0
\$\begingroup\$

I always wondered why ADC is not considered for noise figure measurement in receiver chain? Do ADCs will not have much impact on noise figure in signal processing?

How to derive a Noise figure spec for receiver in relation with the selected ADC?

\$\endgroup\$
  • \$\begingroup\$ Convert Noise dBm min to dBuV min after gain to mV then choose resolution less than Noise level in mV \$\endgroup\$ – Sunnyskyguy EE75 Aug 2 '17 at 17:52
  • \$\begingroup\$ If you digitize at the Antenna, then the ADC noise ---- thermal and quantization and VDD-trash and digital-interface-trash ---- do matter. With ADCs prized for their ability to make decisions, the quality of the decider --- the comparator(s) --- is key. Unfortunately, microvolt signals into a comparator are inadequate overdrive to achieve fast decisions, unless the comparator has lots of broadband amplification internally. And that broadband amplification is serious impact on the noise floor of the ADC's comparator(s), thus RF signal chains perform lots of amplification in NARROW BANDWIDTHS. \$\endgroup\$ – analogsystemsrf Aug 3 '17 at 3:54
1
\$\begingroup\$

Through a receiver signal chain, dynamic range is a more important measure than noise figure. The dynamic range is the difference between some overload measure, like clipping or a distortion figure, and the noise figure. If there is sufficient dynamic range, then the signal level can be chosen to sit well within that, above the noise level of the component. If the dynamic range is not sufficient, then a good noise figure alone cannot help.

This doesn't hold at the very first component in the receiver after the antenna. Here the signal level cannot be chosen, it is what ever is received. This is why the Low Noise Amplifier (LNA) is specified by noise figure as well as by dynamic range.

It turns out that in practice, the dynamic range of a good ADC can easily exceed that of amplifiers. You might spend more (money, or watts, or PCB real estate) on the amplifier that boosts the signal level from your IF strip's -20dBm to the one volt or so needed by the ADC, to match its dynamic range.

\$\endgroup\$
1
\$\begingroup\$

why ADC is not considered for noise figure measurement in receiver chain?

That is because in a properly designed receiver the ADC should not be the main noise contributor. This means that the ADC should not determine the noise of the whole receiver. If the ADC would determine the NF then the NF can easily be improved by using a better ADC (more bits, lower quantization noise).

Does the ADC have much impact on the noise figure ?

No, not if the system is designed properly and the ADC is good enough not to determine the NF of the complete system.

In most systems the NF is determined by the part with the smallest signals and that is at the (antenna) input so the input of the Low Noise Amplifier (LNA). The task of the LNA is to amplify the received signal, while adding as little noise as possible. Once amplified the signal should be larger than the noise level of the other amplifiers, filters, mixers and the ADC so the absolute noise level requirements on those blocks are less stringent than for the LNA.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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