Lasersaber strikes again. A joule thief king ?

[d3x0r]

Picture 1 : high voltage (24V+) looper circuit.... 12mH power-collector, 7mH across LEDs
The collector gets a high flyback pulses, with an additonal rise at the end when the gate is off...


Picture 2 : 7mH power-collector, 12mH across LEDs...
top picture - lower voltage; 7.4V, shows the progressive ring-up after a burst of oscillations
middle picture - high voltage - probably up to 30V with a high flyback on the collector winding
bottom picture - zoomed in, centered better so 0 of yellow trace is at -V, so the 0 line should be equal to V...which shows it gets a voltage above the input.  Being clever, an attempt to put a diode from the collector back to positive actually increases current draw 20-30mA.  Not shown, while paused at the end, i accidentally shorted power to ground and popped the transistor and 2 diodes.


the first picture shows a similar signal to the akula flashlight (v4?) that he was demonstrating on a ferrite core.  Other than being like a 0 duty cycle between pulses... *shrug* Turns out it's basically a useless power, if I just touch a diode to the positive with a finger, that high part goes away.... (further expiriment after fixing it, not shown).


Only thing I do think is good is a diode to catch the low pulse on the gate and dump it to ground... at least keeps the gate from under-voltaging which doesn't seem to affect it... but does reduce power draw and increases brightness.


These white LEDs show much brighter on camera than they do in person... so many times I say 'barely on' and they really are very dim... less than 25% brightness probably, but the camera shows them quite bright.


Don't waste your time watching...
P1 : https://www.youtube.com/watch?v=dPhmjvt8WhI
P2 : https://www.youtube.com/watch?v=L_71bdAI9_E

[conradelektro]

P1 : https://www.youtube.com/watch?v=dPhmjvt8WhI
P2 : https://www.youtube.com/watch?v=L_71bdAI9_E

@d3x0r: Watching your two videos I see similar waveforms which I encountered when playing with this circuit.

I think that one can get a very low power draw if the transformer is just right and no other component besides the MPS18 and the three LEDs are used (but the LEDs will be rather dim).

Once some additional components are introduced (like in your circuit shown here http://www.overunity.com/14591/lasersaber-strikes-again-a-joule-thief-king/msg407881/#msg407881, a resistor, capacitor or diode) the LEDs will become brighter and the power draw goes up.

My experiments showed that about 1 mW are needed to make a LED shine nicely, but one needs the specified 60 mW to make a LED shine as intended by the manufacturer (really eye piercing bright).

Greetings, Conrad

[d3x0r]

@d3x0r: Watching your two videos I see similar waveforms which I encountered when playing with this circuit.

I think that one can get a very low power draw if the transformer is just right and no other component besides the MPS18 and the three LEDs are used (but the LEDs will be rather dim).

Once some additional components are introduced (like in your circuit shown here http://www.overunity.com/14591/lasersaber-strikes-again-a-joule-thief-king/msg407881/#msg407881, a resistor, capacitor or diode) the LEDs will become brighter and the power draw goes up.

My experiments showed that about 1 mW are needed to make a LED shine nicely, but one needs the specified 60 mW to make a LED shine as intended by the manufacturer (really eye piercing bright).

Greetings, Conrad

Ya; going to try with some different coils soon (got some of the tiny ferrite pots).  The cap ended up being required... when I first started I had some copper tape I made the foil windings with; which ends up being that small capacitance.... and it worked just as it was; I had another coil that also worked without the cap and without the foil, but it was really tricky to start oscillating.  If I simplify it back to just the transistor I still wasn't getting the minor current draws... maybe I had excess LEDs...
The 10M resistor helps it start... once it starts, it can be disconnected; but it really doesn't affect the current significantly
but also the type/brand of LED affects the wave forms... those LEDs are at least the same that LS is using in the latest flashlights.

[conradelektro]

but also the type/brand of LED affects the wave forms... those LEDs are at least the same that LS is using in the latest flashlights.

@d3x0r: yes, the LED or LEDs are a big factor. A red LED is really bad in this circuit. Also the direction of the LEDs is crucial. In one direction they are brighter but then the power draw goes up.

A good trick usually is to put a diode in series with the LED because the LED by itself is a bad diode (a lot of reverse current). But in this circuit it destroys oscillation (if no resistor is used, with the resistor it makes the LEDs brighter and increases power draw). So, the circuit seems to need some reverse current through the LED or LEDs if no resistor is used.

In general I have the impression that this circuit becomes a circuit with the "usual power draw" once it is done right e.g. with a diode to keep the base from going negative and with a resistor which biases the base of the transistor in the right way. It then also works with other transistors (not only with the MPSA18).

The most useful part of Lasersaber's JSRLooper experiments seems to be the idea to run a low power Joule Thief from a big electrolytic cap which can be charged by a little hand cranked dynamo or by charging the cap quickly from a battery which is then left behind.

Greetings, Conrad

[d3x0r]

Today I have success! 
Yes the lights are barely lit.  Diode in this circuit kills all function; although I do need to keep cap, which is just 2 wires wound together for like 12 turns... measures something like 200 pF.... (was part of the cap study for TPP something pyramid power unit... )
Used laser saber's small core, primary measures 0.97mH (almost 1) and secondary is 2.5mH on tiny ferrite core.  The advantage of this core is the little wire being used is very low resistance....


I think somehow there is a small capacitance between how laser saber has placed his parts... like the circuit board itself as a dielectric between the top LED (furthest from base) and a connection to either power or ground (although I think cap to power is lower current draw, have not confirmed).
The wave of this coil is entirely different from previous waves.
Ran for near 15 minutes... but that's the full 10M ohm resistor + 400k.... (56mF)
I'll have to edit the video some, was a lot of dead space, eventually turned on some music in the background to kill some time, and eventually got a meter in shot across cap...
Will have to include a picture of my 'cap' too; just cause I like it...
I guess, the power supply was set at 11 volts, but I have a diode to the power supply ground and a diode to the power supply positive, so there's 1.5V drop... when I re-charged the cap at the end, it only went up to 9.5V, and that puzzled me until just now


It's enough light to read by I suppose... could be slightly better with slightly lower resistance.


more probably after many hours...
ya about 20 minute useful runtime I guess; it's still going after I stopped recording, caught up on messages and finished this one... 3 white LED
But they are MUCH dimmer than the target dimness I was going for...

-----
Edit:
LOL Youtube was kind enough to compile a music list for me....



Your video may include music that is owned by a third party.To hear the matched music please play the video on the right. The video will play from the point where the matched content was identified.
Your video is available and playable.
Here are the details:
"Immortal Technique feat. BigZoo-Positive Balance", sound recording administered by:
INgrooves
"Pink Floyd-Money", sound recording administered by:
WMG
SME
"Green Day-Brain Stew / Jaded (Radio Version)", sound recording administered by:
WMG
To learn more about how claims impact your videos ....


Guess I'll have to just truncate it.


https://www.youtube.com/watch?v=30xY6yhOHYg