Hi,
I ran across the RFI rectification issue recently and have read several old papers/notes from TI and ADI. I know little about how the details of the internal implementation of amplifier, but from what I can find, it seems all the explanations are from one BJT transistor or JFET or MOSFET. Though it is understandable how a dc offset is generated at one transistor level, it is not clear how this dc bias propagates to the output of a amplifier. Some related questions ar haunting in my mind. So I list them here, based on my very limited understanding from the feedback control of an amplifier. Forgive my ignorance, could anyone give me some hints for this RFI issue?
1. It is said RFI rectification only happens with high out-of-band frequency input signal. But from the notes explaining this effect, it is mainly due to the nonlinearity of transistor's transfer characteristic curve. So, at lower frequency within the bandwidth of a amplifier, should a dc bias be also generated? If it is not seen at the output of an amplifier, is it because of the suppression from the closed loop control?
2. Then, if so, how is a dc bias generated at the input of an amplifier rejected by the feedback control? It seems that if the output has a dc bias, no matter whether it is rectified by high frequency ac or lower frequency ac, the control loop always has the ability to regulate to the shape of input waveform (before the rectification, if " Vref-Vo" takes effects here)? But apparently, this is not the reality. Or, is it because that the dc bias coexisting with HF output signal cannot be sensed back to the input, whereas dc bias coexisting with the LF signal can be detected?
3. Or, is there an internal special mechanism leading to the difference between HF and LF signals?
4. For a single transistor, this rectified dc bias is positive. Is it possible that the output of an amplifier see a negative dc bias, e.g. in the case of an inverting configuration?
Thanks,
Bo