true RMS power detection resolution

This subject has been partially addressed in another thread:

https://ez.analog.com/circuits_from_the_lab/f/q-a/120301/cn-0399-circuit-note-improvement-related

Regarding which RMS power detector is best:

Consider one of the linear in Volts-per-Volt detectors, for example ADL5501 or ADL5502, because they offer potentially the highest output slope in Volts per volt, provided the input is also high level. For example, ADL5502 can output the equivalent of around 0.15 Volts-per-dB slope up in the 0 to +5dBm region. This means 0.001dB resolution would require 150uV change at detector output. You would need to measure detector RF output noise and stability, and see how it compares to 150uV. 

This subject has been partially addressed in another thread:

https://ez.analog.com/circuits_from_the_lab/f/q-a/120301/cn-0399-circuit-note-improvement-related

Regarding which RMS power detector is best:

Consider one of the linear in Volts-per-Volt detectors, for example ADL5501 or ADL5502, because they offer potentially the highest output slope in Volts per volt, provided the input is also high level. For example, ADL5502 can output the equivalent of around 0.15 Volts-per-dB slope up in the 0 to +5dBm region. This means 0.001dB resolution would require 150uV change at detector output. You would need to measure detector RF output noise and stability, and see how it compares to 150uV. 

hello sir, 

Thank you for the response. 

So, it is difficult to measure 0.001 dB with RMS detector because of the noise and other problems.

Is there any other method to detect a change in power of 0.001 dBm in the signal of 5 GHz.?

Hi amitholey,

I think you will find that simply loosening then re-tightening an SMA connector at 5GHz will introduce much more than 0.001dB uncertainty. 

Could there be some mis-understanding as to what is the real requirement? A requirement such as +/- 0.02 to 0.05 dB or higher would be much more typical. 

Thank you, sir.