Joule Ringer!

[SkyWatcher]

Hi folks, Hi alfu, I checked the frequency rating on typical computer power supplies and it's around 25khz or so.
And it seems likely that the high frequency transformer is usually a ferrite e-core type.
Though I pulled a 2" diameter iron powder toroid from a computer power supply awhile back, not sure if this was used as an isolation transformer, because I'm not sure it is meant for 25khz.
Though I have a ferrite e-core from a different kind of power supply that is probably very similar to the computer ones and I am going to wind both to see how they work with powering fluoro type bulbs.
Also, I wonder why people seem to forget about cold cathode fluorescent lamps as they have no mercury as far as I'm aware.
peace love light

[SkyWatcher]

Hi folks, for anyone interested, I was able to salvage the ferrite e-core by placing it in boiling water and the two halves separated nicely.
Though it is smaller than the iron toroid, so will not fit as much wire, will see how it goes.
peace love light

[NickZ]

@ kcarring and All:
Can you let us know which cheap solar panel you have seen for $13 that could be used with this circuit? As that sounds like a good and affordable choice.
I have made my own toroid coil, that I had come up with some years ago.
It was used before for the TPU projects. Which as we all know never worked as planned. But, I connected it to the Lasersaber type Jt circuit, and it works great.
It is an 8 inch home made toroid coil, using 3 turns of 3/8" steel braided cable, inside a plastic toroid form, and wrapped with bifilar stereo wire all the way around it. It uses a 4 volt input from a lead acid battery, which can be connected to a small solar panel, as Kcarring is mentioning above.
It looks like a Jt- TPU. I'll upload a picture or two, soon.
This circuit is much easier to build than the elusive TPU, although it is still needing an external input. Maybe soon we can get it to work even without needing to be externally fed. As a permanent output Joule Ringer. And if not, then connecting it to several of the crystal power cells that some of us have we've been working hard on, can also provide it with the 6 to 12 volts and 200 mA, that it needs to work.
Best of luck to all builders,
NickZ

[alfu]

Did LaserSaber ever post the scheme for this?

Joule Ringer! Light keeps going without battery. - YouTube

[kcarring]

@ kcarring and All:
Can you let us know which cheap solar panel you have seen for $13 that could be used with this circuit? As that sounds like a good and affordable choice. NickZ

Hi Nick! The best priced panels come and go, but if you use the "Best match lowest price + shipping" in a search for "9v 2 watt solar" function on ebay, you'll bring up one like this one:

($6.97/watt) 1 pcs 9V 2Watt 130*130mm Solar Panel Power Cell $13.95 incl. shipping

To be more correct, I should have called it a "solar laminate", it's not a true "panel". A true 9v panel can be had here:

($11.24/watt) 9V 230MA 2W solar panels solar power panels 2watt $22.49 incl. shipping

Sun-elec.com has some pretty good deals too. Sometimes.
($2.59/watt + shp.) SUN Solar Panel 30 Watts 18.70 Vmp $77.70 + shipping

I also sell a 50 watt / 12v monocrystalline Grade A panel, warrantied, power warrantied, for $140 CDN. ($2.80/watt + shp.) Or, a 140 watt / 12v panel for $300 + shp. ($2.14/watt + shp.)

I have also bought from Thamesmall on ebay, a fair bit: Good Vendor. I like his 6v / 200 mA panels (1.2) for about $7-8 USD delivered. I find if you take a small Walmart mirror and angle it flat in front of these guys, you get about 7.2v / 240 mA. Here is a search of his stuff, he also sells a small 10v, for under $10.

Search Thamesmall for solar on ebay

You can throw together a GREAT tracker for cheap too! It's picaxe / LDR based, the best circuit floating around, I figure (msg me about it if you like, I have the schematic). Here is an awesome 4RPM motor for it, for only $13.00 shipped (The description for this motor is wrong, it's for an ammeter.. but I contacted them directly and have the correct description)! Single Axis tracking gives you 10-35% extra energy.

[alfu]

Thank you for the sources, kcarring

My focus will be to use 12V batteries as storage and source for lighting.

Reason being, after a 12V battery can't crank your car any more, it probably will last another 2 years or so as a space lighting source. The few tenths of an ohm that sulfation adds, that kills cranking capacity, is negligible in the impedances we work with. Is that worth more than the $6 you can get for it as a core exchange? You bet!

Also, solar panels matched to 12V batteries are quite standard and easy to get, and 12V is a good voltage to use for a mix of LED and CFL units.

[NickZ]

kcarring and All:
Wow, looks like you've spent some time on this venture. I will go through the list that you provided, and post a comment later. Thanks, for taking the time to post that information.
I am currently using a 4 volt battery or should I say 4 0.5 Ahr 4 volt lead acid batteries grouped together in parallel. I could place them in series and obtain the 12 volts, but as I am using 3 15 led 4 volt bulbs that are made for big rechargeable flashlights, I can connect three of those bulbs to my version of the Lasersaber circuit, and light all 45 leds or more very brightly, all night long, and maybe even longer. They can be recharged during the day by their own 120v to 12v adapter very easily.
This is working great and gives a good amount of light for emergency back up purposes, or even nightly use. But a small solar panel would be important to have in case of longer lasting black outs, which we have here all the time, or worse, like very prolonged power failures.
Once we find a source for the big ferrite toroids we can make the bigger 12 volt system, and be able to light a small house such as I mine, like Lasersaber is doing with his shop.

[SkyWatcher]

Hi folks, Hi alfu, I agree with you on that, 12 volts is more obtainable, though my car battery that went kaput only has 8 volts, tried for a few weeks to rejuvenate, though must be a couple shorted cells.
I am testing the ferrite e-core i salvaged and rewound, its 1-1/2" high X 1" wide X 1/2" depth, think it was from a dvd changer power supply.
Hard to remember, since I've been salvaging so many parts lately from pre-determined obsolescence devices, wink wink.
I used 24 gauge for bifilar primary and 30 gauge for secondary and it is lighting a 15 watt cfl quite nicely at 12 volts-180 milliamps, hard to say brightness, though when i look at it for a second, it burns a light spot in my eyes, surely half brightness or better. 2n3055 transistor slightly warm.
Using bedini type circuit without the resistor from base to emitter.
Here is a pic of the type of ferrite e-core I'm using.


Uploaded with ImageShack.us
peace love light

[totoalas]

Thanks Patrick
@ 12 v the 3 w led lamps are @ its brightest in the original 10v dc set up om measuring with the lux meter(x100 scale)
warm white smoke diffuser less than 1
cool white smoke diffuser 1
cool white clear diffuser 1.5 the brightest
warm white clear diffuser 1.3
Using lasersaber ckt less cap and resistor

cheers,

totoalas

update 11 11 11
3w 5 led 220v ac will lit 10 pcs in series
3w 20 leds 220 v ac will dimly lit 4 out of 10 pieces
usin 1 2n3055 2 diodes 10 v dc

[xee2]

Some more tests on the 120 VAC LED bulb LaserSaber is using (current reduction tests):

video >>>>> Joule thief with 120 VAC LED bulb - part 2.wmv - YouTube

[totoalas]

Some more tests on the 120 VAC LED bulb LaserSaber is using (current reduction tests):

video >>>>> Joule thief with 120 VAC LED bulb - part 2.wmv - YouTube

low amps thats great have you tried some series lighting

thanks for the info

totoalas

[alfu]

... I am testing the ferrite e-core i salvaged and rewound

I'm benchless for now (typing this when I should be constructing a HVAC return manifold) so I will have to just bug y'all with thought experiments rather than give you real findings.

To wit, it seems to me (in my ignorance) if we are trying to run CFLs (hundred volts and above) from batteries (ones to tens of volts), a switcher PSU transformer which runs a transistorized device (ones to tens of volts) from mains (hundred volts and above) could be used backwards and have something at least light up enuf to take data.

But it seems like everyone here is salvaging the cores and rewinding the coils -- I will have to go back thru the forum posts and find out why.

Boosting voltage from a battery to run a 120V LED lamp, which just has to knock it back down to 2 volts or so internally, doesn't make a lot of sense to me. CFL elements are damned cheap (I have a pile of dead CFLs and I bet you do too), the phosphor area of ONE of them is probably equivalent to 1,000 white LEDs, and various of you have demonstrated dimmability and close to half-brightness with 2W of drive.

Commercially available CFL bulbs use way more power than they should, because just as a car's engine must be sized for its maximum load, so too a CFL's ballast 'must' be sized for creating the starting voltage. To save costs, the ballast's components are non-adaptive, and dissipate this extra energy once the lamp is lit.

I know deriving luminosity data is an inexact science, but I would suggest that any posting of results include some lumens/watt figure. For example, I have an old germanium solar cell-based photographic light meter that I will be using, placed at some reference distance from the bulb.

[SkyWatcher]

Hi folks, Hi alfu, you make very good points I think and lol, my bench is my computer desk for now.
I rewound my core because it barely had any copper on it, probably it was meant more for multiple various voltage outputs, so had to be rewound with more copper.
Also, gas tubes without mercury could be used.
I tested my e-core and circuit with 12 volts to compare to the 16 leds I have in the bathroom and the warm white cfl put out almost twice as much light for roughly the same input watts.
I am rewinding my e-core because the bifilar primary needs more wire and resistance for the 12 volt input, it was .4 ohms per strand of 24 gauge wire.
It needs to be higher than that because the transistor probably can't handle the peak amp pulses and heats the transistor more than need be.
Here is a pic of the ferrite e-core and bulb I've been testing.


Uploaded with ImageShack.us
peace love light

[Lidmotor]

This works

Some more tests on the 120 VAC LED bulb LaserSaber is using (current reduction tests):

video >>>>> Joule thief with 120 VAC LED bulb - part 2.wmv - YouTube

@Xee2
I replicated your circuit using the 12 volt transformer today. I could not find the right LED bulb but I did find a similar bulb by Lights of America that worked. The circuit worked exactly like you showed and is very impressive as far as lumens of light vs power input. We are not going to get the results that Lasersaber is getting without that big Metglas toroid but your solution does work and as easy to make.

Thanks for sharing this with us and for the two excellent "How To" videos.

Lidmotor

[SkyWatcher]

Hi folks, hi lid, could you post any numbers on that replication.
I rewound my ferrite e-core and instead used what seems like 28 gauge salvaged wire for the bifilar, which came out to 1.4 ohms per strand.
Using the same circuit except for the diode as lasersaber's.
It is performing very well using a smaller 15 watt cfl at 12.5 volts-160 milliamp or 2 watt input, looks like 60% of full brightness and the 2n3055 transistor is cold without any heat sink.
The small capacitor in parallel with base resistor boosts the light output.
So it definitely is important to keep the resistance high enough in the primary so that the peak pulse amps of the transistor is not exceeded or it will heat with no benefit added to output.
peace love light