Selfrunning cold electricity circuit from Dr.Stiffler

[EMdevices]

xee,  you realy need a scope meter to see what's happening.    Before I experimented, I thought the diode bridge and cap were pretty good at turning the AC waveform to DC, but I was wrong.   It's all about RF here, and I doubt DR Stifflers measurements now, and mine too. 

So, I'm not so sure this approach produces OU.  We're dealing with inappropriate measurements for the Output power.   We realy should get a supper smooth DC voltage at the output, by a chain of caps and inductors (low pass filter config) so that we can be sure what we are measuring DC and not DC with RF on it.

EM

[fritz]

 
... Have bought some LED samples today ...
will soon come up with nice measuring results ....
LED operated with DC/AC/RF and combination +
differential resistance...
I also "found" an optical power meter in my workshop -
it works up to 999mW - in a calibrated way up to 50mW -
anyway - for comparison/efficiency its usable.
The problem with white LED is that this power meter is
calibrated to wavelength - means I need to measure
at 3 wavelengths for "white"....LED...

rgds,

[DrStiffler]

... Have bought some LED samples today ...
will soon come up with nice measuring results ....
LED operated with DC/AC/RF and combination +
differential resistance...
I also "found" an optical power meter in my workshop -
it works up to 999mW - in a calibrated way up to 50mW -
anyway - for comparison/efficiency its usable.
The problem with white LED is that this power meter is
calibrated to wavelength - means I need to measure
at 3 wavelengths for "white"....LED...

rgds,

@fritz

I am no optics expert by a long shot, so just my two cents worth. Old style LEDS mixed Red, Blue and Green to come up with a White beam, while the newer units now coat the emitter with a phosphor much like an FL tube to enable the emission of the CLOSE to white light.

Most consumer light meters are weighted for the response of the human eye and do indeed have weighting filters added making it difficult to measure LED output with the  same accuracy as the standard sources such as Tungsten, Mercury, Fluorescent and Sodium. You also need to now what type of power measurement you are making, lm/m2, lm/ft2 for example.

Without knowing your input device, I would doubt you can measure an entire LED array do to the beam angles not being convergent. I take a representative number of selected LEDS, measure the output from each and take an average, then monitor only one selected LED and extrapolate over the total number. This method allows for ease is seeing the effects of adding and subtracting LEDS from the change and the approach of optimum matching to the driver.

Looking forward to your results; What circuit and coil arrangement will you be using for the testing??? Just wanting to match Apples to Apples......

[DrStiffler]

xee,  you realy need a scope meter to see what's happening.    Before I experimented, I thought the diode bridge and cap were pretty good at turning the AC waveform to DC, but I was wrong.   It's all about RF here, and I doubt DR Stifflers measurements now, and mine too. 

So, I'm not so sure this approach produces OU.  We're dealing with inappropriate measurements for the Output power.   We realy should get a supper smooth DC voltage at the output, by a chain of caps and inductors (low pass filter config) so that we can be sure what we are measuring DC and not DC with RF on it.

EM

I will be interested to see what measurement protocol you end up with and your filter configuration.

If you plan on being accurate, you will have to consider the high field density around a working circuit and the extent to which it radiates. You might think you have an adequate filter to the measurement points, yet your instrumentation will be affected by the proximity to the field. Extended lead length from filters to measurement equipment is a no/no for sure. Addition of ferrite beads can cause spurious harmonics that are not actually generated by the circuit and will upset the reading, giving false indications, as will inductors couple to the local field.

A realizable measurement protocol is very restrictive and in the process of implementation may indeed detune the circuit to where it is no longer able to provided energy node interface.

[DrStiffler]

Some points of Fact that should be of interest to all following or working on circuits from this thread. Because not all of you or maybe not many of you visit my site often, I will from time to time post here to update on current status.

A) If your SEC amplifier has both ends of either the Primary or Secondary coil connected, SEC power amplification WILL NOT take place. One end of each coil MUST be OPEN for SEC interface to excess energy.

B) The purpose of a SEC circuit is not to see how many LEDs you can drive, rather the purpose is to see how many you can drive where various measurements indicates Pout/Pin > 1. This can be seen in various configurations from 1 to 90 LEDs. The current in the LED string and the average voltage drop across a representative LED time the number of LEDs is what you want to compare to the input. The light emitted energy is a bonus and not a required measurement to insure SEC interface operation.

C) 99% of the working SEC amplifiers are or can be designed by conventional and proven EE methodologies. What this means is that load matching, current limitations, decoupling etc., is all know and fully applies to most of a SEC circuit.

D) If you cannot detect a HV great enough to light a Neon from the AV Plug (75-90volts) then you will not see anything unconventional from your circuit, you will not see power amplification.

E) If you can touch any portion of your operating circuit with the free end of a wire at least 14" long (one end free of any connection) and see either an increase or decrease of LED intensity, your circuit is NOT operating properly and the Cohered energy is not able to be properly utilized within. A stable SEC circuit will show a barely detectable change in LED output with the addition of a wire probe.

Thanks and hope this might help....