are the values and tolerances of R19 and R4, also C17,

I assume the TMP_VCC is the CR2032?

Is the V_TMP output working into a high impedance load?

E

 

Could you provide a link to the datasheet for this CR2032 battery that has an extended temperature range?

What type of resistors are you using? What are their tempco?

 

Hello, Thank you!

R4: 4.7kΩ 50V 62.5mW Thin Film Resistor ±0.1% ±25ppm/℃
R19: 62kΩ Thin Film Resistor ±0.1% ±25ppm/℃
C7: 100nF 16V X7R ±10%
C17: 22nF 10V X7R ±10%

Yes, the TMP_VCC is the CR2032 voltage.

And V_TMP is connected directly to the PA5 pin configured as Analog Input of my STM32L412KBU6 microcontroller, so it should have high impedance.

 

It’s due to the logarithmic response of the thermistor. Itis at its most accurate when the pull-up resistor is about the same value as the thermistor resistance.
if I look at my 8-bit thermistor lookup table, it is about 0.4°C per step around 25°C, and around 10°C per step at the extreme ends of the table.

 

i have implemented the Stein-Hart equation to calibrate with a temperature chamber the device and get the coefficients. I guess I should be able to correct this deviation at the end via software by correcting the coefficients right? However, the other doubt are the spikes that appear at lower temperatures...