Back in the mid-'80s, I owned a boxish Fischer stereo that had a bad turntable, two tape decks, and a radio tuner. The two speakers were separate from the unit.
I discovered that if I left the speakers disconnected but really cranked up the volume, a faint, tinny sound would eminate from the stereo itself, sounding somewhat like listening closely to an unamplified needle on a record, except louder with more distortion (and, no, I wan’t running the turntable – it was the radio and tape players). Though distorted, you could recognize what song was playing.
Any ideas on what part of the cheap electronics could have been turning the voltage meant for speakers into actual sound waves without assistance of voice coils?
Hmm. Not quite the same, but I remember an incident when I had a small bookshelf stereo - tape, CD, radio all in one unit. I was in a location very close to a large number of radio broadcast antennae and I remember that even when the unit was switched to the CD or tape settings, the radio signals were so strong that it was still receiving and playing the radio input.
Can’t help with your question, just thought I’d share.
Transformers can sometimes hum quite a bit. If there are audio transformers at the last stage of the amp, they might vary their hum based on the audio signal. The transformer doesn’t even have to be in the audio signal path (but probably not in your case). The variance of the draw of power can also vary the hum of the power supply transformer.
Set the amp to an unconnected aux input and listen for hum. (Altho you sometimes get signal bleed from other sources, e.g., the radio circuit.)
I was gonna say what ftg said. I don’t know the exact details behind their operation in stereos, but I do know from building simple tone circuits that the audio transformer can reproduce a very quiet, tinny version of the input signal (it usually sounds a lot higher pitch though).
That’s because it is higher pitched. By one octave, in fact. The reason is that the effect which causes this in transformers, called magnetostriction, operates to squeeze the core whether the signal is positive or negative, and therefore the core gets squeezed twice per cycle.
I was going to suggest transistors, but I wasn’t sure if it was accurate or not. I also figured they could produce sound, but wasn’t sure if this was factual or not.