I'm playing around with an oscilloscope and my laptop's headphone jack. I get the tip/ring/ground and measure between them. I also measure between them and an external reference ground. I notice I can see when the battery is charging. That is, if the battery is not in the laptop but the AC is in, then I insert the battery, it runs through a few seconds of a test, and then if the battery is not full, it goes to a certain waveform – if it is full, it goes to another "quieter" waveform. This is best done when the laptop is turned off otherwise the waveform is quite noisy.
I'd like to be able to understand more about what is going on. Might there be any other data present in the signal on the headphone jack if I am not putting out an audio waveform? ie. by virtue of the circuit passing through some of the data processing elements of the motherboard, when the computer is powered on? There is a lot of noise in the signal but maybe it is possible to clean it up some.
I thought this was interesting as it is a similar idea:
http://cs.tau.ac.il/~tromer/acoustic/ (Here the noise comes from a bank of 1500µF capacitors).
Best Answer
If you measure between the signal lines and some external ground, any signal which appears that does not show when you use the device's own ground is ground loop noise: it is the signal representing a fluctuating potential difference between the external ground and internal ground.
The currents which charge the laptop are generating some voltage through the ground connections, which means that these different grounds are not at the same potential, even if they are in fact supposed to be the same ground, electrically.
This is not a valid signal. Audio people pull their hair out eliminating unwanted ground loop noises which sometimes appear where they should not.
You're going out of your way to pick up stray signals by tracing the potential difference between some external ground and an output terminal on the device. Your measurement spans the entire device and its power supply.
The headphone jack's signals consist of the voltage between the tip or ring, and the sleeve. That ground, not any other. If you use the wrong ground, you are not measuring the signal properly across the points between which its voltage is supposed to appear. It is not some "side channel" data, but rather garbage.
This is like measuring someone's height in shoes, and noticing: hey, there is some "side channel" of extra height data that varies with whether shoes are worn and what type!