Lasersaber strikes again. A joule thief king ?

[d3x0r]

coil-cap...
I dunno how long... a few feet of wire coiled together to make a capacitor...

[d3x0r]

Found this using super caps and a joule theif... LEDs in parallel on the collector side
http://www.lafamillecreative.blogspot.com/2012/02/amelioration-de-la-lampe-de-poche.html  He says he found them brighter in parallel than in series... might have to toy with that some...

(background noise; guy talking vaguely about electret effect... http://youtu.be/tZf0yhe4B80?t=11m29s )

was remembering a thing about electrets, and that materials used in supercaps demonstrated electret properties... somewhat related to 'dielecric relaxation' which causes caps to recharge after quick discharge and then allowed to relax...

http://edn.com/design/microcontroller-mcu/4214488/2/Self-adaptive-MEMS-vibration-energy-harvester-targets-low-frequencies
At vibrations of less than 0.2 grams at 50 Hz, Leti's system was able to output 3V, reaching an output power of about 10 microwatts per gram of seismic mass. The resulting mechanical-to-electrical conversion efficiency was an impressive 60 percent.

http://peswiki.com/index.php/Article:Free_Electric_Energy_in_Theory_and_Practice#The_Electret_Effect

More info (from here even) http://www.overunity.com/9878/captret-capacitor-and-electret/195/wap/

http://open-source-energy.org/?topic=1168.0

http://www.printedelectronicsworld.com/articles/big-future-for-titanium-compounds-in-the-new-electronics-and-electrics-00004744.asp?sessionid=1

scholarly articles on 'supercap dielectric relaxation'

------
But; Electret effects are very very low current (probably in the order of nano-amps) or have to be used in very high frequency (Ghz-Thz)?

But; my current setup I put 4 10F 2.7V caps in series (2.5F 10.8V result) and can charge it to from 4V to 8V in under a minute with this hand crank (thinking maybe a heavy flywheel you could rub against something... like this little windup toy truck things kinda) it would be less manual work... at least to top if off... it is quite a bit more torque required to be applied at 4V than at 7V... But then this lasts for an hour to go from 8.4V to 4.(1?)V  .... I was watching the voltage drop at 8.33 to 8.21 and it was about 4 maybe 5 seconds per 0.01V drop. so for 4.00V drop it should have been like 1600-2000 seconds or half an hour... at 30 minutes it was only 6.2V (rough memory)... so I let it sit another half hour and it got down to 4.5V or so... But I had the meter disconnected during the run, rather than when I was measuring I had the meter attached so it probably made the voltage drop faster when measuring...

And, this is running at a high brightness mode, with 4 LED and diodes from collector to power and from base to emitter connection to 10M resistor used to start the circuit and then disconnected.  Also made sure the core is clamped closed tightly...

while testing I had it just wrapped tightly with electric tape, at certain power levels the core would start to 'hiss' at me... so i applied more pressure... and at other power levels that it wasn't making a noise, I found at 14V circuit was drawing 0.01A (as reported by power supply, which is only very rough estimate) but applying additional pressure on the core made it go back down to 0.00A  (probably 0.009 or less reads 0.00).
coils measure 1mH (11-12 turns) and 2.5mH(20-24 turns) lost counts... but it's 5-6 windings across but my outer layers make me skip windings because I wasn't winding with a drill just by hand, and therefore inconsistent pressure.

So I dunno an hour per minute hand charging, might make it useful in caves( http://www.bosnianpyramids.org/ ) ... but maybe just converting to use batteries and run for days

[d3x0r]

Added hand crank generator to charge, and an indicator LED with a zener to avoid over-voltage...
The generator can generate spikes(just a hypothesis) that kicks the oscillation into a higher power draw.  It is a chirp sort of wave whose frequency goes higher and higher until the 1n4148 on collector to power cannot divert the over-voltage back to power.... they go to +26V...


Happens more with a 2n2222
Maybe if I add a smaller cap closer to the rectifier it would help.  Without the rectifier, the caps just turned the generator as a DC motor  

[TinselKoala]

So... UTSource does it again! About a week or so ago I ordered 10 ea. 2sk170 from them. Three dollars for the ten transistors, and four dollars shipping. And they arrived yesterday! They advertise 3 weeks delivery but usually beat that, but this time they really did well. They have a funny pricing system on their own website but now they have an EBay store and that is much simpler to deal with.

So now I have a small handful of special FETs. What to do? So I made a simple little test circuit that is an electrostatic field detector. Remarkably sensitive, it can detect the motion of a charged bit of plastic a meter away.

3V input from a CR2032 to the terminals on the left, and the open end of the 47K resistor on the Gate pin is the "antenna" for the EF pickup, and the blue LED with cathode at the transistor Drain and anode at V+, and Source of the transistor to V-  completes the circuit.

[Pirate88179]

TK:

What a cool little circuit that is.  Can you light the led from "electro-smog"?  Like radio/tv stations?  Would a large antenna have to be added? (and/or an earth ground?)

I think it would be great to be able to light an led for "free" just from energy we are being bombarded with every day.

Bill