Simple Inverter

[Gav]

thanks for your reply
im using a 300v 80uf cap from a camera flash circuit, it was working great with no heat at first.
i found what was doing it though, i noticed a glow on 1 of the connections, after rejoining the bad connection it worked normally again with no heat my bad.
ive made a new coil which works as good too, i used the coil from the fan motor from a microwave oven, i pulled the motor and transformer looking part off it and left the core, then wraped a bifilar over it, much easyer then winding my own

Gav

[SkyWatcher]

Hi gav, glad it's working well for you, sounds like your using the meissner circuit, instead of the joule thief with the capacitor in parallel with base resistor, both work well.
Can't seem to get this flyback core to give good light output, going to try and wrap the 1500 turn secondary directly with 24awg. bifilar and see how that works.
If that does not work well, thinking it may have too many turns on the secondary for this type of core, meaning this core may be working so well, that the voltage spikes may be too high and not giving enough current to the gutted cfl bulb to really get it cooking.
If anyone has any tips or experience on winding ratios for these flyback cores, would greatly appreciate them, otherwise, will do this the trial and error way.
These cores are to nice not to be put to some use.
peace love light
tyson

[Gav]

i followed the joule thief circuit you posted earlyer, im not good with circuits so copying others is the way to go for me, thanks for posting it
i just blew up 5 transistors trying to build another circuit before i found that i had connected it wrong time to desolder more old boards and recycle the parts

Gav

[SkyWatcher]

Hi gav, thanks for the reply.
Yes, have blown my fair share of transistors and also have a few spares with the 3 TV circuit boards here now.
Well, just tested the bifilar wound on top of the secondary using the flyback core and it is much better, though still think it has too many windings on the secondary, so will pull some layers off and try it.
Only thing really different about the secondary on this flyback compared to the bead core, is that it has less wire length, which may be a problem or may not.
It could be the higher resistance is a better match for the bulb or as said, it is just too much voltage for this bulb and not enough current.
Meaning, it is not being efficiently absorbed into the bulb and no matter what adjustments are made with base resistor and capacitor, it is not firing up the bulb sufficiently to take all that voltage and it reflects back into the circuitry, heated that up instead of the bulb.
We will see how less secondary turns works out.
Next thing to try, whether this pans out or not, will be winding up a ferrite yolk piece, just one half, so it is an open core.
peace love light
tyson

[SkyWatcher]

Hi folks, well the theory was correct.
Removed 2 layers off the secondary and it is running very efficiently right now, still more testing to be done to maximize light output for lowest input.
There is 2 layers of bifilar 24awg.-64 bifilar turns, wrapped around the secondary and the secondary now has 8 layers or around 1200 turns of 30awg. magnet wire.
Though as it stands, drawing 280 milliamps at 12 volts and is not comfortable to look at the gutted 13 watt cfl for very long, which is an indicator, that it is close to a 40 watt incandescent as far as light output.
Using the same 1200 volt, 4.7 nanofarad capacitor in parallel with a 4.7Kohm base resistor for now, will be testing different combinations for more efficient light output.
Was hoping this core would perform and it is indeed.
Though the core gap does seem to matter at least with the circuit the way it is.
peace love light
tyson

[SkyWatcher]

Hi folks, for those still interested in lighting up gutted cfls or any fluorescent bulb that has burned out filaments, all these experiments being done here, are for you and me.
Finally was able to get some good light output, though had to go back to the Alexander Meissner type circuit, it seems to be more stable and is not so picky as the joule thief and variants can be.
Another thing that needed to be done, was basically take the flyback core halves and turn them completely away from each other, so pretty much an open ended core again, just seems to work well with these collapsing field oscillators.
The 13 watt cfl in the picture is brighter than the 40 watt incandescent on the right, brighter than the ferrite beads, though using 720 milliamps at 12 volt input, it looks between a 50-60 watt incandescent in light output.
Using a 2n3055 this time and removed another secondary layer, for 7 layers or around 1050 turns, 30awg.
Slight bit of warmth in the transistor.
Will make a few more tests to see if this cfl can output what a true 60 watt incandescent does using this meissner circuit, as it is very close already.
Next experiments will be with the large yoke ferrite cores, probably use only one side, to keep the core open ended, we'll see how it works.


Uploaded with ImageShack.us


Uploaded with ImageShack.us
peace love light
tyson

[Flumen]

SkyWatcher, thanks for keeping us updated. I'm following with interest.

Flumen

[totoalas]

Xmas Charger

Xmas Charger for Samsung Ipad 2 Galaxy Note 2 Phone Light Radio 141212 - YouTube

Merry Xmas to all

totoalas

[SkyWatcher]

Hi flumen, thanks for reply, hope the experiments are of value.
Hi totoalas, thanks for the video link, which circuit is powering the things in the video, notice you said joule ringer, is that just for the light, thanks.

Tested the TV yoke ferrite core and it works well.
Similar amp draw to the last flyback core, though is a little brighter.
Using the meissner circuit again with this one as it is rock solid and the 2n3055 transistor is slightly warm.
The 13 watt gutted cfl bulb is at least 50 watt incandescent equivalent or above as far as light output.
2 layers of around 400 turns of 30awg. magnet wire, then on top of that, one layer of around 40 bifilar turns, 24awg. magnet wire.
Going to try and get this cfl to its maximum brightness, think it is suppose to be equivalent to a 60 watt incandescent, though the light output as it is, is very adequate for most uses.
Nickz, you out there, this is close to what you were talking about, powering gutted cfls without having the transistor go up in flames, hehe.
Here is a video with the TV yoke core.
Alexander Meissner self oscillator using ferrite TV yoke - YouTube
peace love light
tyson

[totoalas]

Hi flumen, thanks for reply, hope the experiments are of value.
Hi totoalas, thanks for the video link, which circuit is powering the things in the video, notice you said joule ringer, is that just for the light, thanks.

Tested the TV yoke ferrite core and it works well.
Similar amp draw to the last flyback core, though is a little brighter.
Using the meissner circuit again with this one as it is rock solid and the 2n3055 transistor is slightly warm.
The 13 watt gutted cfl bulb is at least 50 watt incandescent equivalent or above as far as light output.
2 layers of around 400 turns of 30awg. magnet wire, then on top of that, one layer of around 40 bifilar turns, 24awg. magnet wire.
Going to try and get this cfl to its maximum brightness, think it is suppose to be equivalent to a 60 watt incandescent, though the light output as it is, is very adequate for most uses.
Nickz, you out there, this is close to what you were talking about, powering gutted cfls without having the transistor go up in flames, hehe.
Here is a video with the TV yoke core.
Alexander Meissner self oscillator using ferrite TV yoke - YouTube
peace love light
tyson

Hi Skywatcher
I used your Simple Inverter + JR 1 in a box
I can switch on all except for the ipad at the same time but charging great

totoalas

[Nick_Z]

Sky and All:
Yes, I'm still here, and following what is being done. I'm still working on getting the Cfls to light full brightness, without cooking the transistors.
I'll watch your video now and see what you've done with the yoke cores.

Nick_Z

[SkyWatcher]

Hi folks, Hi totoalas ,thanks for the reply and info. glad your finding use with all these circuits.
Hi nick, modified the 4 ferrite bead transformer, it had 5 layers of 1600 turns-30awg. magnet wire, which was really to much, so took off 2 layers, so it now has 960 turns.
Then put a few less bifilar turns of 24awg. magnet wire, 66 bifilar turns, to try and get the gutted 13 watt cfl up closer to a 60 watt incandescent.
With these eyes, looking at one that is plugged into the wall and this one, it is very close, probably between 55-60 watt equivalent.
Using a TIP42, which for this setup, a 2n3055 is really what is needed or other, though it does work well and the transistor is luke warm, though the small heat sink is warm to hot, going to hook up the 2n3055 to this one and see if the gradual heat buildup remains.
To give an idea, looking at the cfl with these eyes, you can barely make out the spiral tubes because of the brightness and becomes to blinding, though will probably make another video to show how it lights up the wall and things that are lighted indirectly, gives a better comparison.
It is using the A. Meissner circuit and will post an update how this one is wired up.
This one is drawing around 820 milliamps at 12 volt input and probably getting close to the amp limits of this 24awg. wire, as it gets luke warm.
Yes, the ferrite TV yoke core halve is working well and that one only has 400 turns on the secondary, not to mention a higher resistance because of longer turn length.
Going to remove a few bifilar turns off that one, to see if the cfl will get up to a 60 watt equivalent and see how the TIP42 and 2n3055 react to it, heat wise.
peace love light
tyson

Edit: just observed the 13 watt gutted cfl start to fall off in brightness, then switched the wires around on the cfl and it came back up to blinding brightness again, this may still be too many turns and could be causing some kind of damage, though it is odd, that just by turning the connections around, it works fine again.
Edit 2: ok, switched wire connections around again and the bulb is blinding bright again, odd, maybe one of the alligator leads is going bad or a connection somewhere, as it is on a proto board for testing purposes.

[SkyWatcher]

Hi folks, just thought to add one more update for today.
Changed the TIP42 to the 2n3055 and the transistor is ambient to slightly warm and the heat sink is luke warm, so that solved the heat issue and on top of that, it is running from a 13.6 volt-10 amp power supply, for normal everyday use.
Hard to say what the amp draw is, hooked up a 1 ohm-10 watt resistor in line with negative rail and a voltmeter shows 800 millivolts across it, then the normal dmm shows 820 milliamps, though when i touch the coil or the transistor it jumps all over the place.
Will be removing another layer of secondary from this transformer as again, as the cfl heats up, after 15 minutes, the light starts to drop off and i can see an irregularity in the illumination of the spiral tube.
Again, probably too much voltage, so we will see how that works out, unless this bulb is already damaged somehow, we'll find out.
peace love light
tyson

[SkyWatcher]

Hi folks, so indeed it was too much voltage for the bulb and was apparently overheating it probably.
Now has only 2 layers for secondary, 640 turns-30awg. magnet wire.
Using 13.8 volt power supply, it is just about as bright as a 60 watt incandescent, all the light in the video, was only from the gutted 13 watt cfl.
The 2n3055 transistor is ambient to barely warm and the heat sink is cool.
Here is the circuit used and a video link.
A.Meissner Oscillator Variant - YouTube



Uploaded with ImageShack.us
Any and all comments and sharing greatly welcome.
peace love light
tyson

[SkyWatcher]

Hi folks, been testing the latest device that was previously posted, noticed the output was a bit off, then started to dim a bit more, found what the problem has been.
Do not use alligator clips to connect to gutted cfl, as they do not always make a reliable connection and may at first be ok, then may move slightly or any possible bad connection point can cause a large loss of light output.
This most likely means, for those following what is being done here, that the 3 layers of secondary was probably ok, though it seems a bit brighter with only the 2 layers, probably cause a little bit more current potential to flow through cfl, due to lower secondary resistance.
Hope this is of help.
Again all comments are welcome.
Will begin trying to optimize the ferrite TV yoke core to compare with the ferrite bead core setup.
peace love light
tyson