Why does RFI demodulate to a buzz sound?

Question

For RFI problems with audio equipment, why does a dead carrier often demodulate to the familiar mains 60 Hz buzz sound? This occurs on multiple audio equipment around the house. There are lots of references to RFI buzz sounds, and I understand how the RFI gets in and gets demodulated, but I found no explanations as to WHY it often demodulates to a buzz. For example, I key up an HT on 2m next to an old PC/multimedia amp/speakers, and I get a loud powerline buzz sound. The HT is not set to a CTCSS tone, just dead carrier. Seems to me that this RFI should demodulate to just a constant DC level.

Answer 1

This is a more than a bit of speculation, but it's the best I can do without having a set of screwdrivers, disregard for the integrity of your personal possessions and an oscilloscope in your home while being left alone:

Amplifiers have what is usually called a high Power Supply Rejection Ratio (PSRR), meaning that you don't hear the fluctuations of the supply voltage on their output; or to be specific, their output fluctuates by less than the supply voltage by a factor of PSRR.

Now, if you ever built an amplifier from transistors yourself, you will wonder how that works: No matter which basic configuration of transistor you use, the output voltage depends on the supply voltage. Usually it's even worse than proportional!

The solution to that is that you have a feedback between the output of your amplifying transistor and its input, and you build that feedback such that the effective amplification is less than is theoretically possible at any given supply voltage. That way, you not only get a reliable amplification factor, it also becomes independent from the input voltage regulation. (Within limits, but usually sufficiently: Even very cheap audio amplifier chips offer PSRR of > 60 dB, meaning the input increases by less than a millionth if you doubled the supply voltage. Of course, at that point you probably fry the chip, but that's another story…)

So, in short: Feedback between the output is applied such that if the output is higher than the desired multiple of the input, gain is reduced, and vice versa, if output is lower than the desired multiple, the gain is increased. (This is a very fundamental concept in amplifier design.)

This way, you can just use the cheapest transformer supply you can buy that still supplies enough current for what you need, slap a cheap diode bridge rectifier on that, and add a capacitor of the lowest value necessary to not let the voltage drop under high music load below the minimum supply voltage the chip tolerates. You get high supply voltage variation, but together with the high PSRR it's fine! You actually built a good, but cost-efficient multimedia speaker amplifier

Now, if you dig a bit deeper into amplifier design, you'll realize that this feedback needs to be a bit different, in fact, quite different, for different frequencies in your signal spectrum (in this case, audio). For example, you want very strong negative feedback at frequencies at the higher edge of what is audible, so that your speaker doesn't "screech" at you when there's noise at the input.

What happens now when RFI picked up by the output side ends up in the feedback chain? Well, it gets subtracted from the input; if the amplifier wasn't designed to have sufficiently limited bandwidth to ignore that altogether, you might well saturate the whole feedback mechanism, which might act as an involuntary homodyne receiver. But saturating the feedback mechanism means that feedback, and thus, power supply rejection, stops working effectively: You effectively built a receiver that modulates the homodyne reception with the 60 Hz (or 120 Hz if symmetrically rectified) from the power supply. A fuzzy buzz!