"Isolated" power from mains + LED + me = ???

Hi, while playing with a LDO I have found something strange and I have no idea how to explain it. I have an unregulated transformer DC power source, 12V nominal, 17V when open circuit. I believe it is isolated from mains because it didn't kill me so far. I used it as an input to a LDO. I have found that if I connect one leg of a LED to output of the LDO (or GND) and touch the other leg with my hand the LED glows lightly. I does not matter where is cathode and where anode, all LEDs I have tried worked this way. I was "floating" as well as possible - sitting on a wooden chair on wooden floor touching nothing else than the LED. I tried another power source (a switching one regulated to 9V) and it was the same. I tried 9V battery and got no glow. I tried to connect the LED directly to the power source and it glowed but I think it was less than when connected via the LDO. Do you know how it is possible the LED glows?

Because there is stray capacitance, you are picking up some AC voltage from the mains wiring around you. The power supply is either grounded or is picking up a different AC voltage from stray capacitance between its primary and seconday windings.

It doesn't take much stray capacitance to produce a few microamps from stray mains, but I'm surprized you get detectable light from the LED at that sort of current level. Perhaps you are grounded and the power supply has appreciable leakage.

Yeah, that's surprising.

The fact that it's lighting-up in either direction means it's AC.

I just tried lighting an LED "through my body" using my bench supply that goes a little higher than 20V. As expected, I can't see any glow with the few-megohms resistance of my fingers.

And under current conditions I'm getting less than one volt between my body and earth ground, using my body as an "antenna" (without the LED in the circuit). And connecting the LED in series between my body and earth ground, it doesn't light-up.

I was "floating" as well as possible - sitting on a wooden chair on wooden floor touching nothing else than the LED.

That might not be enough isolation, plus I'm guessing there's leakage through the transformer. Try standing with rubber-sole shoes or some other set-up. Maybe jump up in the air to see if the LED goes out. :D :D

P.S.

I believe it is isolated from mains because it didn't kill me so far.

Maybe you just need a better ground... Try touching a water pipe. Just kidding! I stupidly tried that experiment once and it was a very unpleasant experience! I'll NEVER make that mistake again! But, I did confirm the circuit was dangerous!

I had a few minutes for experiments this morning. I tried to jump and it did no difference. The light of the LED is steady, the intensity looks the same when I am moving. The intensity is similar to 1uA DC. When I connect to the LED via 22pF cap it don't light up. When connected via 2.2 nf cap (and greater) it glows as if connected directly.

It keeps glowing for long time (seconds) without any decrease in intensity. I though about (my) capacitance and AC charging it (me) when the AC is in the right part of wave. But since the current is ~1 uA and it should flow in only one direction (because diode) the 2.2 nF cap should charge very quickly and stop the current. The LED should glow only very briefly. Do you have an explanation for this?

The funny thing about electric circuits is that so many components are reversible. A speaker can be a microphone.* A motor can a generator. A step-up transformer can be reversed into a step-down. And an LED can be a photodiode.

When light shines on an LED, current will leak backwards through it due to the photoelectric effect. This will allow the revere cycle to discharge whatever the forward cycle charged, causing a continuous back-and-forth that keeps the LED lit.

  • I once plugged headphones into the microphone jack of my computer when I was in high school and [u]it worked[/u]! I was able to voice chat in AOL with someone in Georgia.

Mystery solved ... somehow. I got at home, tried to do some more experiments and didn't manage to make it "work" again. It worked consistently at thursday evening, friday morning but today nothing. I wonder what and why and how caused the interference...

Smajdalf: Mystery solved ... somehow. I got at home, tried to do some more experiments and didn't manage to make it "work" again. It worked consistently at thursday evening, friday morning but today nothing. I wonder what and why and how caused the interference...

It will have been some form of capacitive coupling. People often think of a transformer as providing almost ideal isolation between primary and secondary windings. It might close to the truth at DC, but at high frequencies it can be a very different story. Depending on the exact configuration of the windings there can be quite a lot capacitive coupling there.