Akula eternal lantern 4

[d3x0r]

Revised a little; this works a little better than the modified tesla-torch schematic...


10mA at 6V; amount of load determines operating range... 20 LEDs can run up to 8V then stops at higher voltage... runs down to about 5V;  [size=78%]10LEDs runs down at 5V... fewer LEDs more current draw...

using same LEDs lasersaber does in telsa torch.
capacitor/choke are experimental values and not really picked...
also don't know if maybe using a smaller drive cap will be more effective... more voltage recovery from choke... needs to be isolated from power source too maybe
2n2222 works better than mpsa18 as transistor


Edit:
Huge brightness increase if I disconnect the choke, and leave the capacitor (which goes to +50 some volts in this case);  but pulls 250mA, and the transistor starts getting HOT.
The voltage at the collector never goes to 0 above 1.2V or so input...
but I'm using a huge ratio coil; going to swap that soon... small capacitance leads to more power draw at similar brightness... inductance of choke needs to be tuned to empty the cap at about the same rate as the duty cycle... too small of an inductance and might as well not even have the cap there... maybe I should remove that whole choke/secondary cap...

[d3x0r]

3 minutes on 68mF with more brightness    ( 42 minutes on 1F projected could probably reduce the resistor on the LEDs at that )
Filled in more values.
With more matched primary/secondary collector does get to 0V above 1.4V....
Have to swap back to my 37uH/27mH coil and try fewer LEDs and remove the extra diode...
When I switched to this 1mH/3mH coil I had to use a resistor to the base for a while... last version did not use a resistor to the base.  Have video proof but it needs to be spliced together... and still isn't a finished product so maybe I'll just keep them as historic reference.


Edit: Remeasured choke it's not 3mH; it's 116mH or 526mH (either works)... it's lower power with such a big choke; otherwise there's no momentum... that is there's no charge to the second small cap.  Could probably go lower... might relate to the inductance of the secondary coil...

[d3x0r]

I don't understand.  The current state of this circuit is boggling my mind. 


got some NTE172A; TRANSISTOR NPN SILICON DARLINGTON 40V IC=0.33A TO-92 PRE-AMP MEDIUM SPEED SWITCH; 7000 gain
( http://www.alliedelec.com/images/products/datasheets/bm/NTE_ELECTRONICS/70215725.pdf )

and BAT41; Diode; 100 mA (Continuous); 110 V (Typ.) @ 25 degC(Reverse); 1 V (Max.); 400 mW;
I got them because of 400mV forward voltage and 5ns reverse time;I guess schottky are just fast)
( http://www.alliedelec.com/images/products/datasheets/bm/VISHAY_SEMICONDUCTOR_PDD/70061664.pdf )

Using original micro-pot, 1mH:3mH coil...

http://youtu.be/q-2gZasEgm8

I dunno... looks like more wire I remove lower power it is... when I started and got it stable I had a lot of extra coils/caps connected, and it was pulling .21A... and the transistor was getting hot... got it down to .17A... and now it's .7A without all the extra stuff connected... actually by the end of the video it's at 40mA... I don't get it.

The power has to go through the 165pF cap at 0.04A but the frequency is only like 40kHz...

without cap, don't get very good voltage; the capacitance was originally chosen to be in resonsant frequency with the coil/transistor before, just ended up keeping them.
I think I should solder this.


Edit: Correction: image indicates 68mF; actually only using a 470uF on the video.
Edit2: collapsed the circuit to 1 board and shorter jumpers between parts; no real increase... the 165pF cap is actually 2 330 caps in series, and I had to add a wire that I could touch to start the high power output on the junction between the caps.  Frequency is less than 7Khz really....
Previously I was touching just an aligator clip that was between it that restarted it.
I added a metal sphere to the low side output just before the LEDs, and touching this sphere reduces power from 70mA to 40mA
(http://youtu.be/i-OyJ2A7cFc) touching the sphere also increases the brightness of the LEDs.  It's not grounding the sphere... it's already -80V (added a string of blue LEDs) when I added the red or orange LEDs the frequency goes up and consumes 210mA.


This method of driving this circuit does not recover as much energy from the coil as it could...well I guess maybe it does; forgot the diode from base to ground... was thinking that leaving that direct to ground as in the original scheme that it would be better...


Edit3: Started again this morning to test this.  It was very difficult to start this morning
Revised schematic to more accurately reflect current circuit.
Attaching scope probes (changing capacitance by even a tiny amount) makes the effect go away and/or draw much more power.  This power is through the transistor, as the transistor does start to get hot.

[d3x0r]

(updated schematic in previous post)


http://youtu.be/c3BGTQ5onow


Experimented with some different coils... higher inductance ratio = greater output for similar current input...
Very low inductance on primary results in higher current draw...


https://www.youtube.com/watch?v=lHqlfgXNYZU


Limited size of these small pot cores limits what ratios I can really get... so I'm going to wind some other cores with higher primary inductance to get back to lower current draw, and even higher secondary inductance...

[d3x0r]

Ya good thing noone picked that up; probably the wrong direction, as simple as it is...


This is lower currrent (higher voltage because the TL494 requires 6V at least)
Need to swap for a more balanced inductance coil; although I get good power... the recover side cannot use a capacitor...


; ( coil notation is wrong it's 100uH:19mH ) will swap to the specified coil...


edit: using the 1mH;3mH coil just reduces power output... turns out one of my diodes was bad, maybe a capacitor can be used with the other coil...
with a large enough inductance, it just drops more current faster as frequency increases... there's no overlap/resonsance to be used... (looks like a saw tooth always).. the smaller the capacitor, the higher it charges... *shrug*


I dunno this coil/core has no double-kickback either...
The secondary coil should be swapped from how it's used with the kacher circuit... works better if the first cycle (turn on transistor) hits the diode splitter with low voltage...