Lasersaber strikes again. A joule thief king ?

[MarkE]

@MarkE:

In this circuit http://www.overunity.com/14591/lasesaber-strikes-again-a-joule-thief-king/msg401872/#msg401872 (which you seem to refer to) there is a LED with the cathode on the base and anode on the emitter limiting the reverse Vbe to -2 Volt. A 1N4148, as you suggest, would be better (Vbe of -0.7 V).

Skywatcher's circuit had nothing.  I see that you suggested an LED which should work provided it is fast enough.

With a 100K resistor in front of the base of the transistor and a 1N4148 to limit Vbe (see the attached circuit and photo, I think you called it the "shunt configuration") I can now run the circuit with 1.2 to 1.5 V from a AAA battery. But the power draw is considerable, 20 to 30 mA at 1.2 to 1.5 V. The LED lamp gives useful light (although far from nominal brightness).

100K seems a lot.  You could try simulations or trial and error to get that down.  I would think that if the circuit was working well before that something more like 5K or 10K would be adequate.  Alternatively a small Schottky diode with an appropriate voltage rating in series would do.

Higher Voltages cause even more power draw like 70 mA at 6 Volt. But this could be a nice flash light circuit with a single AAA battery.

There are some really good LED drivers out there that will run from a single AA cell as well.  Pair one of those with a modern 120L/W + LED and you can have a very nice torch.

Thank you for the suggestions.

I am waiting for the circuit diagram of Lasersaber's latest circuit http://www.youtube.com/watch?v=6B79UJGoNJE before dismantling the coil (in order to add copper strips), I want to do some tests too (without copper strips).

I think only the outermost copper is helpful.

I think that Lasersaber's 2N1304 Germanium NPN transistor can switch with a very low base current, but the BC547C should have similar capabilities. But I did not understand from his video how he connected everything, therefore I have to wait for his circuit diagram. I am using a different coil core than he does, but that should make it interesting. Is it the core or is it the copper strips, which allows for such a low power consumption.

I think that it is the good quality ferrite core and a good match to the transistor.

Greetings, Conrad

I don't have a model handy for the 2N1304.  I do for the BC547, but show that as a Si transistor.

[tturner]

hey guys im a newbie but want to make this devise. i just watched lasersaber latest video of when he just touches the 9v and it runs forever. i would like to replicate that but cant find schematics. also the coil is going to be trouble so any help would be great. thanks

[d3x0r]

hey guys im a newbie but want to make this devise. i just watched lasersaber latest video of when he just touches the 9v and it runs forever. i would like to replicate that but cant find schematics. also the coil is going to be trouble so any help would be great. thanks

wait a day or so; it's the weekend, and time for family...
not forever but 10 minutes; slider just demonstrated 19 minutes...
one would think that if you got LEDs paired with a solar cell that you could recoup more...

but ya; say 100 watt seconds(600 seconds, 0.16W)  from 0.45 watt seconds (100ms touch 9 v, 1A )

[conradelektro]

Based on LaserSaber's "Super Joule Ringer 3.0" http://laserhacker.com/?p=59 and his video https://www.youtube.com/watch?v=6B79UJGoNJE , I could make some progress:

See the attached circuit diagram and the photo, the circuit has a 4700 µF electrolytic capacitor.

Run times with the 4700 µF cap:

With 12 V start up the LED shines well about 1 min 30 sec and last till 2 minutes (rather dim at the end).

With 9 V start up the LED shines well about 1 minute and lasts till 1 min 30 sec (rather dim at the end).

With 1.5 V start up the LED shines dimly about 10 seconds.


Measured power consumption (with a digital multimeter, not very accurate):

0.7 mA at 12 V
0.5 mA at 9 V
0.3 mA at 2 V


I used a different core than LaserSaber, a different transistor and no copper strips on the core. If one turns around the LED it is brighter but run time goes down (power consumption goes up). The circuit does not work without the 1N4148 diode.

The circuit works by connecting the cathode of the LED to the base of the transistor (as shown in circuit diagram), but also if the LED's cathode is connected to the positive or negative rail. If the cathode of the LED is connected to the base of the transistor the on time of the transistor is the shortest, hence power consumption is lowest.

Greetings, Conrad

P.S.: Once I see LaserSaber's latest circuit I will start to modify my coil by adding copper strips.

[alejandroguille]

so?