Battery Free LED Flashlight Driver - SJR Looper V4

JimBoot · 2015.05.01, 08:18:08

I would love to get Tk,tin man slider et al thoughts on laser saber latest
http://youtu.be/1RXolkA08uk
From the description
Published on Apr 28, 2015I have made some more progress with the SJR L circuit. I have some more pot cores coming to test different sizes and materials. I think I need to go back and test using the copper foil within the coil instead of the air capacitor. Those of you who have followed this project will know what I am talking about.
Schematic coming soon at: http://laserhacker.com/
Ready made prototypes at Tesla Maker: http://teslamaker.com/

TinMan · 2015.05.01, 08:32:45

Quote from: JimBoot on 2015.05.01, 08:18:08
I would Tk,tin man slider et al thoughts on laser saber latest
http://youtu.be/1RXolkA08uk
From the description
Published on Apr 28, 2015I have made some more progress with the SJR L circuit. I have some more pot cores coming to test different sizes and materials. I think I need to go back and test using the copper foil within the coil instead of the air capacitor. Those of you who have followed this project will know what I am talking about.
Schematic coming soon at: http://laserhacker.com/
Ready made prototypes at Tesla Maker: http://teslamaker.com/

Watching the video,it looks like that is one of his shortest run times ever. Im not seeing what all the fuss is about Jim?. How is it that his little easy spin motor ran for such a long time on a much smaller cap than this circuit did. I watched cool joules video on the easy spin motor he recieved from LS,and it wouldnt run for more than 30 seconds after spin up. It seems that LS is the only one that can get these extra long run times with these circuit's,and have the LED's so bright,while the rest of us that build the exact circuit's,or recieve one of LS's pre built setups just can get those sort of run times with those sort of results? $:-\$

JimBoot · 2015.05.01, 08:50:53

Quote from: TinMan on 2015.05.01, 08:32:45
Watching the video,it looks like that is one of his shortest run times ever. Im not seeing what all the fuss is about Jim?. How is it that his little easy spin motor ran for such a long time on a much smaller cap than this circuit did. I watched cool joules video on the easy spin motor he recieved from LS,and it wouldnt run for more than 30 seconds after spin up. It seems that LS is the only one that can get these extra long run times with these circuit's,and have the LED's so bright,while the rest of us that build the exact circuit's,or recieve one of LS's pre built setups just can get those sort of run times with those sort of results? $:-\$

I've never built one so I was wondering if this was significant. Thanks for the feedback.

T-1000 · 2015.05.01, 10:32:07

It all goes back to Russian circuit replicated by Delamorto and others..
The principle is same like you see in Dally/akula/Ruslan circuits where static EMP is mixed with magnetic field on coil just in this case is on small scale.
Might be worth trying lasersaber circuit but with some modifications for proper 2 impulses mixing instead of single transistor auto generator type.

Slider2732 · 2015.05.01, 17:43:00

Am just watching it now Gromit.

I did find something very interesting, with a replication the other day of his Tesla Torch circuit.
It was fascinating to find that only 1 coil was needed, not both and then no transistor !
The modded circuit is hyper simple, for very long run times - Just connect a 1000uH axial inductor in series from 1 battery/cap connection, then through the 3 LED's to the other power connection and have a 330uF cap across power input. 6 minutes run when simply spiking quickly with a 9V battery.
Middle LED fades out, then one side and then the other carries on. Finishing measured voltage was 4.39V on the cap, plenty left.

Anyway, back to the vid

TinMan · 2015.05.02, 02:12:45

Quote from: Slider2732 on 2015.05.01, 17:43:00
Am just watching it now Gromit.

I did find something very interesting, with a replication the other day of his Tesla Torch circuit.
It was fascinating to find that only 1 coil was needed, not both and then no transistor !

Middle LED fades out, then one side and then the other carries on. Finishing measured voltage was 4.39V on the cap, plenty left.

Anyway, back to the vid

QuoteThe modded circuit is hyper simple, for very long run times - Just connect a 1000uH axial inductor in series from 1 battery/cap connection, then through the 3 LED's to the other power connection and have a 330uF cap across power input. 6 minutes run when simply spiking quickly with a 9V battery.

Which one is this Mark?

JimBoot · 2015.05.02, 13:16:15

That was a great little circuit there. Reminded of Av plug with an inductor. Very lovely work Wallace. I see we have a new princess overlord this morning too.

Slider2732 · 2015.05.02, 17:50:42

Ah forgot to post the video here.

The actual run time is about an hour, before the last LED finally goes out.
It then has 4.3V left on the 470uF cap.
As pointed out by an unhelpful/helpful commenter, it's similar to a resistor in how it runs...but it's not a resistor now is it ? lol
The measured inductor resistance is 11 ohms, thought i'd check.
The finger capacitance re-fire got me though, how it can seemingly all finish and yet will relatively strongly run again by doing that.

Chet K · 2015.05.02, 19:28:40

Mark
can you post a link to the Vid ?
I get a blank screen on your above Vid

thx

TinselKoala · 2015.05.03, 04:44:49

Quote from: TinMan on 2015.05.01, 08:32:45
Watching the video,it looks like that is one of his shortest run times ever. Im not seeing what all the fuss is about Jim?. How is it that his little easy spin motor ran for such a long time on a much smaller cap than this circuit did. I watched cool joules video on the easy spin motor he recieved from LS,and it wouldnt run for more than 30 seconds after spin up. It seems that LS is the only one that can get these extra long run times with these circuit's,and have the LED's so bright,while the rest of us that build the exact circuit's,or recieve one of LS's pre built setups just can get those sort of run times with those sort of results? $:-\$

That's interesting, I wasn't aware that others, using the same circuit, weren't duplicating his results. I should know more in a week or so.

I did however make a modification to the "sooper looper" circuit some time ago that is verrry interesting...

http://www.youtube.com/watch?v=yjgemF5zpeE

TinselKoala · 2015.05.03, 04:53:10

Quote from: Slider2732 on 2015.05.02, 17:50:42
Ah forgot to post the video here.

The actual run time is about an hour, before the last LED finally goes out.
It then has 4.3V left on the 470uF cap.
As pointed out by an unhelpful/helpful commenter, it's similar to a resistor in how it runs...but it's not a resistor now is it ? lol
The measured inductor resistance is 11 ohms, thought i'd check.
The finger capacitance re-fire got me though, how it can seemingly all finish and yet will relatively strongly run again by doing that.

Actually.... if it's not oscillating, working only on DC, then the inductor might as well be a resistor. Your scope will tell you if it's oscillating or not!
May I suggest that you make another identical circuit except instead of the inductor, you simply put 11 ohms of actual resistors in series with the LEDs and compare the performance of that circuit with the present one? You may find that they work the same. Or not.... but that's why we do actual experiments that compare conditions in order to test hypotheses.

I suppose it's possible that the nonlinear action of the LEDs near their threshold voltages could be producing an oscillation in combination with the inductor and capacitor ... a scopeshot would be nice to see.

About a year ago, I made this video: (turn down your speakers, the "noise" is kind of loud)
http://www.youtube.com/watch?v=GiLVBBdT_uk
Run time appears shorter because instead of dissipating power visibly in three LEDS, it's going into heating the resistor instead.

Slider2732 · 2015.05.03, 15:10:14

Here you go Chet:
https://www.youtube.com/watch?v=pdCGlLgNNcQ
(it's on my regular YouTube channel as 'Public').

The presumption has been no oscillations TK, unless it was refiring in some extraordinary manner from mains in the air or some such. But then, what exactly could be firing, there being no transistor. So, no delusions there, but interesting to witness something other than a resistor, labeled as such, doing resistor things for a long running time. Any coil is a resistor of course.
Resistors and LED's have had me caught out a couple of times in the past...one time where a homebuilt synthesizer was switched off at the wall and the LED trundled on brightly for an hour or similar. It was merely draining from the large smoothing caps and whatnot of the (poorly designed) power supply. But, there indeed a simple resistor was doing the eye candy. Having limited the brightness in any case, it kept the LED at the same brightness for that unexpected run.
With better design, it wouldn't have been a surprise, but the focus was on sound.

A 10ohm was tried and does display similar traits. I'd never have thought that such a low value would do that.
The scope does indeed show a flat trace.

I guess we're all used to knowing that a small cap will deplete straight away with an LED across it. So, when that doesn't happen, it's all very exciting. However, no LED is running at top whack output in any circuit seen or experimented with so far, that I can determine. Factor in more efficient running with the right frequency for the LED itself and things do get interesting for efficiency.
Tuning that frequency to a specific LED, from a specific manufacturer, is something I haven't seen done yet...anywhere.

PhysicsProf · 2015.05.05, 12:41:31

Mark:

Quote
I guess we're all used to knowing that a small cap will deplete straight away with an LED across it. So, when that doesn't happen, it's all very exciting.

I agree and appreciate your inquisitiveness, Mark!

Can you get the simple circuit to work using a toroidal inductor? There are some tests that can be done with a toroid in place, if this can work...

Slider2732 · 2015.05.06, 02:34:36

Thanks Steve, seems it was a case of error though, with not realising that 10ohms could have such an effect.
As a non oscillating system, the toroid ought to work just fine...the resistance of the wire being operative. However, that's not to say that a transistor firing on a second wind of that toroid wouldn't be beneficial. Interesting to think about.

TinselKoala · 2015.05.07, 06:02:17

I thought I'd throw together a rough approximation of a replication.

I didn't have a pot core so I wound a 200 turn outer secondary of #34 onto an existing 7.5mH Litz-wound choke from an old TV chassis. I can't find my old air variable capacitor and all my little variable trimcaps are used in other stuff, and I hate to take working stuff apart. So I experimented with some fixed caps and finally settled on a 180 pF mica cap in place of the air variable. The diode is 1n34a germanium, but I intend to try a Schottky 1n5817 or 1n5711 in there soon. I also changed the electrolytic cap to 1000uF, but 470uF works fine too, just not as long.

This thing is really weird. It has about 5 different modes of operation, depending on what you touch with fingers, how much charge is left, and other stuff. It does run with dim LEDs for long time. But the LEDs can be made to get _very_ bright (but for less time) by touching the right stuff at the right time.

Like many JT variants, it continues to oscillate even after the LEDs have gone out, and they can be turned back on by touching the circuit in certain places.

Here's a pic showing a "touch-bright Mode 3" action, and a scopeshot of typical dimmer "Mode 1" behaviour, after running for about 5 minutes on a momentary 9v chargeup. With a small supercap this thing would run brightly for many minutes.

I should have better materials to play with in a few days, hopefully, and it will be fun making comparisons, to see what matters and what doesn't matter in the construction.

TinselKoala · 2015.05.07, 06:23:50

Changed to the Schottky 1n5817 and the LEDs are a little brighter in "Mode 1" but now it won't do the superbright touch Mode 3.

Here's another shot, what I call Mode 2, which is running along with 9V power supply connected: When the PSU is disconnected, it reverts to Mode 1 as shown above. Then if I touch it in the right place it will go back to Mode 2. Sometimes it will shift into a continuous higher frequency sine wave oscillation, this is when it is about discharged to the point where the LEDs aren't glowing any more, but it is still oscillating and the LEDs can be brought back to life for a little while longer with the correct touch.

Slider2732 · 2015.05.08, 03:14:09

Very nice. Where does the diode go ? between positive and
^ scratch that, have just seen that he's posted the schematic

http://laserhacker.com/?p=491

Will build one.
In fact, I have 2x Litz TV coils (from neck board), have no 34 AWG but will try 40, 200 turns.

JimBoot · 2015.05.08, 14:34:58

Kinda sounds like an aerial of some sort to me. Reminds me of holding onto the TV aerial with one foot up in the air so we could watch the cricket when I was a kid. Have you discounted electrosmog yet? Based on your other stuff I'm assume its on your list

T-1000 · 2015.05.08, 15:31:20

One of few things to note there - if you put N-channel instead of npn transistor in http://laserhacker.com/wp-content/uploads/2015/05/SJRLV4.jpg and protect its gate from over voltage the need of touching might be not needed at all when initial voltage does not go over trigger level. After first pulse it should keep circuit oscillating and recharging cap from BEMF after each pulse.

Just I wonder how long it would run in that mode when there is no energy loss over base np junction an no galvanic current over transistor anymore...

Cheers!

TinselKoala · 2015.05.08, 20:40:51

Well, I tried it with 2sk170 mosfet and it works, in that it definitely oscillates, and at first the LEDs are much brighter than they are with bipolar transistors, but it runs down very fast.

I can also confirm that my build of the circuit does light up in the presence of strong "electrosmog", but not nearly as well as the LariMan MC34063 circuit does.

I've also tried a few other inductors, since I don't have the right pot core on hand at the moment. My LEDs don't seem as bright as LaserSaber's and I'm only getting run times of tens of minutes perhaps, certainly not hours. But like many JTs the circuit continues to oscillate after the LEDs have gone out, and they can be brought back to "light" with the proper touching, but not for very long.

It will be interesting to compare what happens when I have the right core and windings. I really need to find my air variable capacitor, I guess, or get some more little plastic trimmers.

Slider2732 · 2015.05.08, 22:52:24

With the weekend upon us TK, you could do worse than go yard sailing.
Regular AM radios seem to have either 0-270pF or 0-512pF var caps in them.
I know i've nothing like the monster that LS used either...though, you could also do the aluminium on toilet roll type of var cap as an idea.

TinMan · 2015.05.09, 09:36:19

Quote from: TinselKoala on 2015.05.08, 20:40:51

It will be interesting to compare what happens when I have the right core and windings. I really need to find my air variable capacitor, I guess, or get some more little plastic trimmers.

I have one of those air caps around some where. I will have to find it,and measure it's capacitance.

Slider2732 · 2015.05.09, 16:51:06

Perhaps a bit of deserved flattery for TK, in the way of emulation.
I really like the 'bit of wood' building approach of the circuit shown above. It could allow for easy changing out of components, with everything being top mounted.
So, on finding a a circuitboard from an old TV remote, I built something similar

MPSA18, 512pF variable cap from an AM radio, germanium diode, 1000uF electrolytic.

Shown is a coil on test. TV circuitboard Litz coil + 200 wind 30AWG over it.

In practice, the variable cap doesn't seem to do much with this coil..but i'll be testing different configurations and the effects may well be more noticeable. The output is rather low from this one in any case.

T-1000 · 2015.05.09, 17:12:16

Quote from: TinselKoala on 2015.05.08, 20:40:51
Well, I tried it with 2sk170 mosfet and it works, in that it definitely oscillates, and at first the LEDs are much brighter than they are with bipolar transistors, but it runs down very fast.

Thanks for reply, the mosfet has way much lower resistance so there is bigger power draw which drain capacitor faster and bigger BEMF spikes for feedback recovery. So it comes down down balancing circuit between power draw and recovery.

Also one thing I would like to see there like in Delamorto circuit - the foil as capacitor plate(the other one is output coil) on top of windings with another source of BEMF/electrostatic spikes and how it effects the output to LEDs. Just it gets to more complex circuit with 2 signals mixing on right time and right switching of on/off state for impulses to capacitor foil...

P.S> You can see this thing in attached prototype circuit from akula.

Cheers!

Slider2732 · 2015.05.09, 20:41:23

T-1000 I quite agree about the balance aspect.
The circuit I described above has been transformed into a super long runner....by virtue of its low output in original form
By adding just 1 wire, it now runs for over 2hrs on the 1000uF cap !

The mod, is simply to connect a wire, from the negative rail to the top L2/LED's connection.
Whereas before, it would run for about 3 minutes, before fading off, it now has that incredibly extended time.
Best part, the light output remains basically the same, from start to around 1hr 10mins. Then, it fades and will sit again for about half an hour. The circuit is still running now, though barely a with a glimmer of light from the middle LED.
Voltage on the system, after 1hr 30mins was 6.79V, dropping by 1 millivolt per minute or so with the meter attached.