Simple Inverter

[Matthew Jones]

You would be better off just scrapping the trigger wire and going with somthing like a Fixed Frequency Crystal Oscillator.

Also those 3055's are just dated. They are good transistor but for higher speeds with the same SOA curve use a NJW21194G. Its the Direct replacement for an MJL 21194 at way lower price. Lower resistance

The crystal will hook between the 10k resistors (Which can be turned down to 1k if needed) to the base.

If your going to wind it with the trigger make sure to run the numbers and get the frequency up to around 40khz. Your going to be able to use less power and be more efficient.

If your curious do search on Florescent light Power frequencies, or the Lighting research center. Lots of info on what make a florescent tick.

I got parts coming for setup I got in mind, so hopefully I a can stop talking and start showing.

Matt

[SkyWatcher]

Hi matt, thanks for the reply.
Was already researching these fluorescent and cfl bulbs and they did say 40khz, no sure how to test for frequency, though there is no audible noise.
Maybe a little radio like lidmotor always does.
Will look into your other suggestions, thanks.
What setup do you have in mind?

Well, running some duration tests on this cfl oscillator as it is, to see how it runs for a few hours, reliability tests.
Then will try and tweak it for more brightness, though not really needed unless one needs lots and lots of light.
This picture shows 12.4 volt-250 milliamp input or 3.1 watts, was lower, though changed the positive to base resistor to 5kohm for a little more brightness.
Also tried a lower trigger to ground (emitter) non-polarized capacitor, says 1K, think that is 1000pf, though it caused a slight flicker after awhile.
Here is a couple pics.


Uploaded with ImageShack.us


Uploaded with ImageShack.us
peace love light
tyson

[Matthew Jones]

Hi matt, thanks for the reply.
Was already researching these fluorescent and cfl bulbs and they did say 40khz, no sure how to test for frequency, though there is no audible noise.
Maybe a little radio like lidmotor always does.
Will look into your other suggestions, thanks.
What setup do you have in mind?

I am going to setup like this video you posted: TROS alt-lighting VI - "CFL" - YouTube

But the schematic he posted is not all there obviously. He just has the multi vibrator portion. So if thats case why have 2 big chokes in the system? Look at his board. So I am just guessing but I think pre boosts the the voltage before he runs it to the transformer.

Thats kinda what I am looking at. I'll post a drawing this evening. I scratched it out on paper but I don't have any means to put it up. I'm redrawing it as an image.

The prospect of being able to light 6-10 cfls with a 12 volt a battery at or about 1-2 amp of draw sounds pretty motivating. I've done one before but not 6.

The thing is A CFL is basically a real high frequency capacitor, so if you get slick about it you can charge that thing then discharge it. Your recovery on it might not be real high but instead of just burning the power off you can grab little bit of it during your off time.You might could get down as low as 10 watts.

I have always wanted to build a solar lamp shade. Just dreaming, but maybe not. They do make clear solar panel material now.

Matt

PS @Sky. Some meters have frequency testers. I know radio shack has one. Also a scope is a pretty source. If you still have that stamp I can show you how to test with it... A little code and your on your way.

[hello_all]

for $26 you can buy the Digital Multimeter

it has frequency duty and capacitance , one thing that is missing in this one is temperature i think ..

this one is good one , i researched lot of them for a price below 30..

[hello_all]

This is video 75watts CFL on Joule Theif ver 3.0 - YouTube on running a cfl full bright , it took all the amps what it can from the lipo 11.2v 2300 amp

cfl is rated 16 watts (75watts equivalent) , u judge the brightness it is non modified cfl straight from the box.

without connection the volts out were 75 volts
with connection volts our were 110 volts

cannot measure amps but i think above 1amp ..

[SkyWatcher]

Hi matt, thanks for the reply.
Sounds good, plan on returning to that dual transistor circuit again also once this circuit is tweaked better.
Yes, mopozco results looked very motivating.
Unfortunately, stamp chip blew up, was too tired one night and wired it backwards or something.
Hi hello_all, thanks for the replies.
Looks like nice results you are having with the JR3.

Well, rewound the 8" long ferrite bead core with bifilar, 18awg. magnet wire-65 bifilar turns and it is more efficient than the other speaker wire coil, though just realized that it is not 20 gauge, it is 16 gauge speaker wire, thought it looked a little thick.
Also, that test made with the lower value trigger to ground capacitor was actually a varistor, thought it was a capacitor, is blue disc shape and probably explains the bulb flicker after awhile.
Lowered the positive to base resistor again to 2kohm and it is much brighter, drawing 450 milliamps at 12 volt input now.
So this meissner circuit can be adjusted with the trigger capacitor, positive to base, base resistor and of course windings.
Would like to try some thicker gauge wire, like 12 or 14 gauge, though do not have any as the 18 gauge seems to be working well. Do have a microwave transformer, though looks like a pain to get the 12 gauge wire off it.
Also working on a smaller ferrite bead core, since had a few more from surplus shop laying around.
This one has 4 beads glued together, like nickz did and we'll see how it differs from the longer 7 bead, 8" long model.
peace love light
tyson

[Matthew Jones]

Trial and Error

Hey Sky

This is the circuit I am going to try for now.

I might have made mistake when assuming the circuit could be driven straight from the crystal. The power on the crystal is not high enough to drive most transistors so it would take some Darlington setup to drive a good transistor, or an Op amp. To much trouble either way.

For now I am going to drive it with a 555. If it will light 6-10 cfl's on 12 volt 1 amp or less than I will most likely get rid of the 555 and go to a circuit that can use the feed back channel provided at pins 5-6 on Wurth transformer. We might be able to push the thing alot further than one would expect with feed back to the driving circuit.

What you'll notice in the circuit is the the Boost circuit prior to the transformer will run at 40khz 53% duty cycle, the voltage from this will go to 28 volt at or about 1.6 amp, minus diodes then enter the transformer after the Q1 or Q2 are turned off. The voltage coming out of the transformer should be at or around 30 volt at or about 1 amp (Probably less because of the CFL on the output end). But because I do not run it to ground but back to the Battery and the positive of C3, we are now driving the transformer from the potential difference created in the chokes. VERY CONSERVATIVE.... ....Reuse..reuse ...reuse...
Hopefully 40 khz is still slow enough to let this flow around in timely fashion. I have done it far slower and it works like a champ. But I am not sure how the speed will effect it as far as recovery. If all else fails I'll run it to ground and be done with it, hope for the best anyway.

Also if it will run an we can store a fair amount of power in the CFL's there might be a way to switch the remining power off the CFL's amnd back into the system.

But I'll settle for 6- 10 CFL bulbs running for now.

Matt

[mlurye]

12V fluorescent tube inverter 4 - 65W with high efficiency

[hello_all]

today ran a edison bulb , 40 watts ,it was half lit nice light so approx 20watts the system can max , also tried a single 12 inch tubelight and it light quite good , with no modification with blast and stuff.. full bright on that..

one thing i noticed it my lipo battery max out at 24watts so it is not able to go ahead of that.. also tried to run 2 of the 12 inch tubelight it only prefers one and does not light other together .. next i am getting 5 of 2n3055 to play around with the different circuits and i have a hfe from tv and heatsinks from there , would be fun to do something ..

another thing interest me is why to go for 110v ac from 12v dc , why now 12v ac , i see the led 5050 lamps for cars are cheap and run on 12v ac and they consume less ...anyone has idea on how to build something around 100 watts which can run 12v ac.. does 12v ac means 24v dc split in -ve and +ve side.. will have to figure.

here are the 12v ac /dc bulbs
LEDwholesalers Bayonet Single Contact LED bulb Ba15s 8-30v Dc & 12v Ac 60 LED 5w 300 Lumen, Warm White 1422ww : Amazon.com : Automotive

this is very cheap in the lower range

[SkyWatcher]

Hi folks, Hi matt, thanks for the reply, nice schematic, look forward to seeing your results.
Hi hello_all, thanks for reply, not sure 16 bucks is so cheap, though guess it is ones perspective.

Well, have got this 8" ferrite bead core tweaked for now, 600 milliamp, 12 volt input really cranks some light with this gutted 13 watt cfl and the transistor heat sink only has a touch of warmth.
Also got the smaller, 4 ferrite bead running and putting out good light also, though still have to check numbers and compare with the larger model.
It uses the same 30awg., one layer, then tried one layer of bifilar 16 gauge speaker wire and it only lit the bulb partially, then rewound with the 18awg. magnet wire and it fired up to a good brightness.
Think all the plastic insulation is placing the wire a little too far away for good induction or it may also be that it is too few turns, due to all the insulation on the wire.
Though have another 13 watt gutted cfl to use, because hooked the output to it with the guts still intact, seeing if it would light and heard something pop inside.
Probably way too much voltage and popped a transistor or something.
Heck, even had an alligator wire clip start smoking when powering the gutted cfl, never seen that happen with such low current, high voltage.
Anyway, will post a video or pics of the latest results when get a chance after tweaking some more.
Also want to use the long 8" core as the support for a lamp post, hehe, not sure how that will work, we'll see.
peace love light
tyson

[hello_all]

Sky.. I have seen meter show low milliamp on high frequency circuit , one time one guy posted 0 amp draw on exciter coil lighting heated LEDs , it is a fact that only way to measure it is thru watt meter or run a battery and see how soon it goes down... So my guess is you are blowing things ou meaning you have the amp juice flowing in and out ..

[hello_all]

On the other side I am seeing this a way to make blocking oscillator or a Pwm circuit in which the frequency and duty can be adjusted by capacitor resistor or magnets on coil All it is to find a bulb which can run on sawtooth or square wave without blinking ... Added advantage is there is means of getting the negative spike back from collector with diode , which I cannot find any use except to charge a cap and dump it to battery or charge a battery directly.... I will be interested in 12v ac LEDs they are most efficient for home .. And ac making dc losses off...

[SkyWatcher]

Hi folks, been working on a lamp to house this Meissner oscillator.
8" ferrite bead core is in the middle, hot glued in place and is adjustable using pot. form 80 to 630 milliamps, 12 volt input.
Is a very nice light and it has a screw in bulb fixture on top, taken from another light that broke.
13 watt gutted cfl being used.
Here is video of homemade lamp.
Alexander Meissner oscillator lamp with gutted cfl - YouTube



Uploaded with ImageShack.us
peace love light
tyson

[hello_all]

sky looks great , does the pot get hot on run , can you post a circuit diagram

[SkyWatcher]

Hi hello_all, thanks for reply and kind words.
It is the Alexander Meissner oscillator circuit, here is the updated circuit that is running this Meissner oscillator lamp.
When testing, noticed with this setup, that if the 2 kohm resistor was replaced with 1.5 kohm, the bulb started to flicker after awhile, so be aware the circuit may be finicky like that.


Uploaded with ImageShack.us

Oh and used cd case for base, plastic wrap cardboard tube for center pillar, couple parts from other lamps and prettied it up with colored construction paper and hot glue to hold it together.
peace love light
tyson

Edit: oh the pot does not get hot, probably because it is going to the transistor base to control output that way.