Hi,
I am using the TMP468 with a noise filter Cf = 1000p and Rf = 806. In my systems quiet state everythign measures nicely. Then I turn on a noise source with 40 kHz period with 100 nsec duration. Going into the filter I can see noise around 80mV peak and coming out of the filter it is knocked down to 10 mV peak. One would think this is good. What is the expected error for noise in the 10 mV peak range?
However, while the noise is occurring I get a 12C increase or more with Rf = 806 and when I short the resistor out making Rf = 0 the error increase reduces to 6C. This makes no sense to me. If feels like the series resistance compensation calculation is not working in the presence of noise. Also most of the noise spikes are downward and thought this should lower the temperature read no increase it.
According to the spec the Rf and Cf, I use are within recommended limits.
Please advise is this expected behavior?
Thank you,
-Jennifer
