Experiments and Anomalies

[PhysicsProf]

A little role reversal:

Have you ever been really frustrated  trying to decipher and replicate a device in a video, spending time away from your family or friends because you were being deceived.

Have you ever felt the anger for the wasted time when you found out it was a fake?

The video in question was picked up on the featured video page on Stefan's site OU.com. This will be viewed by tens of thousands, maybe a few hundred will spend the time trying to decipher the video and build it.

They may never see the disclaimer over here, and spend valuable time trying to replicate, only to be frustrated and angry at the end of the day.
...

"The video in question" -- are you referring to TinMan's vid, or whose?

@all:  I'm interested primarily here in the vid by LaserSaber, the 5 HOUR run with two LEDs glowing, running on a 1000 uF cap. 
Are you (ION or TinMan or others) suggesting that LS is deceiving us in this 5-hour run?

[TinMan]

"The video in question" -- are you referring to TinMan's vid, or whose?

@all:  I'm interested primarily here in the vid by LaserSaber, the 5 HOUR run with two LEDs glowing, running on a 1000 uF cap. 
Are you (ION or TinMan or others) suggesting that LS is deceiving us in this 5-hour run?

No Steve-they are refering to mine.LaserSaber's devices always seem to work as shown.
Please see my last post just befor your's.

[PhysicsProf]

 Got it, thanks, TinMan. 

  I've had some long run times with Joule-thief type circuits... but FIVE HOURS is truly remarkable!

[Hoppy]

TinMan,

We all share your frustration at so much crap and deception on the fora. However, joining- in is akin to running with the devil, no matter how well intentioned. Just my opinion.

Hoppy

[ion]

Prof:

"The video in question" -- are you referring to TinMan's vid, or whose?

I was referring to the Tinman's video that was later admitted to be a fake.

It's hard to pull a faked video back in, you lose control once you post it and it goes viral, people believe it as truth, they download it and re-post it on their channel or link it to their channel. Only a handful know the original posters intentions or whether it was real or a spoof. Most believe it as a working device and it spreads like fire. Can you put out the flames?

I still see the faked TPU videos by Marco and Dave Clark being re-posted now and again as true working devices by unsuspecting but well meaning newbies. This is still happening over six years later.

Personally I don't want that kind of Karma even if my intentions were good, as it has large, long term negative repurcussions.

Everyone can do as they please, but there are consequences.

[Grumpy]

RomerUK was another one. 

On the contrary, I've never been able to find the video that spherics alluded to of a working device.

[Peterae]

What many don't seem to realize these days is that if someone posts a device and it's turns out to be a fake and was posted as working then they stand a very good chance of being billed for everyones wasted time, especially if they are easily tracked down, it nearly happened to Romerouk and i am sure it's only time before a case comes up, especially big investors in time and money.  

[PhysicsProf]

  Meanwhile, "back at the ranch",  LS has posted another vid.  This time, the run is for TWELVE hours.  He reduced the number of diodes from 6 to 3 and says this helps.  Yes, I think this is exciting stuff.

https://www.youtube.com/watch?v=T5EXisVd9Xs&feature=em-uploademail

[Slider2732]

It's funny how LS's diagram already showed only 3 diodes.
I'm wondering how reducing the number affected the run, but am glad it worked as i'm low on Germanium's lol

[Hoppy]

It's funny how LS's diagram already showed only 3 diodes.

Its also funny how it stays lit for 12 hours.

[ion]

Hoppy said:

Its also funny how it stays lit for 12 hours.

Never believe what you see in videos as if it were truth, regardless of the banner ads.

I have been simulating the LS circuit in LTSpice and my waveforms look fairly identical to the ones he has shown, and as expected the circuit produces a very sharp narrow pulse of about 400uA through the LED's.

Once you understand what is going on, it is not so mysterious. The LED's are being hit with a high peak current pulse to get those photons moving, once they are shocked out of the nest, there is no bringing them back. Of course, they are fairly dim in reality, despite the theatrics at the beginning of the latest video. Note that LS charges the cap, then puts his fingers on the board to add some bias from body resistance, which puts a lot of current through the LED's for the "bright shine on the wall" part of the show, but the cap is quickly discharged. He then charges it again, repeats some more theatrics before the final charge for the long run. Now the LEDs are fairly dim compared to earlier.

It is the capacitance coupling between the windings that is not shown on his schematics, but it completes the circuit of what would be a normal blocking oscillator with very weak coupling to the base of the transistor. This produces the narrow pulse.

There are more stable means of generating such a pulse without having to rely on the hit or miss properties of the inter-winding capacitance. The leakage in the diodes help to get the transistor biased up a bit, but they are not really needed, you could just add a high value resistor to the plus supply to add some bias.

Green trace is current through the LEDs, blue trace is base of the transistor.

[PhysicsProf]

It appears that for the 12-hour run, he is still using one 1000uF cap (not sure?) and two LED's (green this time).  So I get:


E=1/2CV^2 = 1/2 1000uF (81)   = 40.5 mJ  

 The two LED's are run for 12 hours on that, or roughly

40.5 mJ/ 12*3600s = 0.94 micro-watts  (average) for both LED's

   I've suggested to LS a few tests.

[ion]

Lasersabers video title:

12 hour run on 3500uF - SJR Looper V2 LED Driver

http://www.youtube.com/watch?v=T5EXisVd9Xs

Everyone should watch this video very carefully. Note how dim the LEDs are until he puts his thumb on the back of the board to "bias up" the transistor. After all the playing with bright light on the wall he takes his thumb off the circuit and then it goes back to very dim. Those LEDs are at the threshold, I estimate 3uA to 6uA average current.

[Hoppy]

Lasersabers video title:

12 hour run on 3500uF - SJR Looper V2 LED Driver

http://www.youtube.com/watch?v=T5EXisVd9Xs

Everyone should watch this video very carefully. Note how dim the LEDs are until he puts his thumb on the back of the board to "bias up" the transistor. After all the playing with bright light on the wall he takes his thumb off the circuit and then it goes back to very dim. Those LEDs are at the threshold, I estimate 3uA to 6uA average current.

Also carefully watch and note how when the LED's extinguish, he reconnects the cap to re-charge the device. Is he using the TinMan trick??

[PhysicsProf]

Also carefully watch and note how when the LED's extinguish, he reconnects the cap to re-charge the device. Is he using the TinMan trick??

  Please explain what you mean.  How does the "trick" work?

[Hoppy]

  Please explain what you mean.  How does the "trick" work?

By using a super-cap disguised by sheath as a standard cap. In the LS device another perhaps smaller super-cap would be needed concealed in the pot core, which is re-charged by the external cap. Unlikely but possible thus my ??

Hoppy

[ion]

I don't think LS is faking his device, it is real with a lot of theatrics to mislead the watcher. The numbers seem to be in the right order of magnitude for LED's at the threshold of lighting. (3500 uF and a 10 volt fresh battery)

I am very wary when scientific presentations begin to resemble carnival acts, banners and all. Shades of Joe Newman and Dennis Lee.

If you are not familiar with their stories, watch their original presentations in the mid 1980's compared to those much later in their life and you will see what I mean.

Why is it that the spotlight and attention is so alluring that people abandon good science in favor of a "wow" presentation?

Human Nature?

[PhysicsProf]

I don't think LS is faking his device, it is real with a lot of theatrics to mislead the watcher. The numbers seem to be in the right order of magnitude for LED's at the threshold of lighting. (3500 uF and a 10 volt fresh battery)

I am very wary when scientific presentations begin to resemble carnival acts, banners and all. Shades of Joe Newman and Dennis Lee.

If you are not familiar with their stories, watch their original presentations in the mid 1980's compared to those much later in their life and you will see what I mean.

Why is it that the spotlight and attention is so alluring that people abandon good science in favor of a "wow" presentation?

Human Nature?

I don't find it to "abandon good science in favor of a "wow" presentation" ; LS says he wishes to develop a flashlight; nothing wrong with that.

But what types of measurements would you like to see?

[TinMan]

Quote Hoppy: Its interesting that everything was telling me your device was a fake, especially the square waveform.

The wave form was real Hoppy-nice and clean as you seen it. The pic below show's a current trace aswell-blue trace. This is across a 1.5 ohm resistor-seems to have some inductance there as well-bad resistor?.

[Hoppy]

I don't think LS is faking his device, it is real with a lot of theatrics to mislead the watcher. The numbers seem to be in the right order of magnitude for LED's at the threshold of lighting. (3500 uF and a 10 volt fresh battery)

Its this one with the tiny SMD ceramic cap that really gets me wondering: http://www.youtube.com/watch?v=FOUrRXINDqY&feature=em-uploademail

[ion]

I don't find it to "abandon good science in favor of a "wow" presentation" ; LS says he wishes to develop a flashlight; nothing wrong with that.

But what types of measurements would you like to see?

Agreed, nothing wrong a wow presentation if it's real and not just showmanship, posers abound on youtube, as you probably know.

I would like to see him show the measured current per pulse into the LED, both peak and average, pulses per second (frequency), duty cycle, Initial measured charge on the capacitor, measured value of the capacitor.
As a bonus, he could include the effects of dielectric absorption in the capacitor. This alone could extend the run time of a barely lit LED. Dielectric absorption is a battery effect that goes beyond the normal 1/2 cv^2 formula and is present to a large degree in big electrolytics.

Developing a bright, usable flashlight using pulse mode drivers has already been done many times over, the science is well advanced.

Is LS attempting to develop a flashlight for the blind, which would be an interesting project.

I saw such a project while employed at "Bionics Inc" in the mid 60's, but it used ultrasonic transducers.

Or is he attempting to build a flashlight which utilizes a source of as yet unknown energy? That would be more in line with this forum.

He really has stated just a long run time as his goal, usable light output has not been mentioned.

There is always hope.

Hoppy said:

Its this one with the tiny SMD ceramic cap that really gets me wondering: http://www.youtube.com/watch?v=FOUrRXINDqY&feature=em-uploademail

Yes, that one caught my eye also, and if real would point to some extra energy being captured from somewhere.

SMD ceramic caps would not have that kind of duration, though it be short....but SMD electrolytics or tantalums would!

[Hoppy]

Hoppy said:
Yes, that one caught my eye also, and if real would point to some extra energy being captured from somewhere.

SMD ceramic caps would not have that kind of duration, though it be short....but SMD electrolytics or tantalums would!

I've had another much closer look at the ceramic cap and can see with the small LED device, he has three SMD caps stacked on top of each other and they could be tantalums, looking at their body colour. The large device also has stacked capacitors, two or could be three, its hard to see clearly.

[Grumpy]

http://articles.latimes.com/2013/apr/27/business/la-fi-made-in-california-eye-20130428

It's a start.

[PhysicsProf]

Agreed, nothing wrong a wow presentation if it's real and not just showmanship, posers abound on youtube, as you probably know.

I would like to see him show the measured current per pulse into the LED, both peak and average,

How do you do this, assuming we're in the uA range?  A simple meter might give some sort of average.  Would you suggest voltage drop across a tiny test resistor, or what?
  I'm really interesting in measuring the output power... may not be so easy.  Light output could be measured however.

pulses per second (frequency), duty cycle, Initial measured charge on the capacitor, measured value of the capacitor.

These are good points, these one can do quite easily I think. Thanks.

As a bonus, he could include the effects of dielectric absorption in the capacitor. This alone could extend the run time of a barely lit LED. Dielectric absorption is a battery effect that goes beyond the normal 1/2 cv^2 formula and is present to a large degree in big electrolytics.

Perhaps better to stay away from electrolytic caps, just use ceramic caps?

...

Hoppy said:

Its this one with the tiny SMD ceramic cap that really gets me wondering: http://www.youtube.com/watch?v=FOUrRXINDqY&feature=em-uploademail

Yes, that one caught my eye also, and if real would point to some extra energy being captured from somewhere.

SMD ceramic caps would not have that kind of duration, though it be short....but SMD electrolytics or tantalums would!

Wait, are you guys referring to the early part of the "5 hour" vid by LS where he shows the device running on the ceramic cap alone? (no electrolytic cap)

I would try severak tests running on that cap alone, no electrolytics.

[Hoppy]

Wait, are you guys referring to the early part of the "5 hour" vid by LS where he shows the device running on the ceramic cap alone? [/b] (no electrolytic cap)

I would try severak tests running on that cap alone, no electrolytics.

Yes, except that I see two or maybe three stacked caps in parallel and possibly tantalum, not ceramics.