Lasersaber strikes again. A joule thief king ?

[conradelektro]

Using the prior (simulation) circuit and substituting a 50 to 350 pF variable mica capacitor for the fixed 200 pf, I can tune the drive capacitance to look just like the LS circuit on the bench with a real circuit. I wound the pot core  using 15 mH each, winding on a split bobbin to minimize inter winding capacitance. The circuit then draws about 20 uA and delivers about 15 to the LED.

The aim was to prove that by  minimizing inter winding capacitance, I could simulate it external to the pot core winding capacitance and get the same waveform. So nothing special about the LS circuit, it is a current starved blocking oscillator.

TK: DALM, but what is the LM? Lamp for Miners?

@Vortex1: I looked a bit more at your very nice circuit which you published here  http://www.overunity.com/14591/lasersaber-strikes-again-a-joule-thief-king/msg403216/#msg403216

I have all the components (even the variable cap) and could wind the two 50 mH windings on an other core I have:

http://at.farnell.com/ferroxcube/etd49-25-16-3c90/ferrite-core-half-etd49-3c90/dp/3056417?Ntt=3056417
FERROXCUBE - ETD49/25/16-3C90 - FERRITE CORE, HALF, ETD49, 3C90

Have to do it soon, everything is really clear and well specified, a joy to replicate. I wish everybody would publish his work in such a way.

The pot cores are nice (because one gets high inductance with a small size, but I do not think that it is indispensable (if size is not an issue).


I wonder how well "wound components (coils)" can be simulated with TSpice? I would like to be able to get a coil right at the first try?

You say "current starved blocking oscillator"? I guess you do not agree with my strange theory about "negative current feed back from the base to the emitter or collector"? The strange wave forms stem from the weak current at the base? Which leads to the extremely short pulse? So, if one feeds more current to the base, power draw of the circuit goes up, because the transistor will switch nicer ore more in an ordinary way (longer pulse)?

Greetings, Conrad

[Vortex1]

Greetings Conrad

You can get free LTSpice here: http://www.linear.com/designtools/software/

It is so easy to use, the learning curve is not so steep. I had help from Poynt99 with the inductor and transformer stuff and a few other helpful hints.

I will simulate exactly the LS circuit you describe ( I have already done this but will go back to it).

What does not show up on the LS circuit is inter winding capacitance, and resistance leakage which gets the oscillator started (LS sometimes uses his thumb). To fully simulate a circuit you need to add things which don't often appear on the elementary circuit, but this is not hard.

You can find more of my work regarding this topic posted here as user "ION": http://www.overunityresearch.com/index.php?topic=2436.msg38773;topicseen#msg38773
and on back pages of this and other threads.

Besides simulating, I like to check the results in the real world on the bench, to be sure. The real learning is in circuit analysis and proper analysis of scope shots.

I guess you do not agree with my strange theory about "negative current feed back from the base to the emitter or collector"?

I do not disagree with it, I will look into it deeper, you may be right.
Short on time right now, I will get back to any questions I have missed.

Kind Regards
Vortex1 / ION

[conradelektro]

I could replicate the circuit described by Vortex1 in his post:

http://www.overunity.com/14591/lasersaber-strikes-again-a-joule-thief-king/msg403216/#msg403216

My coil is a bit different and has only 36 mH instead of his coil with 50 mH. (This is the self made decoupling transformer for my Function Generator, the 50 Ohm resistance at one winding was not used in this experiment.) The coil is bifilar wound (also a difference to his coil). It needs 12 Volt because the coil has less inductance. The strange waveforms appear at all Voltages but the base does not go as far negative (as one wants to see) at lower supply Voltages. The ration between primary and secondary winding should be higher, e.g. 1:4, not 1:1.

And I could show the difference between "normal switching" (only 10 pF cap) and "current starved switching" (about 900 pF). See the attached photos.

The same strange waveforms (trace over the base and trace over transistor) as in LaserSaber's very low power circuit appear. Note the same very narrow pulse followed by ringing.

I seems that Vortex1 has explained what is happening and it can be demonstrated with his circuit by "current starving the base" by help of a variable capacitor.

My very old variable capacitor (from a 1960ies radio) has about 10 pF to 1100 pF (in fact it has three sections of 3 pF to 360 pF in parallel).

I am impressed by the fact that the strange wave form can be produced with a circuit which is different to LaserSaber's and more easy to replicate. Note that I used the 2N2222 transistor and a very conventional core.

Also note, that the circuit has a very low power draw (and rather dim LEDs) if the strange wave forms are present, it is only around 41 µA at 12 Volt.

In "normal switching mode" the power draw is about 220 µA and the LEDs are of course much brighter.

I think we have made progress thanks to Vortex1, no unusual coil and no unusual transistor is needed to reproduce a circuit doing in essence what LaserSabers circuit is doing. And the variable capacitor is a nice tuning possibility (much more convenient than playing with coil windings and cores).

Greetings, Conrad

[d3x0r]

Nice

[TinselKoala]

Yep, I can demonstrate the same kind of waveform shift and different LED brightnesses and current draws on several of my JTs too. I dug out the DALM (the one with the round PCB and the 4 blue LEDS in series) and with MPSA18 it shifts waveforms if the supply voltage is too great, gets quite a bit dimmer and the transistor heats up. It stays in this mode until the voltage drops below about 1.6V then it may, or may not, shift into the brighter mode. I think mine actually draws less current in the bright mode than in the "overload" dim mode though.