Simple Inverter

[Flumen]

croc clips

"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."

I've had several croc clips fail and give me false results. But I like the way I can change configs with them so quickly. I took the croc leads apart and soldered them up.

Flumen

[SkyWatcher]

Hi flumen, thanks for the reply.
Actually, all mine have been soldered for some time already, due to bad connections there.
Actually referring to, the actual physical metal contact between the croc clip and cfl wire, that is what has been unreliable for me, which was just realized today, by re-clamping the connections and instantly, the cfl started to light back up to the blinding light mode. hehe
peace love light
tyson

[SkyWatcher]

Hi folks, wonder if anyone can shed any light on this.
Have 3 nice high voltage NPN 60 watt transistors, salvaged from TV's.
Unfortunately, in the Meissner circuit, they only light to around half brightness and this with a very low, 100 ohm base resistor and the transistor gets real hot fast.
Notice that a diode is built into these transistors, across the collector and emitter and also a built in resistor across the base and emitter.
Anyone know a way to somehow bypass the base to emitter in some way, as it may be the reason for poor oscillation in this circuit.
It's HFE or gain is similar to the TIP42 and 2n3055, so that should not be the problem, maybe.
Don't think the built in diode would cause an issue, though any help would be appreciated, hate to see these nice transistors go to waste.
Here's a link to one of the salvaged transistors datasheet.
2SD2578 datasheet - Color TV Horizontal Deflection Output Driver (with
peace love light
tyson

[SkyWatcher]

Hi folks, an update on the ferrite TV yoke core.
Have achieved full brightness lighting the gutted 13 watt cfl, using 950 milliamps at 13.6 loaded voltage or 12.92 watts input.
Primary coil is slightly warm after an hour use.
Transistor heat sink is warm after an hour use.
Using A. meissner variant circuit.
Using same 47 nanofarad capacitor, same 1 kohm resistor from positive line.
Base resistor is 470 ohms.
Secondary is 2 layers, 400 turns total, 30awg. magnet wire.
Then on top of that, primary coil is 40 bifilar turns, 24awg. magnet wire.
Here is a pic of the gutted 13 watt cfl at full brightness.


Uploaded with ImageShack.us
Here is a pic of our mini led christmas tree, merry christmas, hehe.


Uploaded with ImageShack.us
peace love light
tyson

[SkyWatcher]

Hi folks, here is the updated circuit for using the ferrite TV yoke core.
Also making a small e-core, similar to the fuji transformers, probably a little bigger, we will see if it fires up a gutted cfl at the lower 1.5 to 3 volt inputs.


Uploaded with ImageShack.us
peace love light
tyson

[Nick_Z]

I am still working with the ferrite bead rod core, as well the the Tv yoke core. I think that this kind of oscilator circuit can not light both gutted as well as the un-gutted Cfls, unless there are different taps points on the secondary windings. Or the Led or Cfl bulb will burn out with higher voltages, that are neccessary to make the gutted Clfs, light up bright.
So I recommend leaving tap points, especially between the different layers of the secondary winding. This way you can use the same transformer cores to light both the gutted-ungutted Cfls, incandescent bulbs, as well as the Led bulbs.

[SkyWatcher]

Hi nick, yes, very good idea you have there.
And with the ferrite yoke core used in this setup, it has 2 secondary layers, 200 turns per layer, so tapping right at the middle point would be easy, even with this one here, already wound, thanks.
Was testing the yoke core meissner variant last night, to use less power and still have decent light, like for a bathroom.
Of course a potentiometer can be used before the 1 Kohm resistor shown in circuit, though those can be twitchy sometimes and cause bulb flicker.
So tested with regular resistors in series.
Got the input down to around 120 milliamps at 12 volt battery input, using 17.2 Kohm resistance from the positive rail to the transistor base, with very useable light and the DMM amp reading is reliable with this yoke transformer, it does not bounce around like some other cores do.
peace love light
tyson

[SkyWatcher]

Hi folks, this is for totoalas and anyone else interested, did not want to clutter up the joule ringer thread any further.
Wound another 30awg. secondary layer, when trying to power a single bulb off each of the 2 secondary windings, one lights up, taking all the power away from the other, so the other one is not lighting, maybe it would work if they were identical from the same package, not sure.
Though the good news is, wired the secondaries in series and also placed the gutted 13 watt cfls in series and they both are lighting to a very good brightness.
Though to get better brightness using this Meissner circuit, had to lower the positive rail, 1 Kohm resistor and just placed a 470 ohm in parallel, whatever that works out to ohms wise will be the resistor to use.
Will test for awhile to see if the transistor or anything else heats up too much.
Not sure if the circuit will be able to light 3 or 4 bulbs in series by adding one or two more secondary windings, that might push the transistor too far and primary.
Though maybe using thicker primary bifilar would do the job, if the transistor can stay cool.
peace love light
tyson

[SkyWatcher]

Hi folks, here is a couple pics of the 4 ferrite bead transformer running two gutted 13 watt cfls in series, running from a 12 volt battery.
Transistor is slightly warm.
Would like to try 3 or 4 in series, though not sure if there is any more warm white 13 watt cfls laying around here, have some 13 watt cool white ones that might work in series, not sure.


Uploaded with ImageShack.us


Uploaded with ImageShack.us
peace love light
tyson

[SkyWatcher]

Hi folks, just wanted to add another observation.
The two gutted 13 watt cfls in series together are actually a little brighter it seems than a 60 watt incandescent bulb used to compare.
The advantage of this, each bulb stays cooler, the bulbs color is more incandescent like and not so white and the transistor stays a little cooler.
Of course, each bulb alone is not equivalent to a 60 watt and is not blinding to look at.
Though when put in the same room, they cast the same or more light brightness/output than one 60 watt incandescent.
The circuit is drawing just about 13 watts from a 12 volt battery.
Hope this of value to anyone.
Will add another secondary coil and see how that works with 3 cfls in series, maybe it will become more efficient at a point, maybe not.
peace love light
tyson

[Flumen]

Hi SkyWatcher
Have just started on the 4-bead circuit, built one yesterday and couldn't get it to work, but I was using a coil with different gauge wire and turns ratio (a coil that works with the SJR 3.0). I'm using a TIP42c, though have got some 2N3055s and similar knocking about. (I gather you prefer the 2N3055). And maybe my gutted CFLs are different. I'll replicate your coil soon.
Quick question about your success running the two CFLs in series with the 4 ferrite bead core - just to confirm re your posts on 23rd December, you wound another secondary on to the 4-bead coil (so, at first, you wound another layer of secondary of 640 turns, but then you wired the secondaries up in series, so now you have a single 4-layer secondary of 1280 turns running the CFLs in series)? From the photo you posted on the 23rd - I can see what looks like four wires (2 each end) of the secondary coming out from the coil, which would indicate that you are using two separate secondary windings.
Sorry for rambling on.
By the way, have you got any further with the low-voltage Fuji-type coils?

Regards
Flumen

[SkyWatcher]

Hi flumen, thanks for the reply and interest.
The 4 ferrite bead core has 2 layers of 30awg. magnet wire, each layer has 320 turns for 640 total turns, connected each layer in series.
Then on top of that, 65 bifilar turns of 24awg. magnet wire, that layer is a little less than 3" in length, from end to end of core.
Had to lower the 1 Kohm resistor coming in from positive rail, down to 500 ohms to maintain brightness in bulbs.
Your TIP42C transistor should work fine with a heatsink, though will probably be a little warmer than using the 2n3055, which this is using at the moment.
The winding turns ratio is optimized for using the 13 watt cfls, though using the ferrite TV yoke core transformer, was able to light a cool white 23 watt gutted cfl to good brightness, though warm white color is more pleasing to the eyes.

In regard to fuji type transformer, was able to light a 13 watt gutted cfl to low brightness, using 2 AA nimh in series or around 2.5 volts, using a ferrite TV flyback core.
Hope that helps, feel free to ask all the questions needed to get the cfls fired up.
peace love light
tyson

[Flumen]

Many thanks, Tyson for your detailed description and for your encouragement. Much appreciated.
So, I understand that you have two layers. Are they made by one long wire wound up and down the ferrite (in the same direction), or did you make one layer and then decide to add another in series by extending the first one?
Great stuff, cheers

Flumen

[SkyWatcher]

Hi flumen, thanks for reply.
Oh yes, typed that into comment yesterday, but computer froze up while writing message.
Started with one secondary layer directly on top of ferrite bead, (with a layer of packing tape between ferrite and winding) to test and see how each additional secondary layer would work.
So yes, each of the two secondary layers are separate, just twist connected in series aiding, each layer.
Though you could just wind one long, 2 layer secondary winding, if you only plan on lighting only 2 gutted cfls in series.

It is interesting, considering the turns ratio for one secondary coil to primary, is only 4.9 to 1, to light a single gutted 13 watt cfl to just about full brightness.
Though using the 4 ferrite bead on 12 volts, seems to perform better, using the 2 cfls in series, which is a little brighter than a 60 watt incandescent.
Though the TV yoke core, can light a single gutted 13 watt cfl to full brightness with the 13.6 volt input, maybe because it is a 10 to 1 turns ratio.

That means, if we were to apply the standard transformer rules, of primary to secondary turns ratio.
According to those rules, then this 4 ferrite bead core transformer with 4.9 to 1 turns ratio at 12 volt input, should only be outputting around 60 volts AC.
Though in this case, since it is unidirectional pulsed, we must be getting over 400 spiky AC volts to run these gutted cfls at good brightness and it has to be the collapsing field of primary coil, that is mainly powering these bulbs.
Look forward to seeing your setup running.
peace love light
tyson

[Flumen]

Thanks, Tyson. All clear now
Got the beads glued together. Tomorrow, I'll get the wire.
I'm a bit slow on actually building stuff, but I'll let you know how I get on.
Mega thanks man.

Flumen