How to Make Joule Thief Light a CFL - Jeanna's Light

How to make a Joule Thief circuit that powers a Compact Fluorescent Lightbulb (CFL) using batteries, in this case the Jeanna's Light approach. I go step-by-step on how to make the circuit, including details on how to wind the coils which is the trickiest part. This uses just two AA batteries, a switch, a 2N3055 power transistor, a 100 ohm variable resistor or potentiometer, a ferrite toroid core, and for the coils, some 30 gauge magnet wire along with some more thickly insulated wire for the high voltage side. I also do some demonstrations lighting up my office, outdoors and for reading.
This video has correct English captions. Click on the CC button at the bottom of the video to see them.
For all sorts of measurements and demonstrations of this joule thief and CFL see my follow-up video "Fun with Joule Thief Powering a CFL":
• Fun with Joule Thief P...
For details on how a Joule Thief works, see my "How a Joule Thief Works" video:
• How a Joule Thief Works
For how to make a simpler Joule Thief that lights an LED using a dead AA battery, see my "Make a Joule Thief for Zombie Batteries" video:
• Make a Joule Thief for...
To see how to attach the transistor to a heat sink to prevent overheating, see my "How to mount/attach transistor to heatsink (2N3055)" video:
• How to mount/attach tr...
And to see how to make a very simple AM radio transmitter see my "How to Make AM Radio Transmitter" video:
• How to Make AM Radio T...
For more about the Joule Thief circuit, see my website here:
rimstar.org/sdenergy/joule_thi...
And here's Jeanna's channel:
/ jeannacav
To follow me on Twitter:
#!/RimStarz
rimstar.org
Music used for the first part:
A Darker Heart - music by audionautix.com
Music used for the second part:
5 Cents Back - music by audionautix.com

[@OverlandOne]

The original Jeanna's Light circuit used a TIP 3055 and a 3.375" ferrite toroid and output over 700 volts from a "dead" AA battery. Jeanna was an online friend of mine and designed a brilliant circuit. I built 3 of those lights here and they still work all these years later. Your videos always show great detail and easy to follow instructions for replication. I admire your ability to do that. Keep up the good work. PS The Jeanna Circuit was developed over at overunitydotcom. Bill

[@RimstarOrg]

I'll have to get back to you on all that. I did observe the voltage at the CFL on my oscilloscope at one point during R&D and the peak was around 800 to 900 volts. Still watching your latest videos - lot's of stuff there.

[@nowyoutryit1275]

Man I'm learning so much from you.. Thank you for making these videos. I can't wait for my son to grow a little older so I can share these projects with him.

[@ronaldgerth5120]

Outstanding! Great practical application for one of the Internets most cherished circuit. Clever build and nicely compact. Thank you for sharing.

[@Slider2732]

A simple rule of thumb with it all, is that the lower the number of primary turns, the brighter the light but more the current draw. A balance can be reached and that's a main interest in such circuits :) Yours is great for what it was designed for, reading, emergency light, night light perhaps and fun !

[@RimstarOrg]

Thanks. As usual it's the kind of tutorial that would have helped me. But this way I got to have the fun of figuring it out! The coil was tricky to get right.

[@RimstarOrg]

Thanks, Dan! You're last one was pretty cool too. I'm looking forward to your next one.

[@RimstarOrg]

Thanks. If you have any problems building something from one of my tutorials, don't hesitate to ask for help. I'm always delighted when someone finds them useful. And I don't know the battery life yet. I'll let it run this weekend while I watch it.

[@emlokkolme9504]

Thanks So Much!! Finally found a video that helped me forward on my project. Thanks!!!

[@Aaron-rn6pg]

Very interesting video, thanks for making both it and the video describing the operation/principles behind the Joule Thief circuit :) .

[@dhruvshah7488]

You make the best electronic videos!

[@pleaseyourselfsir]

Great work!.... Thanks for taking a chunk of your time to male these kind of video's, I learn more from your videos than I do reading a book for 7 hours!! ;-)

[@RimstarOrg]

Be sure and upload a video when you do. BTW just enjoyed your videos. Some cool yet simple stuff you have there.

[@RimstarOrg]

Thanks! And thanks for feedback re the music. It sure beats just hearing me talk and is hard to select so that it helps rather than hurts.

[@RimstarOrg]

It was a conscious descision to put only the one 1.5v battery. I started with a joule thief lighting an LED which normally has just one 1.5v battery and I decided to leave it like that since the CFL circuit could have any number of 1.5v batteries depending on your resistance, core and windings. And I think I found the variable capacitor mention you're talking about. It's in the labeling at 0:47. I added an annotation. Easy typo for me to make since I've used variable capacitors so much. Thanks!

[@RimstarOrg]

You should tear open the CFL and give it a try that way. Even a broken CFL might work. Of all the CFLs I've opened that "didn't work" anymore, the tubes were fine and lit up with this circuit. Thanks for sharing what you did.

[@RimstarOrg]

By "without using electricity" I guess you mean lighting the compact fluorescent light without plugging into the wall. I've gotta admit it is a cool feeling holding the CFL in your hand with no chord and walking around or sitting reading a book. Like magic. :)

[@RimstarOrg]

I guess. I actually did it the equivalent of 3 times in all my experimenting. :) Glad you like the video.

[@RimstarOrg]

Thanks. And heck, there wasn't even a video like this around for me when I made it just recently! Now there is, though. I just needed a time machine to go forward and find it useful for me! :)

[@Mast6r90]

Great idea. You are the best on KZbin. Thank you.

[@RimstarOrg]

Glad to hear it! Thank for letting me know. Cheers.

[@DiyEcoProjects]

Thank you Rimstar... youve videos are always very clear. Readers, you may like to try this with a LED bulb (light of amerca 4w) as you often get more amp hours that way. 5.11 thanks for saying this about the CFL, i had to research this a while back Chears Rimstar, your a star!

[@PranabLodh]

Okay We'll wait for it.......but you and all of your gadgets are just AWESOME....

[@RimstarOrg]

It'd be cool to see how well it works with your ion cells. I love it when as much of it is homemade as possible.

[@dsvc4591]

thank you man for the info. i love your videos and the quality. thank u man. keep it up

[@MYNICEEV]

And as always you did it. Will be watching for the next one.

[@RimstarOrg]

Thanks! Yeah, it helps when it's in a more engaging format. Thanks for watching!

[@RimstarOrg]

Thanks! The ASZ15 transistor is a PNP type transistor whereas in this video I'm using an NPN transistor. However, you can use your PNP transistor if you simply reverse the positive and negative of the battery in the circuit. Everything else is the same. The ASZ15 is a lower power than the 2N3055 I used so hopefully that's not a problem. I just tried with an NTE219 PNP transistor which is close to a 2N3055 in power and it worked. Sorry for the delayed response. I wanted to find and try one first.

[@RimstarOrg]

Glad you liked it, DiyEcoProjects. Yeah, it took some hunting and reading for me to figure out what to do with the CFL's wires too. At least now, anyone who watches this video will know.

[@scottdavis5739]

Really good responsiveness to questions. Good videos. Subscribed.

[@RimstarOrg]

Welcome! And thanks!

[@RimstarOrg]

Cool. Sounds like you're building one. Have fun with it!

[@RimstarOrg]

I don't want to try LEDs right now since LED flashlights powered from batteries exist, makes it less fun. But powering CFLs from batteries is exciting! I chose my core completely at random. I selected it because it was the biggest I had. It's also an iron powder core which people seem to think wouldn't be great. Maybe it isn't but it worked and it stayed cool while the transistor and batteries heated up. So I'd say, just try what you have. And I'm glad you liked it. Thanks for watching.

[@Ozzy3333333]

Why not run the lamp from the collector current coil? What kind of run time you get with a new pair of AA cells?

[@MYNICEEV]

Fantastic tutorial. Many many thanks.

[@RimstarOrg]

Thanks mate! Glad you liked it!

[@pronmaster09]

thanks a lot for this educational video! i had many hours of fun trying to reproducing this. i never lit up a CFL, but i didn't take it apart. i did however get a high AC voltage from my joule theif, which is about 1/3 of ur size core and 200+ magnetic wire winds and about 5 winds for the two primaries. i don't have a oscilloscope, but my volt meter vibrates at HV so it's pretty high(when passing 4.5V through). using 2 c1846-r transistors in parallel. the green/blue core is a bit under an inch.

[@RimstarOrg]

Thanks. I try to keep them helpful.

[@KyleCarrington]

Very well done. I wish this video was around bout the same time we were all building them the first time hehehe. Excellent job

[@RimstarOrg]

I was wondering if anyone but me noticed that! It was purely a happy accident and I love it. :)

[@RimstarOrg]

+Nishi Chemie If you don't have a variable resistor then try different fixed value resistors since your resistance might be too high. Also, see my follow-up video where I measure the resistance kzbin.info/www/bejne/r6vCamyli7KVias Your problem may also be in the wiring. Double-check the way you connected to the coils since it's backward from what you'd expect. It's also possible that you had the problem I talk about at 2:16 if you didn't put thick insulated wire at that point. For an easier one, try powering an LED instead. You can see how to do that in this video here kzbin.info/www/bejne/eGeUdYhtrqqmhq8 PS There's no Reply button under your comment because of your Google+ settings. - go to your Google+ page, - in the top, right corner click on your thumbnail icon, - in the popup that appears, click on "Settings". - for the 2nd question down "Who can comment on your public posts?" set it to "Anyone".

[@RimstarOrg]

Thanks. Glad you like it.

[@RimstarOrg]

Thanks. I've tried only the one core so I really don't know if any random one will do, but it was a random selection on my part so I'd say go for it. Judging by all the other joule thief powered CFLs out there I think most, if not all are randomly selected. Most are actually likely iron powder cores and not ferrite as I said in this video. I didn't know that when I made the video. I should put an annotation. I'll have more about it in this week's video.

[@siliv653]

very descriptive, really nice, thanks for the video

[@RimstarOrg]

Thanks! How are you calculating the efficiency?

[@GREENPOWERSCIENCE]

EXCELLENT VIDEO!

[@denny8548]

Thank you Mr. Rimstar for all wonderful educational videos. That IBM compatible PS1 keyboard scare me. LOL.

[@RimstarOrg]

+Denny Smoke LOL I have a newer keyboard stored away but I like the deep press of the keys with this one. And I recently cleaned it so it's good for another frightening 20 years or so! And thanks, I'm glad you're enjoying my videos.

[@GC-wc8wj]

Another great video... thanks

[@RimstarOrg]

I don't turn up any good search results for either dry crystal solar cell battery or sn305. Can you elaborate?

[@RimstarOrg]

It's definitely 18 AWG. It says so right on the spools and I used the 18 gauge location on my wire stripper to strip the insulation. Must just not look the same through video or just be a different one than you're used to.

[@RimstarOrg]

+Nishi Chemie You can increase the output of your solar panel by giving it more light. Keep in mind that your solar panel can give out only so much voltage. If you want to get more voltage and amps from your solar panel but without increasing the light then you can't. There are circuits that will increase the voltage while dropping the amps or that will increase the amps while dropping the voltage but you can't increase both at the same time without supplying more energy, i.e. more light.

[@RimstarOrg]

***** I buy my 2n3055s at local electronics stores, though it's cheaper to buy them online (e.g. ebay.com). I don't understand the rest of your question. If you're asking for a different circuit to power your fluorescent light, I don't have any other circuit off hand.

[@RimstarOrg]

***** No, I've never heard of it, so I probably can't help you with it (at least not with my available time.) But I did a search and found a few things, but nothing black and heavy. Do you have a link?

[@RimstarOrg]

***** Thanks. Looks like an automotive ignition coil and a battery. I guess the flame comes from having an ignitable gas flow through the arc? That's what it looked like in some of the videos I saw when I searched earlier. I don't have any circuit to add to it to power a fluorescent light. Possibly something with a 555 timer chip and the ignition coil will do it. Try looking for a circuit with those two things. I know I made one once but I never documented it. I didn't try a fluorescent light though.

[@RimstarOrg]

***** You're going to have to search for that yourself. I'm not even sure it if will work since I haven't done it. It's just a suggestion so that you have something further to look into. I only know that I once managed to make a high voltage power supply using a 555 timer and an automotive ignition coil, but I never used it with a fluorescent light. I don't have the circuit anymore.

[@AtomicElectronCo]

RimstarOrg Actually, using an ignition coil+555 chip ought to work REALLY WELL to light up a CFL although it won't be so compact any more. Should be interesting to try.

[@RimstarOrg]

I don't know yet. I'm going to open it up and do some testing this week (voltage, resistance, current, oscilloscope, ...) and put it all in a video.

[@RimstarOrg]

Using the 13 watt CFL in this video the two batteries lasted 25 minutes at full brightness and 30 minutes dimmer. Using a 5 watt CFL the two batteries lasted 90 minutes at full brightness and 15 minutes dimmer. I show these tests in my follow-up video "Fun with Joule Thief Powering a Compact Fluorescent Light". There's a link to it in the description below this video or you can find it in the Videos section on my channel page. Others have different runtimes based on the coils, resistances, ...

[@RimstarOrg]

Mine doesn't make any noticeable ringing sound. I do recall at points while I was still experimenting it, hearing the ringing that I hear in a lot of other videos but it wasn't very loud. I don't recall what the circumstances were.

[@RimstarOrg]

I'm testing that this week. It'll be in Friday's video.

[@RimstarOrg]

You're welcome. Thanks for watching.

[@mhr1311]

great created,easy explain--thanks

[@RimstarOrg]

The potentiometer controls the brightness of the light. I'd already adjusted it before filming. However, you can see it in action in my follow-up video "Fun with Joule Thief Powering a Compact Fluorescent Light". There's a link to it in the description below this video or you can find it on my channel page.

[@RimstarOrg]

Oh, I like that idea!

[@RimstarOrg]

+Nishi Chemie You can't do it with a joule thief circuit. But if your input is AC then you can look into "step down tranformers", and if you want DC output then add a full wave bridge rectifier after that. If your input is DC then one solution is a DC-DC converter. I don't have anything to help you further, but maybe knowing the terms to look for will help.

[@BlenderGameArtist]

nice! music is very good, works very well!

[@RimstarOrg]

I'm not too familiar with trigger coils - just had to look it up. Wouldn't that give you just one brief flash with a long wait before the next one? I don't know, they're new to me.

[@RimstarOrg]

Thanks, Samimy. And thanks for the tip.

[@RimstarOrg]

The lower the resistance of the resistor, the lower the frequency and the higher the voltage and visa versa. You can see me do this by adjusting the potentiometer (variable resistor) in my follow-up video "Fun with Joule Thief Powering a Compact Fluorescent Light" at around 3:12 into it. There's a link to that video in the description below this video.

[@efrem055]

Thaks very much. I made it and it works well. God bless you.

[@silenrocc]

Great job !

[@RimstarOrg]

Actually, comparing it to some I've seen, I thought the light level was pretty good. I was surprised I could easily read by it! I did try a 9V one early on and it was brighter than whatever else I was doing at that stage. I think I didn't keep using it because I wanted to keep the battery voltage and power small, since to point is to get the best results with the least power. I started out trying just one AA for a long time without success.

[@RimstarOrg]

I didn't have success with a cylindrical core, but I tried it first before I realized the issue with voltage breakdown through enamel insulation. Others seem to have had success with cylindrical cores though.

[@RimstarOrg]

I suspect a ferrite rod might work, though I don't know how to layer the coils or if you should. I started out with trying my ferrite rod and didn't have any luck so I switched to the toroid. But I was trying only one AA battery when I was experimenting with the rod so it may or may not have worked had I tried two AA batteries. I've seen videos of others using similar circuits and longer rods.

[@RimstarOrg]

The first thing to check is your connections to the coil. They're connected the opposite of what you'd expect so compare them carefully to the diagram. The other thing to do is try a higher input voltage. I was trying for a long time with just one AA battery but when I finally added a second AA battery connected series (positive of one battery connected to the negative of the next battery) it started working. You could even try 2 or 4 AA batteries.

[@Ziurkyss]

Hey +RimstarOrg it work with bf869 transistor? I readed about it is NPN high voltage transistor

[@RimstarOrg]

You can get cores in many things: microwave ovens, power supplies, TVs, ... I think stacked washers would be susceptible to eddy currents across their surface. But make sure to put something thin and non-conductive between each one to prevent eddy currents through the stack (waxed paper, thin plastic from a plastic bag, ...). It probably won't be as good as a ferrite core but it's worth a try to see how good they'll be.

[@bakupcpu]

Awesome work :) Thanks for sharing :)!!!

[@Graemereserve]

Xlent video mate !!!!

[@chems]

I have done everything exactly as you have and I am getting a voltage on the small windings but its not enough for a CFL, i can barley light an LED

[@RimstarOrg]

+chems Try adding another battery in series (+ to -) for more voltage. One AA battery didn't work for me so I added a second one and it worked.

[@chems]

+RimstarOrg I have used a 9V battery. - UPDATE - I remade the ferrite core to be EXACTLY as you've done, 30 gauge wire with thicker wire where ends meet, some packaging tape, 10 turns of (admittedly 20 AWG, not 10) and 5 turns of the same wire. Basically what you've done. I connected it with the exact same circuit except I'm using a Tip41c which with all my tests works almost the exact same as the 2n3055. I use a variety of variable resistors, 10k 1k and 100 ohms. Heres the odd part: Once powered the CFL doesn't light. To see if the circuit even works I connect a 20W LED (because previously I got a 20W to get insanely bright off a 9V battery with a different core) and it immidetly brighten and then quickly dies down to nothing...? I adjust the ohms and it does the same thing but at different brightnesses. If you didn't get that: I powered it, it turns on brightly, then slowly dims to no light in about 2 seconds. Possible reasons: 1. Tip41c may not work the same as the 2n3055 in this circumstance. 2. The 2 10 gauge wires are at different coils. I was wondering why one was 5 turns and the other is 10, why is that? I think this might be the problem because of the inbalance in turns, the magnetic field gets weaker every cycle. Until it turns off? Maybe then I have to switch the wire so the inbalance works the opposite... I don't know if anything I'm saying is making any sense. But I'm not sure. I find it really awesome that even though this video is 2 years old you are still helping people out!! Your awesome and smart! :D

[@RimstarOrg]

+chems The number of coils for the 2 10 gauge wires were arrived at by trial-and-error, basically trying different numbers of turns and seeing what worked best. You can try different numbers too. From the datasheets the Tip41c looks comparable to the 2N3055 as you say. Maybe you should try lower input voltage like mine, only around 3 volts. Possibly your input voltage is too high causing a voltage at the wires that's too high and breaks down the air. The insulation is already probably damaged from the first breakdown. That would explain it working only briefly. When you turn it on, the corona would take a little time to build up between the ends of the wires before the air becomes conductive enough to act like a short circuit. When you turn it off, the corona goes away. If that's the case, using a lower input voltage might still work as the lower high voltage at the wires might not be enough to form a corona in the air. The voltage reached before the breakdown may not be high enough to ignite the CFL but might be for the LED (I'm not familiar with the LED circuit, though I doubt it requires high voltage.) If it still doesn't work with the lower input voltage then remove the 2 10 gauge wires so that you have just the enamel wire and apply some more tape, basically to plug the holes from the breakdown. Or you could paint on some corona dope. You might also want to look at my follow-up video for some ideas kzbin.info/www/bejne/r6vCamyli7KVias.

[@chems]

+RimstarOrg The LED wasn't connected to the high voltage, it was connected to the Collector and Emitter to see if the "joule thief" part of the circuit even worked. I did try putting the LED on the high voltage and got the same results. But you may be correct with the voltage damaging the wires. I'll go take off the 10 gauge wire and try to fix the insulation of the 30 gauge and see if that works, otherwise I might rewind wire onto a new core. Thanks.

[@RimstarOrg]

I was originally thinking of embedding it in the top of a wooden staff (ala Gandalf's staff or something) but I love how the cola can turned out instead. Interesting stuff of JonnyDavro's. It seems if you look at just the basic principle of 'store up energy and then release in a burst' there are a huge number of ways to do it - probably even mechanical! I only measured current at the early prototype stages and it was over an amp at one point. Don't recall which coil iteration that was though.

[@RimstarOrg]

To make your own cores you could try using large washers stacked on top of each other with a layer of plastic or paper in between each to separate them electrically. I don't know how well or even if it will work since there will be large eddy currents compared to a manufactured ferrite core. Sorry for the delayed responses. Email notification of comments doesn't seem to be working.

[@ifoxrider]

nice video. You must be very patient to do that coil around the core :P

[@RimstarOrg]

My toroid is 1 1/4inch outer diameter and 5/16inch inner diameter or 33mm outer diameter and 15mm inner diameter. But I just used the biggest one I had. If you have a bigger one then use it so that you can get more turns on it. It'll also be easier to wind. There was some question if a ferrite core would be better than an iron powder core (like mine) but since mine stayed cool during the tests in my "Fun with Joule Thief Powering a Compact Fluorescent Light" video, iron powder cores are okay.

[@TumpaTalapatra]

Nicely explained. Can i use mosfet irfz44 i have bunch of those.

[@RimstarOrg]

No, a mostfet would require a different circuit.

[@RimstarOrg]

I don't know how long it lasts yet. I don't have any LED comparisons. I tried a reflector and it didn't help much. I don't think it was bright enough to matter. Same for solar cells I suspect. But the fun was just in getting it lit as much as I did, more than I expected in fact. Adjusting the potentiometer does adjust the brightness. I turn it down to min resistance and then increase the resistance while the brightness stays the same, using as little current as needed.

[@martinlicht1969]

I've heard of versions of this circuit that claim to self charge the battery running it ...? Thanks for the sharing of your projects.

[@RimstarOrg]

Thanks. I don't know yet. I'll reply again when I do.

[@nishichemie9282]

hi ok!do you know a way to increase amps and decrease voltage.i like it to be 9v and 0.8amps

[@harshbawari414]

Have you tried to make a tesla coil. The last post was also mine. Now I am using a different id. I am having a bar type core can I use. Also can you give some formulas to calculate the windings.

[@RimstarOrg]

No, the distance is the length of just the tube itself, from the end of the tube where one set of wires connect to it and the end of the tube where the other set of wires connect to it. In order for the tube to light up, the gas inside the tube has to be ignited from one end to the other. For a spiral tube that's the length of the tube as if you'd straightened it out and measured from one end to the other.

[@RimstarOrg]

Thanks, my friend!

[@AscanioBorsato666]

wonder work maaaaan! amazing!

[@RimstarOrg]

It's not only counter intuitive, but it's the opposite of what I get. Are you sure you're measuring the resistance correctly - though I don't see how you can be?

[@RimstarOrg]

I've gotten it working with 2, 3 and 4 AA batteries and there wasn't much difference in brightness so I suspect 1 would work too. The biggest difference might just be in battery life. The more batteries, the longer they'll run for.

[@GamerKid1024]

Appreciate your reply.

[@RimstarOrg]

It's a biography, "Robert H. Goddard - Pioneer of Space Research" by Milton Lehman. Goddard was an American who did rocket research in the 1920s and 1930s and is known as one of the fathers of rocketry. A fascinating man and a great read.

[@RimstarOrg]

I don't know how you'd adapt a camera flash circuit. It's the same principle though. Build up a store of power and release in a burst. Just a question of increasing the frequency.

[@brianmoon1777]

2 Questions: Does the size of the ferrite core matter? and does how close the wires are on the ferrite core change the voltage?

[@RimstarOrg]

If you're not getting a voltage reading from the many-turn coil then it sounds like you may have a short circuit, maybe due to high voltage when you had only one battery, not enough to light the CFL but enough to get high voltage. How thick is the insulation for you short coils? It's possible to have a short between the short coils and the many-turn coil. If that's the case, you might be able to just change the wiring for the short coils and the short on the many-turn coil won't matter.

[@xerxesdacat]

Thanks for letting me know, im going to try the wax paper spacers and im going to try different windings now that you tell me its not going to be as efficient, quick question, galvanized washers wouldnt make much of a difference now would they?

[@sullivanzheng9586]

Very COOOOL invention

[@mikaeljiskovkristensen7861]

hello. now i had made this joule thief. can it be done with a sox(low pressure sodium lamp bulb). my friend has send one to me, so can you explain about the possibilities to drive sox lamp by a joule thief and the necessary circuit.

[@SAMIMYS]

very nice project, educational and fun ! i am sure that little AA battery is very Happy among those monster size components :) Steve i don't know you smoke or not, but you should keep a bottle of Zippo Lighter Oil in your lab cause nothing can remove that sticky label residues rather than that oil. i had that problem with HDD label removals, nothing helped until i found the magic :) good luck with next one.

[@nishichemie9282]

hi! i need a way to boost voltage and amps output from my small solar panel. it's 0.17 amps and 18 volts.i need it to be about 30 volts and 0.8 amps.is there a way to do this?please tell me if you know.

[@RimstarOrg]

I don't know any other transistor numbers off-hand that'll work but one way to find some is to do a search for "2n3055 equivalents". You can also use PNP transistors if you reverse the battery in the circuit while leaving everything else the same (the 2N3055 is an NPN transistor.) I do that in my "Fun with Joule Thief Powering a Compact Fluorescent Light" video (see link the the description below this video) using an NTE219 PNP transistor.

[@RimstarOrg]

A frequent error is to get the connections to the short coils wrong. Very carefully compare the diagram to how you've done it. Notice that the two short coils are connected to the circuit oppositely each other. If that's okay then add more batteries in series. Mine wasn't working with a single AA battery but when I added a second one it started working.