Selfrunning cold electricity circuit from Dr.Stiffler

[hartiberlin]

Maybe LEDs do create their own power. Then again, maybe Ohm's Law does not apply to LEDs with constant current sources. These are real measurements. I offer no explanations.

Hi Xee,
how did you measure the 2 mV across the LED and resistor ?
Maybe your voltmeter was jammed by the RF power ?

What does it show without the 100 Ohm resistor ?

[DrStiffler]

If you are suffering 'HIGH' LED failure in SEC circuits driving LED's, the problem has been solved or greatly reduced. It does although require additional passive components but the cost is less than a handful of good LED's.

The circuit was added to the bottom of the SEC page; www.drstiffler.com/ce4.asp

One thing to note that is not discussed in the circuit overview is that the 22uH coils must be separate from the balanced inductor, just selecting an inductor of higher value will not work. The input chokes to the inductor are required and attempts to circumvent this will result in the matching and loading not being met and proper isolation from the load and safety diode and the driver are not achieved.

[xee]

@hartiberlin,
Reads 0.94 volts across diode without 100 ohm resistor and LED lights brightly. RF could be problem but test is at output of full wave rectifier and seems to be DC. I do not have scope so I can not check using that. If diode leads are reversed, LED does not light and voltage across diode reads 41 millivolts without resistor. Using a 0.1 uF capacitor in series with forward diode prevents diode from lighting and voltage reads 0.77 volts across diode (without 100 ohm resistor). So there seems to be some kind of voltage being generated by RF. Maybe from parasitic coupling. LED is 15,000 mcd super bright. Meter is digital Radio Shack. Very strange. I will try some more tests today but do not have a lot of time. Just wanted to see if anyone else had seen something similar.

[jonesbeene]

Thanks for banning the toxic negativity of certain skeptics, Harti.

I fear that the incessant mindless chirping of nay-sayers, who add nothing relevant into the mix - like that one, has once again caused DrS to abandon further contribution to the group for a while.

Funny that the skeptic chose to ignore the last contribution by xee, which is provocative and indicative of the strange kind of LED input, which eventually derives from series negative resistance, but the simple test is not conclusive of anything admittedly  --- except of what we have been observing in operating circuits. Normally negative resistance is a decrease in current with rising voltage, but the reciprocal of that (near reciprocal) is closer to what seems to be happening. IOW in a series loop (the AV plug), and with an open-ended secondary to supply the capacitive coupling (which a few experimenters have dropped unfortunately) there seems to be a strange kind of stasis which develops in the loop, to push the light output - such that the expected voltage drop of series LEDs seems to disappear, at least for part of the cycle.

The more series LEDs which are added, up to a point, the more actual work which is produced from the DC input which actually drops instead of rises. The drop in P-in is a result of added resistance, true, but the added light output is the big surprise. Lest we not forget -- the "work" being done here is solely in *light output* and if you have followed Stiffler's testing you have noted that the expected heating does NOT occur, commensurate with the light output. This is not standard electrical engineering, nor is it standard physics.

Please keep this in mind - this is a LIGHTING device first and foremost. It may have, on occasion, showed some indication of being amenable to self-power - but it is not optimized for that, and I fear that a few experimenters are looking for that alone and neglecting the fact that the energy output here is in photonic; and that lighting is the MOST factor to optimize. Do not attempt to build this with the anticipation that it will magically self-power. Personally, I think the only way to achieve self power will be to focus the light output onto a very efficient photocell (triple junction solar cell) and use that DC from the photocell to power the oscillator.

The downside, and the fly-in-the-ointment, is that these LED devices are NOT engineered for higher voltage, and the failure rate limits the number which can be added in the plug. My guess is that between 30-50 is optimum for sustained operation at moderate failure rate. Plus the best "quantum well" diodes, like those from Cree and Nichia - are far more expensive than the eBay garden variety rainbow.

Although the run-of-the-mill red and green diodes will show the series negative resistance effect, it is to a lesser extent. If you can afford it- go with quantum-well diodes direct from either Nichia or Cree. Those two companies own ALL the important patents on quantum well LEDs, and any other suppliers will be just adding on a surcharge.

Jones

[hartiberlin]

Dr. Stiffler has added some interesting new
output load circuits.

Have a look at his website, cause he has a strict copyright
notice, so I don?t copy it over to here:

http://www.drstiffler.com/ce4.asp

Regards, Stefan.
P.S. In this moment I don?t have much time to experiment
unfortunately, so before christmas all is hektik again...
too much other work needs to be finished..
I hope some other people find the time to experiment on with this..