1
\$\begingroup\$

This is a typical extract from a spectrum analyzer data sheet.I want to know the reason behind the increasing uncertainty of level measurement with frequency response specification.enter image description here

\$\endgroup\$

1 Answer 1

1
\$\begingroup\$

The effect of stray reactances, usually shunt C and series L, gets worse as the frequency rises, it becomes an increasingly large fraction of the total impedance. This means that everything has less gain, less stability, more uncertainty, as frequency rises.

All instruments these days are processor based, and during manufacture are compared against known standards in a process called calibration, so you might expect that even if the RF hardware in the instrument itself had a frequency dependent slope on it, it would be possible to calibrate out.

'The Standard' 1mW of power is realised at low frequency.

Although calorimetry is independent of frequency, in principle you can heat a resistor with DC or with RF to make a substitution measurement of power, in practice the uncertainty of the input match of your resistor will create an uncertainty in power.

This means that the laboratory standards themselves are increasingly inaccurate as the frequency rises.

The uncertainty of match for the standards, cables, and the instruments being calibrated means the power reference is transferred to the instrument at less precision at high frequencies than low frequencies.

The uncertainty of match for the instrument means that the manufacturer has to allow for even more uncertainty when you use it at high frequencies.

This all assumes the hardware is perfect, and doesn't drift with time or temperature. At low frequencies, where the processes used to make amplifiers and the like have a lot of gain, ratios can be determined by feedback, based on stable resistors. At high frequencies, the increasing stray reactances reduce the available gain, and the instrument gain is more 'what you get with the process', which is inherently less stable.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.