Lasersaber strikes again. A joule thief king ?

[Pirate88179]

The TS3002 is difficult to get and very difficult to handle (so small). If one reads the data sheet it becomes clear that in any useful application its power draw is a few µA (4 µA, not only 1 µA).

One would need a breakout board like this: http://uk.rs-online.com/web/p/clock-timer-development-kits/7877201/ (very expensive)

I think that the TLV2401 (used as an astable multivibrator, http://at.mouser.com/Search/Refine.aspx?Keyword=TLV2401IPE4) is best, also because it is available in a practical TIP T8 package.

With the TS3002 or the TLV2401 one needs an additional transistor to drive a coil (back EMF drives LED). With the TS3002 one could get away with only one battery, which would be nice. May be there will be a cheaper TS3002 breakout board available or a 1 Volt OpAmp.

It is not easy to work at 1 µA and 1 Volt. That is the reason why LaserSaber uses the 9 V power source (9 times more Wattage, 9 V and 1 µA --> 9 µW, 1 V and 1 µA --> 1 µW).

If one uses any integrated component, it is hard to beat what I did at (with very careful design one could maybe get at half the power draw, 15µW to 20µW)
http://www.overunity.com/14591/lasesaber-strikes-again-a-joule-thief-king/msg402574/#msg402574
http://www.overunity.com/14591/lasesaber-strikes-again-a-joule-thief-king/msg402695/#msg402695

Greetings, Conrad

P.S.: I do not want to diminish LaserSabers great work, I juts look at the numbers. A big advantage of his approach are the few components, the drawback is the special coil and the difficult transistor choice. But at 1µA any component is a difficult choice, even today.

Well, I got lucky and listened to Gadgetmall when he told us about this chip and the demo board on the other JT topic.  I asked for, and received the demo board and about 6 chips to play with...for free!  I actually had forgotten that I had it and dug it out the other day.  I just found the pdf online which explains how it works so, when I get a chance, I will hook it up and see what happens.  You are right, those chips are very, very small.  I am glad one was already installed on the demo board as I am not sure that I could solder one on there correctly.

Bill

[conradelektro]

Strange switching in LaserSaber's low power circuit with a 2N1304:

(This is not criticism, just some measurements.)

I got some germanium 2N1304 transistors and replicated LaserSabers circuit, only the coil and the two red LEDs are different (see the attached circuit diagram).

The circuit makes the LEDs glow for about 6 minutes with a 1000 µF electrolytic capacitor started at 9 Volt. It works better with two LEDs and the LEDs have to point in the direction given by LaserSaber.

Please see the scope shots for the strange switching going on. The base goes below the emitter (never above) and causes nevertheless a very short pulse.

For comparison I put a diode (1N4148) between base and negative rail to see "normal" switching, which makes the LEDs shine brighter and causes much more power drain because the pulse is much longer.

I think this clarifies why it does only work with certain transistors like the 2N1304 (and may be the MPSA18, not yet here) which also switch very shortly when the base goes below the emitter.

Greetings, Conrad

[conradelektro]

Did some more tests. LaserSaber added three feed back diodes to his circuit, and they seem to help a bit. I could extend the run time of my replication by about 20%.

My electrolytic capacitors are very leaky, where do I get ceramic caps with 1000µF or more?

Greetings, Conrad

P.S.: Just did a 30 minute run from a 4700 µF cap starting with 9 Volt.

7 minutes from 1000 µF should be about 33 minutes from 4700 µF, the 3 minutes loss were probably leakage. So, it is no fluke, the thing runs consistently (although the switching is peculiar). Will work on the coil in the next days.

[hotrod34]

Hi all,

Here is a question from a newbie. Have anyone tried some of Linear Technologys energy harvesting step-up converters
for example, LTC3108 ? It operates from Inputs of 20mV. Maybe we can combine it with a Joule thief and make something useful.

Regards,
hotrod34

[Paul-R]

This is an inspiring and impressive project but how can we set out to scale the business up so that there are inputs and outputs that are more readily measured?