Oh, come on! the TIS1040 puts out 3kV at 160kHz, probably a nasty waveform at that.
You and tooki may get off on this constant argument but it's getting kind of tedious for the rest of us. That 'demonstration' has put you down in my estimation.
P.S. It only takes a couple of pF inter-strip capacitance to pass enough current (the TIS1040 is limited to 7uA) to perform that little trick without needing any insulation breakdown.
You omit to mention he other very important characteristic: the source impedance. That is what will determine the voltage seen at the device terminal. The manual, such as it is, is unclear about the source impedance. Certainly touching the exposed termjnal with ringers produces no sensation whatsoever.
Of course the leaf capacitance is of to be of the order of a few pF. That's obvious from simple school physics. But that is sufficient to cockup many analogue circuits, especially those with high internal gain reduced by heavy feedback.
I've no idea why you've invoked insulation breakdown; that never been suggested as relevant.
Please don't make it worse...
Source impedance of the TIS1040 is irrelevant in this situation, it's designed to be effectively a current source (for safety reasons). The voltage would have been the breakdown of the Neon (or LED) plus the drop across the reactance of the inter-strip capacitance at that the operating frequency.
What is the reactance of a 1pF capacitor at (go on, I'll be generous) 150kHz pure sine wave? Take your time. Now how much reactance, and so, how big a fraction of a pF do you need to pass the entire current limit (7uA at 0-5Meg) of the TIS1040 output at 3kV?
I mentioned breakdown, subsequently, because Omega Glory expressed surprise at the 3kV, it's likely that he was thinking about insulation breakdown.
Posting you little neon demonstration without indicating the magnitude of the source was unforgivable. Credit to him for asking the question.
Just for fun, yesterday I was inspired to spend 5 minnutes demonstrating the fun effects solderless breadboards can have. Getting the photo took another 5 mins
This is an unmodified solderless breadboard, a power source, and a neon bulb. As you can see, the power source is connected the left side side of the neon, but not to the right side. The neon bulb is getting enough power to light up, despite being unconnected to the power source.
Such power transfer is more than sufficient to screw up an op-amp feedback loop, etc, etc, etc.
You may claim it was "Just for fun" but you posted it to make some sort of dramatic demonstration of how bad breadboards were, otherwise why bother, when a simple fractional-pF series capacitor would have achieved exactly the same result.
Why did you use emotive language like "power transfer" and "the neon bulb getting enough power to light up, despite being unconnected to the power source." when what you actually meant was, there is a tiny bit of capacitance, as there is in anything. You intended to deceive.