I'm doing an online course in electronics and I have been set the following question:

Calculate the maximum value of Rn if the resolution of the ADC is not to
be adversely affected by thermal noise. Assume the ADC operates at
25°C.

Now, I know how to calculate the thermal noise. I know how to calculate the resolution of each bit of the ADC. What I don't know is what constitutes not "adversely affected". I can't find it any where in my materials or any of the recommended sources. I assume that the thermal noise at most should be equal to or less than the quantization error but maybe it should be less? Is there a standard value in terms of the resolution?

 

Hi Fnord,
This PDF may help define, Ref Pages #14,, #70+
E

 

Doesn't it depend on the sampling rate? V=√(4kTRf) so the noise is proportional to the square root of the bandwidth, which would be set by the anti-alisaing filter.

 

Sorry for the delay.

Hi Fnord,
This PDF may help define, Ref Pages #14,, #70+
E

I'll take a look. Thanks!

Doesn't it depend on the sampling rate? V=√(4kTRf) so the noise is proportional to the square root of the bandwidth, which would be set by the anti-alisaing filter.

Yes, but that isn't the issue. I've been given the temperature and the frequency and I need to calculate the resistance. However the only information regarding the voltage is that it shouldn't "adversely affect" the resolution of the ADC. I assume it would be proportional to the voltage of the LSB but I've seen both ≤LSB and ≤½LSB.

 

Sorry for the delay.


I'll take a look. Thanks!


Yes, but that isn't the issue. I've been given the temperature and the frequency and I need to calculate the resistance. However the only information regarding the voltage is that it shouldn't "adversely affect" the resolution of the ADC. I assume it would be proportional to the voltage of the LSB but I've seen both ≤LSB and ≤½LSB.

If it's not the issue, what value for F are you going to plug into the Johnson noise equation?

 

If it's not the issue, what value for F are you going to plug into the Johnson noise equation?

It is 1Mhz.

 

It is 1Mhz.

So you are sampling at >2MHz?