Simple Inverter

[Nick_Z]

What we may not have figured out yet is, are the gutted or un-gutted Cfls the better way to go??? Both work, but which works best for our application? As not everyone can afford the 7.5 watt or higher wattage led bulbs.
What appears to be the case so far is that the 12 volt led bulbs are the most efficient, over all other bulbs, more so than the 110/220 Cfls, or incandescent, or even the higher voltage leds bulbs. As there is no need to jack the voltage way up, just to bring it back down, thus waisting energy. The more components a circuit has, the less efficient it will be, as each resistor, diode, or even transistors will produce a loss. But, when I mention transistor heat, I'm using 25 to 65 watt Cfl bulbs, and 50 watt incandescent bulbs, those do get a bit hot. Not the 13 watt Cfl, or small 7 watt incandescent bulbs, or any led bulb, as they all stay cold.

All these inverter circuits are working in a very similar way. They are ALL just simple pulsing oscillators, including the open ended Exciter circuits. So I think that the most important aspect is the permeability of the cores, and the proper turns ratio. In the link below is a list of the different perm materials and frequencies that ferrite cores work at. In order to maximize our output, we need to consider these things, also.
There are ferrite cores with much higher perm, not just a few hundred, or a couple thousand U. As well as the air cores, which may be best used for the Cfls. Lynx Steam proved that even the air cores can work very well, and can produce even higher watt/lumen output than the best commercial transformer/driver made.
You can see in Mopozco's 6 Cfl video that he actually went too far on the secondary turns ratio of his hand wound fly-back transformer and his output can blow the lights out, when using 12v. So, it's important to give them the right and proper voltage and current, and not too much voltage and too little current. But, that all depends on the input source. What works for an AA is not the same as when using a 12v car battery, or a 24v, or 36v source.
My goal and intention is to light my house, all the time, day or night, and to do this on practically nothing, no budget.
Not an easy task... but, I'm getting there. Thanks for your help.

Ferrite cores materials and permeability:
Fair-Rite Materials

[hello_all]

"My goal and intention is to light my house, all the time, day or night",

buy some deep cycle batteries and solar panel with inverter and charge controller and be happy don't have to worry about the heating and checking amp and volt draw each moment. LaserSaber setup is the same , he just use some led bulbs to experiment with the power transistor and coils... In the end he always relies on the solar panels..for power...

[boguslaw]

What we may not have figured out yet is, are the gutted or un-gutted Cfls the better way to go??? Both work, but which works best for our application? As not everyone can afford the 7.5 watt or higher wattage led bulbs.
What appears to be the case so far is that the 12 volt led bulbs are the most efficient, over all other bulbs, more so than the 110/220 Cfls, or incandescent, or even the higher voltage leds bulbs. As there is no need to jack the voltage way up, just to bring it back down, thus waisting energy. The more components a circuit has, the less efficient it will be, as each resistor, diode, or even transistors will produce a loss. But, when I mention transistor heat, I'm using 25 to 65 watt Cfl bulbs, and 50 watt incandescent bulbs, those do get a bit hot. Not the 13 watt Cfl, or small 7 watt incandescent bulbs, or any led bulb, as they all stay cold.

All these inverter circuits are working in a very similar way. They are ALL just simple pulsing oscillators, including the open ended Exciter circuits. So I think that the most important aspect is the permeability of the cores, and the proper turns ratio. In the link below is a list of the different perm materials and frequencies that ferrite cores work at. In order to maximize our output, we need to consider these things, also.
There are ferrite cores with much higher perm, not just a few hundred, or a couple thousand U. As well as the air cores, which may be best used for the Cfls. Lynx Steam proved that even the air cores can work very well, and can produce even higher watt/lumen output than the best commercial transformer/driver made.
You can see in Mopozco's 6 Cfl video that he actually went too far on the secondary turns ratio of his hand wound fly-back transformer and his output can blow the lights out, when using 12v. So, it's important to give them the right and proper voltage and current, and not too much voltage and too little current. But, that all depends on the input source. What works for an AA is not the same as when using a 12v car battery, or a 24v, or 36v source.
My goal and intention is to light my house, all the time, day or night, and to do this on practically nothing, no budget.
Not an easy task... but, I'm getting there. Thanks for your help.

Ferrite cores materials and permeability:
Fair-Rite Materials

Stupid question but please give me advice : can I use old speaker ring magnet as a ferrite toroid core ? Maybe there is a way to heat it to convert into soft ferrite core ?

Why am I asking ? Simply, if you want to experiment with any power above 100W you need a good core and size is the price also. I would like to test big JouleThief with a ring ferrite core but I can buy speaker magnet 10 times cheaper then any good sized ferrite core.

[SkyWatcher]

Hi nick, thanks for the reply.
Indeed, what you say makes much sense and we all probably would like the same goal as you nick, at a minimum, to light our houses on very little juice.
Know regular warm white christmas leds work great, as have been running 16 in series parallel off a wall wart for over half a year now and no problems with wall AC spikes trying to kill them, still the same nice brightness, have had to turn on the incandescent bulbs only a few times for a couple minutes.
Though since many, many people probably have a bunch of these cfl's, sure some of those people would like to keep using those bulbs forever, even when the cfl element inside bulb blows.

Hi boguslaw, thanks for reply, now that is a great question, have a bunch of fairly large ferrite doughnut magnets laying around here.
Maybe if we put them in the oven at high temp, not sure if that is high enough temp to demagnetize, will try and research your question on net.

Well, making some experiments here trying to light these gutted cfl's.
It seems the simple inverter just cannot get up in frequency high enough or it may be because it is not allowing for the sharp flyback voltage effect that typical blocking oscillators can provide, like the joule thief.
Soo, wound 24 bifilar turns of 20 gauge speaker wire to replace the 24awg. that was on there in video.
At 12 volt input, it fired up the gutted 13 watt cfl right away, would say 85% full brightness, though the 2n3055 started to get hot, probably because the primary is now very low ohms, though the frequency is not audible.
Using 1kohm base resistor.
Though for sure, this circuit needs a good ferrite core, maybe next test will try the beads or a medium size ferrite e-core.
Edit: realized the parallel transistors joule thief configuration is still getting hot, so back to testing.
peace love light
tyson

[Matthew Jones]

You can use these type of multipliers...
Voltage doubler
with square wave stuff. But then you have to modulate it again if thats what you need.

Matt

[Nick_Z]

@ Hello All:
Yes, I know that the solar panel, charge controller, inverter route, is the normal way to go. But, if you add it all up, it would come to more than I can presently afford. In anycase, I'm going that same direction, sooner or later. But, I'm presently looking for cheaper alternative, like free alternatives.

Boguslaw: My advice is to forget the speaker magnet core idea. That has been tried before and did not work well, and is not a good alternative.
I have tried to show two free alternative for ferrite cores that work real well. The CRT ferrite yokes, and the ferrite beads, which are both free, at least they were for me. It's better to make an air core, than to use old magnets for the core. I have tried it, and I don't recommend it.
The idea is to use the BEST ferrite core materials, made for the RIGHT frequency bandwidth. As ferrite is really where it's at.

@Sky:
I tried the two transistor set up, and found that there was lower light output. Yes, they share the load, and the heat build up, but, they still heat up, and at the expense of the light output.

Ufopolitics was using a circuit with 4 mosfets to light big huge Cfl bulbs. I may try that one also, sometime. Maybe not all those fets are really needed, possibly just 2, or so. His bulbs were something like 2000 plus lumen output type.

I like the simple one component circuit, if I can get it to work for me. As I really feel that the trick is in the turns ratio. Maybe further development is needed to run with it, from where Lasersaber left off.

I'm now working on getting several of the gutted Cfls to light brightly, as that has always been the hardest test of all. They do need higher voltage, and much higher frequency to light, compared to the ungutted Cfls. As the ungutted Cfls have a E-core transformer inside of them, which is helping to raise the voltage, along with a pulse circuit.
What I've done is to combine both the ferrite rod's secondary coil to the yoke core's secondary, for increased voltage output. That will light the gutted Cfl. Then I return the end of the Cfls to the battery positive rail, instead of the transistor C or E. An additional E-core added to your secondary coils output, may also work.

[totoalas]

I ve tried microwave magnet with sjr with no success even with slayers for 2 yrs now
Then I got the idea same as you guys to put the ring magnet in a stove for 30 minutes to one hour and see if any change in the magnetic properties
will try this tonight

for the SJR / SOLin1 heres my wish list
Im planning the 4 ckt
1 by skywatcher - simple inverter
2- Slayer using the solar alum case for wireless lighting / battery av plug charging to alkaline batteris 9 v dc
3- JR 1 by xee2 for lighting @ 200 mA
Batteries - 100 pieces 9 v dc batteries

two solar panels as 10 watts 36 v dc series output to a Radiant oscillator
for charging two 12v 7 ah batteries
im looking for cheap led flat panel if its cheaper to put on the sides as cover

Hope SANTA is listening lol

[SkyWatcher]

Hi folks, Hi matt, thanks for the link, will have to think how to apply that.

Hi nick ,thanks for the reply, again think the dual transistor setup needs the right winding ratios and core to perform well, the one here with the iron toroid and 13 watt ungutted cfl is running very well, cool to touch transistors.
Like simple better also, whatever works for us makes most sense.

Hi totoalas, read briefly online, something about the ferrite permanent magnet material being a hard form, whereas the ferrite used for cores is a soft form.
Though if it can be completely demagnetized over the stove, that would be interesting to see if can be used as a core.

So, was making many more tests with blocking oscillators and the flip flop type circuit.
The best circuit so far is the bedini solid state oscillator, it lights the 13 watt gutted cfl the brightest for the least heat in transistor, though still gets hot if given enough on time.
Though this heat may simply be the fact that the primary winding resistance causes a peak pulse current that well exceeds the 2n3055 threshold and slowly heats it up.
Or another reason may be that these gutted cfl's seem hard to drive, meaning higher impedance if not fully lighted and pulses end up going back into the transistor, heating up, whereas, with the typical flyback diode charging a battery, all the juice is soaked up in the batteries lower impedance.
The ferrite bead core performs better than the iron toroid, was able to light the gutted cfl with the flip flop dimly, only problem is most of the pulses are back feeding into the transistors and getting hot and also gave me a little zap.
So conclusion so far, gutted cfl's are not a piece of cake to drive to good brightness and keep the transistor cooler.
Though will be making more tests with the bedini oscillator for now.
It's the one with the 18kohm resistor from collector to base and 10kohm from
base to emitter and trigger base resistor.
Used 1kohm for base trigger, 2kohm for collector/base and 10kohm for base/emitter resistor.
Draws around 600 miiliamps at 12 volt input, maybe 70% of full brightness.
Not sure if this circuit will work on cfl bulbs that have blown coil element inside, though probably.
peace love light
tyson

[Nick_Z]

To gut, or not to gut... that is the question.
It may be that gutted Cfls are better when used for higher voltage Exciter type circuits, that have an open circuit. And that the ungutted Cfls are better for inverter circuits that have closed circuits. There may still be a chance of additional extra energy being drawn into the circuit, due to the BEMF, in both cases. Although that may be easier said than done, and even harder to prove. The self runner may be the ultimate proof, IF its even possible.
In the meantime, we have solar panels, wind, and hydro, to call on, although none of them are very good options, in my case.

I'm devoting 100% of my time and effort to this end. As I feel that it's this focused effort that is what really counts ultimately, and not just the final result.

[totoalas]

To gut, or not to gut... that is the question.
It may be that gutted Cfls are better when used for higher voltage Exciter type circuits, that have an open circuit. And that the ungutted Cfls are better for inverter circuits that have closed circuits. There may still be a chance of additional extra energy being drawn into the circuit, due to the BEMF, in both cases. Although that may be easier said than done, and even harder to prove. The self runner may be the ultimate proof, IF its even possible.
In the meantime, we have solar panels, wind, and hydro, to call on, although none of them are very good options, in my case.

I'm devoting 100% of my time and effort to this end. As I feel that it's this focused effort that is what really counts ultimately, and not just the final result.

hi nickz
.neon lamps used thin wires on the glass for hv ignition of gas maybe you can try as well on the cfl just like in Slayer my pulse motor is consuming all my extra time for the past 2 months lol

[hello_all]

As per my tests about 2n3055 , it will remain cold till 10watts , and then it is all upto heatsink upto 50 watts , u may try some water cooling on heat sink with drops of water , after 50 watts it is sure to blow, why because the primary gets saturated and high frequency it will not be able to handle...

how to avoid that , you need to seperate the transistor from the primary using the mosfets or flip flop circuit , but some point a big mosfet will take heat, i think that's why inverter has a huge heatsink with fan on it..

best of luck , i am gonna try some flip flops..

[totoalas]

Just cookrd my ring magnet for 15 minutes abd it becomes brittke as I pickep up the ring broke into two
the good news is found the nethod of cutting the magnets into half round
the magnetic strength seems a little weakrr

totoalas

[Matthew Jones]

If all you wanna do is neutralize the magnet get real big neo and push the 2 together.

I built a small one these for testing and it worked out, but that what I had to do is force the ferrite into a neo.

Matt

[SkyWatcher]

Hi folks, thanks for the replies and sharing.
Hi nick, well the gutted tests have paid off, found a circuit that have seen before, someone posted it again at the other forum.
It is powering a gutted cfl at close to maybe half brightness at 12 volt-200 milliamp input.
And most importantly of all, the 2n3055 is only getting slightly warm, though not sure if that lower transistor temp. will maintain when attempting to get more brightness.
Either way, this warm white cfl is putting out some very useable light, which is what was hoped.
Have to rewind 8" long ferrite bead rod back to 30awg. secondary winding, as was testing 24awg. and the 30awg. gives more brightness, we'll see if that remains the case with this circuit, if not, will just put the 24awg. back on again, hehe.
Here is the Alexander Meissner circuit, used 330volt-120uf capacitor at trigger winding and added 1kohm base resistor and left out primary coil capacitor, also left out resistor in line with primary coil.


Uploaded with ImageShack.us
peace love light
tyson

[SkyWatcher]

Hi folks, made a drawing of how this cfl is being powered in this setup.
Going to try and increase brightness a bit by using less turns, then try 18 gauge magnet wire, then 16 gauge lamp cord and hope that the heat remains low in transistor.
The way it is now, with the 13 watt warm white cfl, it is outperforming the leds in the bathroom and giving a much more incandescent quality and is drawing the same input power, around 2 watts.


Uploaded with ImageShack.us
peace love light
tyson