Overunity Device by Tanju Argun (Moderated)

[gyulasun]

I will be doing the pure resistance experiment tomorrow because it is midnight here now.
LED spec says 28-30 volts and 700 miliamperes. That gives me an internal resistance of  40 ohms. 3 in series is 120 ohms and 7 parallel branches give me an overall internal resistance of 120/7= 17.14 ohms
So instead of the LEDs I will connect a 17 ohm resistor and see what happens.
I must also mention another peculiar thing which I witnessed.  In the video I have 15 LEDS and my my output current was still 0.3 Amps I later added 2 more parallel branches which brought the LED qty to 21.
And to my surprise the current did not increase. Does this prove that radiant energy "loves" load?

Hi Tanju,

Well, it is a good question but I would say as long as we can explain things by conventional science no need to turn to 'fancy' things. 

[I just noticed member cheors mentioned the Zener like nonlinear LED behaviour while I prepared this answer.]

I mean the followings: you surely know LEDs have a relatively steep VI characteristics i.e. assuming voltage source_like drive a small forward voltage change across them may cause high forward current change in them i.e they behave like Zener diodes do.
You connected 3 LED arrays (having 28-30V forward voltage each) in series to get a 84V-90V forward voltage range and you also paralleled 7 such arrays and you measure 80V forward voltage at 300 mA overall forward current.

Now you wish compare this LED power data to a resistor power data. It is not comparable, because a resistor has a linear VI characteristic and no Zener diode-like behaviour. I know you know this and sorry to mention but we tend to think that way.
You measured 80V across your LED array, dividing this by 3 it gives 26.6 V forward voltage drop for a single LED out of any 3 in the series strings. This means such LED receives less than its specified 28-30V hence its own current draw must also be much less than the specified 700 mA of course, you measured 300 mA for the 7 paralleled branches. We need to divide 300 by 7 = 43 mA to get current in any of the branches. This is possible due to the Zener-like characteristics. 

From your resistor load tests the inner resistance of your 80V 'voltage source' comes as if it would change between 23 Ohm and 48 Ohm values: is there any series component like a diode in the output going to the LEDs that may change nonlinearly with load current? If there is not any series components, then what may cause the change?

Could you show a scope shot across your 80 V (presumably DC) voltage source (that collects the 'radiant') while feeding the LEDs? First use DC coupling for the scope input and just for curiosity, check it in AC coupling to increase AC amplitude resolution to see how pure DC the high value capacitor(s) provide (i.e. how much ripple voltage is there if any).

Gyula

EDIT I attached a LED VI characteristic curve from the data sheet member Itsu provided earlier. Your LEDs may have even steeper VI curves than the type shown.

[Tanju]

Hi Tanju,

Well, it is a good question but I would say as long as we can explain things by conventional science no need to turn to 'fancy' things. 

[I just noticed member cheors mentioned the Zener like nonlinear LED behaviour while I prepared this answer.]

I mean the followings: you surely know LEDs have a relatively steep VI characteristics i.e. assuming voltage source_like drive a small forward voltage change across them may cause high forward current change in them i.e they behave like Zener diodes do.
You connected 3 LED arrays (having 28-30V forward voltage each) in series to get a 84V-90V forward voltage range and you also paralleled 7 such arrays and you measure 80V forward voltage at 300 mA overall forward current.

Now you wish compare this LED power data to a resistor power data. It is not comparable, because a resistor has a linear VI characteristic and no Zener diode-like behaviour. I know you know this and sorry to mention but we tend to think that way.
You measured 80V across your LED array, dividing this by 3 it gives 26.6 V forward voltage drop for a single LED out of any 3 in the series strings. This means such LED receives less than its specified 28-30V hence its own current draw must also be much less than the specified 700 mA of course, you measured 300 mA for the 7 paralleled branches. We need to divide 300 by 7 = 43 mA to get current in any of the branches. This is possible due to the Zener-like characteristics. 

From your resistor load tests the inner resistance of your 80V 'voltage source' comes as if it would change between 23 Ohm and 48 Ohm values: is there any series component like a diode in the output going to the LEDs that may change nonlinearly with load current? If there is not any series components, then what may cause the change?

Could you show a scope shot across your 80 V (presumably DC) voltage source (that collects the 'radiant') while feeding the LEDs? First use DC coupling for the scope input and just for curiosity, check it in AC coupling to increase AC amplitude resolution to see how pure DC the high value capacitor(s) provide (i.e. how much ripple voltage is there if any).

Gyula

EDIT I attached a LED VI characteristic curve from the data sheet member Itsu provided earlier. Your LEDs may have even steeper VI curves than the type shown.

Very informative input. Thank you very much. But still does not answer my questıon. How can I get such high luminious intensity even though the Leds are subject to a voltage lower than  the low threshhold voltage 26.6 vs 28
Tanju

[gyulasun]

Well I do not know it yet.  Perhaps a scope shot across the 80V capacitor and an actual schematic on the output parts could shed some further lights on an answer.  And also, a dependable light meter to check lux emitted would also help. Brightness to our eyes can be decisive, unfortunately.

Gyula

[e2matrix]

Well I do not know it yet.  Perhaps a scope shot across the 80V capacitor and an actual schematic on the output parts could shed some further lights on an answer.  And also, a dependable light meter to check lux emitted would also help. Brightness to our eyes can be decisive, unfortunately.

Gyula

Hi Gyula,  I'm guessing you meant "Brightness to our eyes can be deceptive" ?   Rule of thumb I've learned from flashlight forums is it takes almost double the brightness in lumens to be a noticeable difference to the human eye.   
Also I'm not sure if I'm up to speed on the discussion of LED brightness here with regards to this circuit but LED's can look quite bright running on pulsed DC or AC so unless you have a scope of the voltage it might appear less than what the lower input voltage an LED needs. 

[e2matrix]

I will ad that Tanju's setup is very interesting and I once again think it may be a confirmation that flywheels in combination with the right motor and generator setup can do some fascinating things probably including OU.