NTC Thermistor noise disturbance

[funnyKedougou]

Hello, I am new in this community and I am glad to a membre.
Actually I am bloked since almost 1 mouth on a problem. Look at the circuit below

the NTC thermistor returns wrong values due to disturbance coming from nearby MOSFET. see below

I tried to create a filter by changing the C105 value to 200n or to add a 10k R to create a PI RC filter. But noise reduced but stay high.

Can please give some new ideas ?

Thanks

 

[crutschow]

I tried to create a filter by changing the C105 value to 200n or to add a 10k R to create a PI RC filter. But noise reduced but stay high.

Is the filter close to the MOSFET input, where it should be?
Have you tried increasing the value of the capacitor?

 

[funnyKedougou]

Yes the CTN and RC very close to the Mosfet, the reason is to have better measure of the Mosfet heating.
increasing the C105 value was my first idea, but this reduce the noise just a little.

 

[crutschow]

increasing the C105 value was my first idea, but this reduce the noise just a little.

What value did you use?
So what's limiting using an even larger capacitor?

 

[rjenkinsgb]

Your circuit shows it having a ground remote to the MCU. If it wired like that, it could well be ground noise.

Try power and ground connections directly at the MCU, and no connections to anything else away from the MCU.

 

[funnyKedougou]

Hello thanks for this idea, I will try it and let's see the progress. I let u know the result

 

[funnyKedougou]

I have used 3 values: 10n, 222n (100+100+22n) and 400n (200+200n). In the final try, I added a 10k R and I have now a better result but not good yet :-I

 

[crutschow]

I have used 3 values: 10n, 222n (100+100+22n) and 400n (200+200n). In the final try, I added a 10k R and I have now a better result but not good yet :-I

You can go a lot higher than that since the thermistor resistance changes slowly, so even a large RC value won't affect the measurement accuracy.