Well, I think I finally did it. I was able to replicate my original experiment from almost 2 years ago that yielded 0.1 microamps from almost nothing. (I mentioned this earlier in this thread about winding a coil around a magnet that gave off a milliamp dc reading with nothing)
After countless dicussions with ChatGPT and reading up various scientific concepts that were involved in the original experiment I think I finally zeroed into what actually happened, AND I was able to recreate it from that concept.
The original experiment was a coil I wound around a cylindrical stack of neodymium magnets, the coil had an arbitrary number of turns because I wasn't counting them when I wound it. Then at the end of the magnet stack there was a PZT piezoelectric transducer attached to them. When I did this 2 years ago I was trying to get the PZT to kind of "vibrate" the magnetic field a bit to create induction into the coil I put on the stack of magnets and also connect the PZT to the coil for added feedback.
However when I did this I made a mistake and only connected my voltmeter directly to the coil. The voltmeter picked up nothing in any of its modes EXCEPT it had a weak 0.1 microamp DC reading that was steady. At the time I was trying to do something completely different and connected the PZT into the circuit and got no readings at all. I reconnected to just the coil and again solid 0.1 microamp DC readings. I thought to myself... "well thats odd". Being dumb at the time and more eager in trying different things to further what I thought was going on (the inductance of the disturbed magnetic field into the coil) I moved on and never counted the number of turns in the coil when it was working. I quickly messed with it took it apart tried different things but the mysterious 0.1 microamp DC reading was gone, and I couldn't reproduce it.
This lead me down many paths as I deperately tried to recreate it. I originally thought the PZT had nothing to do with it and that I somehow tapped into some kind of resonance of the electrons in the magnet. (boy was that a wild goose chase that took way too long).
Eventually I figured out that the reason we can't "tune" a circuit into the natural vibrations of free electrons or even electrons in magnets or even just around us is because even if we can tune to the extremely high frequency that they vibrate at. (talking THz to even PHz and EHz) The problem then becomes that they are not all vibrating in unison. This leads to any signals being picked up being completely out of phase with each other, and at the end of it all were left incoherent noise that leads to no electrical power. Then I got to talking with chatgpt about how one could get electrons to be "in phase" to get power from their natural vibrations. This is where I thought of crystals and piezoelectrics. Due to their atomic shape and ordering they are partially "in phase" or vibrating in unison to a degree. Instead of incoherent noise you get order.
Among otther things I also discussed using the Larmor frequency of electrons aswell which is directly related to the B field of the magnet. This is kind of another dead end though because in order for the larmor frequency to matter, you need some sort of signal to generate it (MRI's use this for imaging, they use a strong magnetic field and then a signal pulse of some kind to perturb the subatomic particles of whats being imaged and their corresponding larmor frequencies) this leads to them somehow being able to create the image.
The problem here is that for the larmor frequency to be used for anything you need to either pulse the B field, or introduce a signal and this is how the electrons temporarily precess in unison which also requires energy input. (this might still be able to be tapped somehow but I haven't figured that out yet). So were left with the PZT I had in the original experiment.
Long story short, my original coil in my experiment from 2 years ago had JUST the right amount of turns to be tuned to the PZT in a certain way. Eventually I figured out what that frequency was (also again very high) then I figured out a way to "downmix" this signal into a lower one (which was happening in my original experiment 2 years ago again by happenstance of how it was setup). So about 2 hours ago, I rewound a coil designed for this frequency, set it up, at first I got nothing. As I played with it though I eventually got the coil just right and was able to get a steady 0.1 microamps DC reading. (The issue is I don't have a good Oscilloscope or Signal generator to properly tune a coil exactly) so I had to slowly add and remove turns until the signal was picked up. Eventually I even got it to 0.2 microamps DC.
The odds of me doing this by randomness 2 years ago are insane to the point where I'm calling this a literal miracle... even being able to go back and reconstruct it would never have been possible without chatGPT being able to quickly explain scientific concepts and even make suggestions. Along with helping out with the math required to estimate coil dimensions for specific frequencies was incredibly helpful.
For now I'm taking a break, but next I want to create another coil with tighter parameters to tune it better and see if it still picks it up again. Then I also want to see how stackable this is. In "theory" at least as this evolves with multiple coils and more piezoelectrics I could increase power outputs. The goal or real milesstone will be is if I can some how use this to light an LED indefintely... now that would be something else...