The 1/f (or flicker) noise of an LNA can strongly limit the sensitivity of a low-frequency detection chain. JFETs are well known as the best transistors, in terms of low-frequency noise. However, "modern" integrated circuits (ASIC) are more and more limited to the only MOS transistors, dramatically affected by 1/f noise (corner frequency of 10-100kHz usual).
In such situations, 2 techniques are used to counteract the impact of readout 1/f noise :
Modulation: the analog signal is shifted in frequency (modulated) higher than the 1/f corner frequency. Then demodulated after amplification.
Reference sampling/Chopper: the readout system periodically (faster than the 1/f noise) readout a fixed reference to subtract long-term noise contaminations. Also called: correlated double sampling - CDS
Have you a clear understanding of any fundamental differences/similarities in the use of these 2 techniques? a comparison?