Joule Ringer!

[br0ns0n77]

Hey guys, still using the 2n3055 with my SJR 3.0 setup with CFLs and it's still getting very hot - so much that I burnt myself testing it. Unless there's some way to keep that thing cool I don't see how this particular transistor can hold up to the load needed to light several CFLs. Outside of putting the transistor in a heat sink or cooling it with a fan - as cheap CFLs are today, just don't think they are a viable option for widespread lighting.

Any ideas? Would placing a load somewhere else in the circuit help cooling? If so, where? Thanks for the help!

[Zardox]

I still do not have a transformer that will run a cfl. I can get a 40 watt incandecent to light very very dim. However my 2n3055 does not get hot. I do not see why anyone would not have the transistor mounted on a heat sink just for general perposes.

[Nick_Z]

By adding a magnet to the rigger's ferrite yoke I was able to alter the frequency high enough to practically not be able to hear it, only a slight very high pitched whine.
I was also able to light a 50 watt incandescent bulb along with a 65 watt Cfl. Although the Cfl was lighting dimly when used along with the 50 watt incandescent, which was about 1/2 brightness. But, that's a good level of light for late night kicking back while on the pc, or watching tv, etz... I love that flame colored golden light from the incandescent.
I do use a heat sink, even though I don't need it when using the Cfls. With the incandescent the transistor does get warm to hot, but not burning hot.
I understand that a 2n3055 can boil water if mounted on a proper heatsink, but won't last long by itself. I'm having a couple of those 2n3055 brought to me from the U.S. So, I'll have a more accurate replication when they get here in a week or so.

[br0ns0n77]

Nick, it sounds like your transistor is better suited for this setup being that it's not getting hot at all. That's pretty impressive on a 65w CFL. I guess technically we could build multiple circuits to run on a couple of big batteries to keep the temp down on those things.

What is the transistor model you are using?

[Nick_Z]

I have been using the TIP 3055 from Radio Shack, which is actually a MJE 3055T. Adding a diode between the base and emitter does increase the output on some bulbs.
If you haven't tried the gutted Cfls, I'd give those a try, also. It makes a big difference the bulb that you use. Florescent tubes are also worth trying, as they work differently than the Cfls.
You can also mount the transistor on a pc heat sink, the ones used on on the motherboard processors, that uses a 12v fan on top. That should take care of the heat issue.
My system will not light several Cfls BRIGHTLY. It can light several but dimly. I'm still working on that aspect, looks like more input is needed.
By adding a second magnet on the yoke I was able to TOTALLY eliminate the ringing noise. Very good news for me...

You'll notice on Lasersaber's newest solar ringer unit that the transistor is mounted on the outside of the aluminum panel's frame, for heat dissipation. But, the whole panel is going to get very hot once placed in the sun, so I don't think much cooling is going to happen like that, nor will it be weather proof, or able to withstand freezing temps. All components should be placed in a separate vented sound proof box, so that the ringing won't be heard, but the transistor and other components should still be able to vent but not be exposed to the elements.

[SkyWatcher]

Hi folks, Hi br0ns0n77, decided to take the risk again with trying this circuit with the 2n3055.
Tried the different joule ringer trigger coil locations.
It seems that connecting the trigger wire with the positive input lead, results in the lowest heating of the transistor and yet it lights the non-modified 13 watt cfl to full brightness, probably brighter.
Not using heat sink, so can see what trigger wire location runs the transistor best, though the heating may be caused by too low a resistance in primary winding.
This one is 25 turns-24awg. primary (actually wired it bifilar, 2 layers for 50 turns per strand, only using the one strand of 50 turns) and the secondary (base trigger winding) has two layers for 340 turns-30awg. wire.
Multimeter shows 1.5 amps at positive lead location, though if the transistor casing is touched, it drops to around 800 miliamps, probably because this circuit acts like the exciter and messes with the meter.
Think this circuit will be very efficient, if this used a larger core and ferrite would probably be best, like nicks setup, this one is an iron powdered toroid if not mistaken.
Here is a pic of it running and transistor slowly getting warm.


Uploaded with ImageShack.us
peace love light
tyson

[Nick_Z]

I ran my yoke core ringer for 6 hours last night and at the end the transistor was not even luke warm, and using two snake egg magnetite magnets on the center bottom of the secondary windings the frequency was so high that there was NO ringing that I can detect. 1n4148 diode was used between the base and emitter.
The 15 volt, one amp wall adapter was hot though, and the 65 watt Cfl bulb was warm. It lights a 50 watt incandescent bulb 1/2 brightness, also.
Most any computer repair, or Tv repair place will have some kind of ferrite core that can be used, for cheap, or free.

[nicknewbie]

Ive developed a combo of parts that make it really easy to make a joule ringer for people who have never done it before. Ill be uploading pics tomorrow but here is the part list.

One "dead" battery
6 Led flashlight head
2 3/4 by 1/2 inch toriods wound bifilar 5 wraps
Tip142 transistor

thats it. The circuit draws around 44ma @ .95 volts

Ill upload a diagram in a few hours.

heres some pics to show that the led - can be hooked to any points in the circuit. The middle connection between the coils is grey to grey and orange to orange. grey on the second coil goes to middle leg of the tip142 orange to the left and ground on the right

[Nick_Z]

Nice replications guys...

Here are a couple of pics of my different types of light bulbs, as a comparisons. First picture is of both the incandescent 50 watt bulb, and the 65 Cfls, both running together, and their different light qualities.
A nice mix of lights between those two types of bulbs can be obtained by careful mixing. Last pic is of the small led bulb running wireless on an small inductor placed next to the coil.

I'll show some other bulbs in my next pictures.

No heat on the transistor running the 50 watt incandescent bulb, and the Cfl both together. No ringing, either.
Nick_Z

[br0ns0n77]

Guys, I finally just took the thing apart and wound it exactly by the specs and it's amazing what happens when you follow directions precisely huh? I didn't have enough turns of magnet wire because I was out of it - was waiting on more to arrive and was impatient. The turn ratio apparently is very important.

The circuit now pulls around .8 amps with a 4w CFL. I had 3-4 18w CFLs plugged in and it was pulling nearly 3-4 amps... not sure that's the load I want to be running haha.

Anyway, I'm glad to report that using a 120v 7.5w LED bulb, it's drawing next to nothing amp wise and the transistor is cool as a cucumber. I'd imagine I could light 30 or so of those LED bulbs to get around 1 amp draw.

Here's a video showing it working:

Super Joule Ringer 3.0, Lead Alum Car Battery Powering CFLs - YouTube

[SkyWatcher]

Hi folks, thanks for sharing nickz and bronson77.
Turns is definitely important, because hooked up a regular transformer yesterday after testing that toroid successfully and it fried another transistor.
The primary was the about same turns and ohms, the only difference was the secondary (trigger) winding, it was around 3 times the ohms, not sure what wire gauge is on it.
Something about the impedance match between the trigger winding and load to be driven may be responsible for the transistor graveyard effect, as nickz also experienced.
Since have no desire to risk anymore 2n3055's, going to try and see if any of these many salvaged mosfets collected, will work on this circuit.
peace love light
tyson

[Nick_Z]

After a several hour run last night I seems to have fried my 15 volt, one amp wall adapter, that was getting pretty hot on this set up.
Now my other adapters with less output levels just don't run this system as well, showing less light output from the circuit.
I think that for this Ringer circuit to really "shine", it's going to take some amps, or we'll be restricted to a single bulb, or two, at most. As you can't light your house on miliamps
I also think that each bulb may need it's own transistor, and any other components as well. Too many bulbs on the same transistor will cause too much draw, heating and other problems. Especially with anything other than led bulbs of lower output values.

[SkyWatcher]

Hi folks, Hi nick, thanks for that information.
Does anyone have any more details about this video slayer posted.
He said he used 9v-110v transformer.
Found one laying around, though the primary from the wall lead is 450 ohms and secondary step down is around 6 ohms.
Using a tip32c, it can light the 13 watt cfl and 3 watt led bulb, only problem is, the frequency is too low, so the cfl just blinks, the led shuts down because of micro-controller, though a 4 watt incandescent glows fine and led christmas string kind of lights 3/4 of leds and flickers a bit.
Slayer said he gutted the 1 or 3 watt led bulb he was using, anyone know if those leds from the bulb he gutted are wired for 120 volts, or are they wired for lower voltage, like 9-12 volts, thanks.
Lasersaber Type Exciter - YouTube
peace love light
tyson

[Nick_Z]

He mentioned that the bulb was gutted, and it was an Ac bulb, 2 or 2.5 watt.
The amazing thing is the 28 mA draw, looks pretty bright for only 28mAs. He uses the 1meg resistor, as if it were an Exciter, along with the 1n4148 diode. I believe that to be the reason for the very low draw. 10 of those bulbs would only draw 280mA, or less.

I just filled the second half of my crt yoke with secondary windings, but the frequency is too high and it won't light any bulbs, except the neon. The neon is so bright it looks like it's going to explode. And the core rings again, even with the two magnets. Oh well...

[SkyWatcher]

Hi nick, thanks for replying to my post, too bad the extra windings didn't work out.
So yes, slayers bulb was gutted, the question is and makes sense to me, do those bulbs have small step down transformers inside normally to feed lower voltage to the leds, which may be in series/parallel arrangement.
Since experience shows, that leds perform and last longer when powered by lower voltage series/parallel wiring.
Guess is, that since it looked like 3 sets of 3 leds or 9 total leds, so each 3 led cluster is probably in series and then in parallel with each other.
Think slayer/gbluer was using the transformer in a step down fashion, meaning the 9 volt winding of the transformer might have been placed as the trigger base winding, maybe.
Your thoughts appreciated, since his results seem worth trying to replicate, maybe someone can contact slayer somehow and he can shed more light on this, thanks.
peace love light
tyson