Coils have some capacitance, but important here is that a magnetic field builds up. When the magnetic field has built up enough to saturate the ferrite core, the interaction between the coils shuts off the transistor fast. That causes the magnetic field to collapse quickly, building up more than enough voltage at the LED to turn it on. There is an interaction between the two coils that controls the transistor to make this work well. I haven't thought if the same could be done with capacitors.
I'm delighted you're enjoying this world of electronics. It's hard to make a videos that'll keep the interest of someone totally new to it. Thanks for giving it a watch!
Hi Debasis, You can most likely use the core. I don't see why not. However, without knowing the part number of the transistor I can't say if it would work or not. I'd suggest you try it and see. Keep in mind that if it doesn't work right away you should first check the way you wired the coils into the circuit as I mention at 4:28 in the video. Also, some LEDs are not very bright (the mcd value) or shine in one direction mainly so look carefully at the LED to see if it's lit. Have fun! -Steve
The resistor protects the transistor from too much current. At the same time, the higher the resistance the dimmer the LED, so if you have a potentiometer (variable resistor) then you can adjust the brightness.
The voltage of the solar cell will vary just as the output of the joule thief varies. An LED lighting up for short bursts but very bright can have the same overall power output as an LED lighting up for longer bursts but less bright. A better approach would be to try charging a capacitor for a fixed period of time and then disconnecting and measuring the voltage of the capacitor. The energy accumulated in the capacitor is 1/2 x capacitance x voltage x voltage. This gives a result in joules.
Yup. Many people power compact fluorescent lightbulbs with it, and still with just a single 1.5 volt battery. But there is a limit to what it can power, both because of the available power and given the voltage waveform I showed in the oscilloscope output.
Transformers are for a number of things e.g. stepping a voltage up or down, electrically isolating parts of a circuit. If you transformer has 4 prongs then there are probably 2 coils, 2 prongs for each coil. Measure the resistance of the two coils. Apply voltage to the lower resistance coil and the voltage on the higher resistance coil should be increased. Note that this happens only when the voltage is changing. Straight DC will transform only when being turned on and when being turned off.
Did you try reversing the wires coming from the coils? That's usually the first thing to check. Also, if you were using a breadboard like I did, go back and check that the connections are all where they should be and that they are fulling in the holes, also very common issues. Also double check that you have the transistor legs right, they're not always the same for all transistors. Also double check the polarity of the LED, one leg is positive and one is negative, they're not interchangable.
I don't really know for sure. Ferrite seems to have a few meanings. The ferrite core here is a poor conductor and is a ceramic whereas ferrite also refers to iron. A ferrite core, being a poor conductor doesn't have as a much of an issue with eddy currents being induced in it as iron would. I guess you'd just have to try.
Ah, I see. So you have the output of the peltier module go into the joule thief. Good to hear that you're getting some result so far. Thanks for clarifying.
A crystal radio puts out fluctuating DC at the output. Someone suggested I try powering a joule thief using a crystal radio in the comments to one of my crystal radio videos (hmm... can't seem to find it) and that's why I made this. So far the radio's DC output is too small to turn on the transistor but I'm still trying. My sources are small local electronics stores and Radio Shack (The Source in Canada) is the only more widespread store I can think of that might have germanium diodes.
mA is amps/current, not power. It's also extremely difficult to measure current for this type of spikey output so the 31.3mA is probably not accurate. But anyway... People who've played with the resistor value have seen it affect the LED brightness so trying different resistors or using a variable resistor (potentiometer) should help I think because it affects the timing. Decreasing the turns may work too since the power flows through the coil that's not connected to the resistor.
Ah, so that's what it was. I've made only very simple power supplies. There's a heck of a lot more than I'd have expected. Thanks for clarifying. Happy to hear you like the video. Thanks.
I haven't tried more than one. I guess more than one might work. The LEDs need current too so after adding some number of them you'd start to notice they're not as bright.
Yeah, without an oscilloscope, if your battery voltage is high enough to light the LED anyway, then it's hard to know if the joule thief is doing it's stuff. I once lit an LED with the joule thief circuit before only to find with my oscilloscope that the joule thief wasn't working at all.
Sure, if you can find them. Some electronics stores might sell ferrite cores. Usually the ones I see in the stores are pretty small, but then again, we don't have really good electronics stores here. I just had a quick look at ebay.com and saw some there.
I don't know. I never tried one. I'm talking really tiny cores, ones that are for threading a single wire through, so the ones I have in mind wouldn't work anyway because you couldn't wind the coils on them.But it wouldn't hurt to go see what's available in the stores near you.
Also, the insulation of the coils may not be able to handle the high voltage. In that case the insulation will break down and the transformer will be ruined. You need to know the voltage rating of your transformer coils, or take a chance. :)
You cannot rely on the face of a transistor for a correct pinout guide. You should ALWAYS look at the datasheet for the part number of the transistor. Also one should consider the wire gauge and amount of actual turns on the core, this is important and will make a difference between running one Red LED or several White LEDs. I have built these things using a coil wrapping design for a power transformer (primary coil and secondary coil) and I had success running small DC fans for vehicles and even a car stereo off two AA batteries. Of course you want to use transistors rated for higher volts and currents and heat sink them properly.
So, after many hours, i threw out all the peices,and got new ones, finnally got it working, must have been something broken, sorry for being so hopeless, great advice, great vids, keep up the good work and thank you!!
so this is the single one that worked, i had reproduced other 3 "youtube diagrams" and this is the working one. the diference being the resistance placement.thnx man
Not a 1.5v battery. A car's battery is 12 volts but that then goes through an ignition coil to produce a much higher DC voltage to produce the spark in the combustion chamber. I'm no car expert though.
I'm no expert on measuring these odd waveforms. The only way I know to measure voltage like this is to use an oscilloscope and even there I've seen mention of possible errors - but it's close enough.
The coil winding really is as simple as the little I showed. You can find a diagram in my "How a Joule Thief Works" video. There's a link to it in the description below this video and you can find it on my channel page. The wire sizes don't matter much, just make sure they have insulation on them i.e. not bare wires.
I'm not an expert on transformers. I just know a bunch of stuff about them from research for projects. But go ahead and ask. If I don't know the answer, maybe someone else will. And if it's too off-topic from joule thief then we can try private messages.
+Razia Binate Saleh I've looked in stores and the only one I've ever found on in was a local electronics store (Gervais Electronics if you're in Ottawa, Canada.) You could try science stores like the ones attached to a museum or something like that. 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".
Thank you. Sort of like a set of points collapsing the field to an ignition coil. but instead of the primary field collapsing mechanically, the joule uses the transistor triggered by the saturation point of the ferrite core.
Yup. I just fixed it. Thanks! 820 ohm or 1kohm are both fine, though having a potentiometer to try different values will allow you to find an optimal one for brightness.
Good on you to keep at it until it worked! Did you light an LED with the peltier powered joule thief? You should upload a video - it'd be great to see.
I have been thinking or replacing the 1000 ohm resistor with an avalanche diode. I hope it would cause a voltage drop while making a jump in current. Votage may not matter as the led only needs 3 or 4 volts.
New RimStar Video...woot woot! Holy hell he's like freaking MacGyver, he can rig up anything! I liked the start too with the zombie batteries! Always very interesting how you can make all sorts of appliances and other electronics by jimmy rigging up some wires and batteries! I need to get my head into the books more dude, you are so smart! Liked and will share!
Just made a Joule thief myself and it works very nice :) I winded the coils with different length on the ferrite core (Longer for the LED and collector) it can light up a 3W 3,7V powerled with only one AAA battery that is as low 0,7V (0,5V under load) even if just a tiny bit of light. a battery with 1,2V (0,6-0,7V load) is enough to get a pretty strong light from that led. running 2 parallell depleted AA batteries gives even more light :) thanks for a good video sir
RimstarOrg Hi it's me again :) when I put in those 2 AA batteries that had 1,4 volts in them has been lit 24/7 and enough to light up the whole room in the night ever since I posted that last comment. I never thought it would run for so long before the batteries would die. they still give light but now so weak the led just glows. really decent light for 8 days and weaker the last from already weak batteries. sick
I made one version with a 10K potentiometer to set the brightness, and with half of the brightness the led lasts extremely long (I'm using a diffused blue led that shines bright enough even with half the normal brightness). With full 10K of resistance it shines a lot dimmer but since that is under your control, it can be very efficient.
I actually forgot to mention that I still have a resistor connected in series with the potentiometer: A transistor needs some resistance or it can get really hot.
I'm using the peltier+joule thief to turn heat into electricity that can be collected from readily available sources like hot coffee cups, the sun heated roof of a car ect., i think they are also called seebeck generators, they are very simple devices, i wanted to make my own after seeing a camping water kettle with a peltier built in.
Thank you for making videos like this, im happy to see that there is still a corner of the Internet thats actually usefull & worth watching in the era of reality television, jersey snore, & honey boo boo! Thank you from the bittom of my heart for giving us something worth watching! That being said i just assembled this very same joule thief circuit and im testing it with a $10 peltier cooler i bought at amazon, so far its kinda so so, i need to adjust it more.
Oh, I did at least look at it and even attend a workshop a number of years ago, while attending a fair for something else. No one there, even the long time builders had succeeded so I didn't see any point in pursuing it further myself.
get all the transistors out of and google the parts number. get a general use or switching type "NPN". I know the transistors from a camera flash circuit work. but not nearly as well as something like the 2n2222 or 2n4401, etc.
I cant figure out pm's, so i guess my main question is, is the main point of the transformer to have a current flowing one way and then a current flowing the other way? I have a very simple 4 prong transformer, actually i have a few differant ones with all with 4 prongs, and also im having trouble with resistors because im not sure what one to use, i am working with 2 aa's and have 5 1kohm resistors, and 4 smallish transistors, the idea is making a very simple high voltage driver
Thanks very much on the tutorial, my slayer exciter didn't work for te upcoming science fair on 6-th of April, so i remembered on your joule thief. I suceeded to make it on the first try!Thank you veeery much!
+Andrew Kimble I get my parts from local electronics stores, from taking apart electronic devices (microwave ovens that people have thrown out are great, but so are a lot of other things), and less frequently, from searching online.
can i use a mosfet ? because i only have 2 pairs of irfp250s right now,or i have to buy a power transistor i guess i'll try itmyself,with irfp250,flyback transformer core ,and some 22 awg copper wire
A 2N2222 General Purpose bi polar transistor is the one recommended Though you can look into other higher powered types like those used in power supplies. Also you do not have to use an NPN type, you can use a PNP type with slight modification to the circuit.
24 volts huh? Question. Could you use the 1.5 v battery to start or jump a dead car battery? If not then...Could you use the car battery and the joule thief circuit to start the car?
Adam Fattal I don't even know if it's possible for a DIYer to make a particle accelerator. :) It would be cool though. It's certainly not something I'd do in the near term. Off the top of my head, it would involve at least a vacuum, which I don't have.
RimstarOrg Thanks, it is fine, I have been building one since last week and thanks a lot, your videos are one of the reasons that I got into physics (I am only 14 :p). But can you at least do a video on how to build a cathode-ray tube? Because my younger brother wants to build one for a science fair and is not understanding from all of the videos out there!
+Adam Fattal Why build a cathode-ray tube? That's a pretty tall order. You will need to have a phosphor screen, and envelope and an electron gun. Maybe not as hard as I thought. Again google provides: www.instructables.com/id/DIY-Electron-Accelerator-A-Cathode-Ray-Tube-in-a-/
Great tutorial, Just did one and happy that box of batteries collected over the year for disposal got second life. :) I use night lamps, to see the way in corridor :) Thank you!
hi i have a small question, I have another circuit that produce low millivolts (200mv - 700mv) with a very low current. Is it a good idea to use this circuit to power up loads or charge a rechargable battery? Any modifications are needed thank you very much and best regards ! :D
nice thanks,yeah it was me that made the comment your refering to lol. im just trying to find out as much as i can before i start buying stuff,thanks again man.
Yeah it works, kewl project thanks. I swapped out the NPN with a PNP, flipped the battery, flipped the LED. Nothing. Do I have a Dud transistor?, need to flip one of the windings, or just give up? Regards Bruce
Are you sure your battery is completely drained? For example, a battery may not have enough power to run a small motor but it may still be able to power a remote control for a TV. If all you do is test it with the motor then you'd think it was drained. But if it can't power even the lowest power device then I guess you'd call it drained. In that case you need another battery.
Sir it's not about using a device , its just about storing even a very small amount of energy even if its negligible and store enough and give an energy burst , so my intentions are to store even small amount of energy and use is like a burst just like the joule thief circuit, so do you have any alternative suggestions that might help me in me research.
And also could i use differant resistors, im not sure how the resistor if effecting the circuit im thinking knowing this will help me decide, my end goal is to make a high voltage power suply out of this, but im starting with just simply making this.
hello +RimstarOrg ! I need 3.5V for my project but the source is capable of giving only 2V.......... so will this joule thief be able to solve the issue?
+Siddhant Rai Probably not. It depends on what you're trying to power. The output of the joule thief is not flat DC, it's very spikey and pulsed, as you can see from the waveform I show on the oscilloscope near the beginning of this video. However, you can add further parts to make the output be DC like one fellow shows here on my website rimstar.org/science_electronics_projects/joule_thief_powering_clock_ls.htm. But the current will be very low, so it again depends on what you're trying to power. He's just powering a clock. Also, you'll have to play around with the part values to get your 3.5 volts.
Mine isn't working for some reason. I used a BC547 transistor because they are more common in my area, and when I shortcut the LEDs leads with the negative wire and let go, the led flashes so I think that the rest of the circuit works, the transistor just doesn't pulse like it should. What can I do?
PunakiviAddikti You might want to check the specifications for your transistor. I think the emitter leg and the collector leg might be the reverse of what my transistor is.If that's not it then the most common problem is getting the wiring to the coils wrong. They're not connected to the circuit in the way you'd expect.Lastly, try different resistance values. The higher the resistance, the dimmer the light. You shouldn't have to go over 100 ohms.
thanks, I'll try that. Does it matter how much wire is wrapped around the ferrite core? I have less than a meter of wire around it, and the wires are the same diameter.
PunakiviAddikti I don't have any feel or formulas for how many turns there should be wrapped around the ferrite core. You will get best results if cover the whole core though; that keeps the magnetic field inside the core as much as possible. Though it's probably good to play it safe and leave a small gap of core with no turns around it where the ends of the coil meet.
I got it to work, thanks! I actually used a pre-made toroid from some old electronics and changed the 1K resistor to the base of the transistor. I actually found one 2N3904 transistor and used it and it works great!
does a crystal radio make ac or dc. if it makes ac do you know of something like this that will work on ac. and do you know any stores i can walk in and find a germanium diode.
My guess is that you have the connections going to the coils reversed. Compare what you've done very carefully with the diagram in the video. If you don't have them wired correctly then your 9v battery is just powering the LED directly like normal. If that's not the problem then it could be that your resistor is too big. I used a 1 kilohm resistor. If it still doesn't work after you've checked the above then tell me which transistor you used, what the voltage rating of your LED is and how many turns you've put on the coil.
I'm using 2n3904 transistor in the same circuit but the output voltage is less than that of the input voltage. I can't seem to identify the problem what do you think?
Are you sure you have it wired up correctly? The wiring going to the coils is the opposite of what you'd expect and is where most people go wrong. If you use thinner wire then you don't have to increase the number of turns.
Will I like your Joule Thief it works great. I played around with it after I made it. I put a new 1 1/2 volt battery on it and powered a Blue LED for 7 days yes it went dim but it stayed on for that time. Your layout is the only one I found that work good. I learn a lot from your videos about them. Now I am going to try it with solar for a light. I am new to this so a lot of errors but figure it out. the only thing I did different was to use a 10K adjustable resistor in place of the 1K it work. thanks for the good information.
Everyone talks about using the long lead of the LED to denote polarity. BUT ... What if the LED is scavenged and the legs are both the same length? Do you know how to tell which is the negative and which is positive? Great video, by the way! Thanks for the information!
Thanks! It depends on the LED but there are often other differences. The 3V white light ones I have have a flat edge to the plastic base on the side with the short leg. But since I often can't remember which way even with the long and short legs, I usually just connect it to a battery with a suitably sized resistor and try it both ways. Once way lights up and the other doesn't.
So i made everything like in the diagram, and it didnt work, so i tried connecting the transformer differantly and one combonation worked when i diconnected the resistor and conected it again repeatedly i got the light to blink but thats the only way i got it to work, so i tried making a rind farrite core coil, and still cant make it work, iv tried 7 fidderant transistors( i have an old tv and its just full of great stuff) but iv got my circuit exactly the same as yours, tried differant l e ds
@@RimstarOrg alright! It worked very well I managed to get a 7.5 volt output from a single double A. Anyways I used the same circut with a dead battery and for some reason my multimeter told me 0.5 volts output even though it could power two LEDs. Whats up with that?
The voltage output is neither flat DC nor sinusoidal AC, meaning that your multimeter can only show you samples of the fluctuating voltage. You really need an oscilloscope to see the output voltage. I show what it looks like near the beginning of the video. If your multimeter has a min/max function then you could use that to get an idea of the maximum value. Regarding the dead battery, it's of course not dead and that's the point of this circuit. If you again look at the oscilloscope output which I showed, you'll see that the circuit produces voltage spikes and its just after those spikes that the LEDs are powered. They just appear to be powered all the time. I explain how the circuit works in this video th-cam.com/video/0GVLnyTdqkg/w-d-xo.html.
+Roush Hawley You can tell from the transistor symbol. If the arrow on the emitter is pointing toward the base then it is a Pointing iN Pointer (PNP). If the arrow is pointing away from the base then it is a Not Pointing iN (NPN) type. I've used that method to remember the difference for over 40 years ... and still have to say it to myself to remember. :)
Hi i recently made a joule theif and was wondering why a 1.27 volt battery didnt power my 3 volt LED, i used a 3 volt battery through the mechanism, and it lit the led, i used a multimeter to measure the output voltage of the entire circuit, but it was the exact same voltage of the 1.27 volt battery, did i mess something up on the torroid? Is there a specific amount of loops around the core? - Thanks
Scotlahn Mccallister The number of turns I put on the coil was randomly chosen. The most frequent error is in connecting to the coils in the wrong way. Recheck that since it's the opposite of what you'd expect. Another possibility is that the value of your resistor is too high. The lower the resistor value, the brighter the LED.
RimstarOrg Hi again, i tried what you told me to do, checking the connections, i changed the resistor value to 1 ohm and it still didn't work, i'm starting to wonder if my transistor is the correct type to use for this circuit, my transistor type is PN2222A6E. - Thanks
Scotlahn Mccallister PN2222A should work. I can't find anything that tells me what the 6E means. My guess is it should work. Try the same number of turns on your toroidal core that I use, around 13 turns.
RimstarOrg Okay, i checked those, i think it might just be my connections on my breadboard, and i see that the voltage may be traveling directly to the LED and not through the transistor first, i would like to see a picture of how you connected everything to the breadboard so i can check if i am correct or not. - Thanks Again
Scotlahn Mccallister I don't have any picture of how I did it. Try figuring it out from the video and let me know if there's any placement that you can't figure out.
Coils have some capacitance, but important here is that a magnetic field builds up. When the magnetic field has built up enough to saturate the ferrite core, the interaction between the coils shuts off the transistor fast. That causes the magnetic field to collapse quickly, building up more than enough voltage at the LED to turn it on.
There is an interaction between the two coils that controls the transistor to make this work well. I haven't thought if the same could be done with capacitors.
I'm delighted you're enjoying this world of electronics. It's hard to make a videos that'll keep the interest of someone totally new to it. Thanks for giving it a watch!
Hi Debasis, You can most likely use the core. I don't see why not. However, without knowing the part number of the transistor I can't say if it would work or not. I'd suggest you try it and see. Keep in mind that if it doesn't work right away you should first check the way you wired the coils into the circuit as I mention at 4:28 in the video. Also, some LEDs are not very bright (the mcd value) or shine in one direction mainly so look carefully at the LED to see if it's lit. Have fun! -Steve
The resistor protects the transistor from too much current. At the same time, the higher the resistance the dimmer the LED, so if you have a potentiometer (variable resistor) then you can adjust the brightness.
The voltage of the solar cell will vary just as the output of the joule thief varies. An LED lighting up for short bursts but very bright can have the same overall power output as an LED lighting up for longer bursts but less bright. A better approach would be to try charging a capacitor for a fixed period of time and then disconnecting and measuring the voltage of the capacitor. The energy accumulated in the capacitor is 1/2 x capacitance x voltage x voltage. This gives a result in joules.
Yup. Many people power compact fluorescent lightbulbs with it, and still with just a single 1.5 volt battery. But there is a limit to what it can power, both because of the available power and given the voltage waveform I showed in the oscilloscope output.
Transformers are for a number of things e.g. stepping a voltage up or down, electrically isolating parts of a circuit. If you transformer has 4 prongs then there are probably 2 coils, 2 prongs for each coil. Measure the resistance of the two coils. Apply voltage to the lower resistance coil and the voltage on the higher resistance coil should be increased. Note that this happens only when the voltage is changing. Straight DC will transform only when being turned on and when being turned off.
Did you try reversing the wires coming from the coils? That's usually the first thing to check. Also, if you were using a breadboard like I did, go back and check that the connections are all where they should be and that they are fulling in the holes, also very common issues. Also double check that you have the transistor legs right, they're not always the same for all transistors. Also double check the polarity of the LED, one leg is positive and one is negative, they're not interchangable.
I don't really know for sure. Ferrite seems to have a few meanings. The ferrite core here is a poor conductor and is a ceramic whereas ferrite also refers to iron. A ferrite core, being a poor conductor doesn't have as a much of an issue with eddy currents being induced in it as iron would. I guess you'd just have to try.
Ah, I see. So you have the output of the peltier module go into the joule thief. Good to hear that you're getting some result so far. Thanks for clarifying.
A crystal radio puts out fluctuating DC at the output. Someone suggested I try powering a joule thief using a crystal radio in the comments to one of my crystal radio videos (hmm... can't seem to find it) and that's why I made this. So far the radio's DC output is too small to turn on the transistor but I'm still trying.
My sources are small local electronics stores and Radio Shack (The Source in Canada) is the only more widespread store I can think of that might have germanium diodes.
mA is amps/current, not power. It's also extremely difficult to measure current for this type of spikey output so the 31.3mA is probably not accurate. But anyway... People who've played with the resistor value have seen it affect the LED brightness so trying different resistors or using a variable resistor (potentiometer) should help I think because it affects the timing. Decreasing the turns may work too since the power flows through the coil that's not connected to the resistor.
Ah, so that's what it was. I've made only very simple power supplies. There's a heck of a lot more than I'd have expected. Thanks for clarifying.
Happy to hear you like the video. Thanks.
The thing you got from a junk shop was a switching power supply. Full of neat stuff.
Another good video.
Ah! I knew it was similar to what happens in the ignition coil but I didn't know what caused the collapse. My turn to say thank you!
Yes. Over time it collects energy in one of the coil's magnetic fields and then releases it in a quick burst, which provides the voltage.
You're very welcome! I know how that feels, when something doesn't work out and you have to switch to something else. I'm happy to hear it worked out.
I haven't tried more than one. I guess more than one might work. The LEDs need current too so after adding some number of them you'd start to notice they're not as bright.
Good question. I'm guessing it because the voltage voltage is there for only a very brief time and then it's off again until the next voltage spike.
Yeah, without an oscilloscope, if your battery voltage is high enough to light the LED anyway, then it's hard to know if the joule thief is doing it's stuff. I once lit an LED with the joule thief circuit before only to find with my oscilloscope that the joule thief wasn't working at all.
Sure, if you can find them. Some electronics stores might sell ferrite cores. Usually the ones I see in the stores are pretty small, but then again, we don't have really good electronics stores here. I just had a quick look at ebay.com and saw some there.
I don't know. I never tried one. I'm talking really tiny cores, ones that are for threading a single wire through, so the ones I have in mind wouldn't work anyway because you couldn't wind the coils on them.But it wouldn't hurt to go see what's available in the stores near you.
Try reversing the polarity of the battery and the LED. i.e. Put them in backwards to the way shown in the circuit diagram.
Also, the insulation of the coils may not be able to handle the high voltage. In that case the insulation will break down and the transformer will be ruined. You need to know the voltage rating of your transformer coils, or take a chance. :)
This video is way better than the other jtc video . Worth a sub
You cannot rely on the face of a transistor for a correct pinout guide. You should ALWAYS look at the datasheet for the part number of the transistor.
Also one should consider the wire gauge and amount of actual turns on the core, this is important and will make a difference between running one Red LED or several White LEDs. I have built these things using a coil wrapping design for a power transformer (primary coil and secondary coil) and I had success running small DC fans for vehicles and even a car stereo off two AA batteries. Of course you want to use transistors rated for higher volts and currents and heat sink them properly.
So, after many hours, i threw out all the peices,and got new ones, finnally got it working, must have been something broken, sorry for being so hopeless, great advice, great vids, keep up the good work and thank you!!
Thanks. I think we've all been in that situation at one time or another. I'm glad to hear you got it working.
I don't know. The magnet's magnetic field might be too strong, but then again maybe not. Maybe give it a try and let us know.
so this is the single one that worked, i had reproduced other 3 "youtube diagrams" and this is the working one. the diference being the resistance placement.thnx man
I couldn't find anything on how long. The one's I've seen remove all the electronics from the CFL, so you're powering the fluorescent tubes directly.
Not a 1.5v battery. A car's battery is 12 volts but that then goes through an ignition coil to produce a much higher DC voltage to produce the spark in the combustion chamber. I'm no car expert though.
I'm no expert on measuring these odd waveforms. The only way I know to measure voltage like this is to use an oscilloscope and even there I've seen mention of possible errors - but it's close enough.
The coil winding really is as simple as the little I showed. You can find a diagram in my "How a Joule Thief Works" video. There's a link to it in the description below this video and you can find it on my channel page. The wire sizes don't matter much, just make sure they have insulation on them i.e. not bare wires.
I'm not an expert on transformers. I just know a bunch of stuff about them from research for projects. But go ahead and ask. If I don't know the answer, maybe someone else will. And if it's too off-topic from joule thief then we can try private messages.
+Razia Binate Saleh I've looked in stores and the only one I've ever found on in was a local electronics store (Gervais Electronics if you're in Ottawa, Canada.) You could try science stores like the ones attached to a museum or something like that.
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".
You're welcome. Your project sounds interesting. How are you using a joule thief and a peltier cooler together?
Thank you. Sort of like a set of points collapsing the field to an ignition coil. but instead of the primary field collapsing mechanically, the joule uses the transistor triggered by the saturation point of the ferrite core.
That's cool. As long as it works. I know others have used transformers too.
Hey Rimstar. Great info on the JT. I'm assuming "an 820 kilo ohm resistor" as shown in your description should be 820 OHM or 1 Kilo ohm ???
Yup. I just fixed it. Thanks! 820 ohm or 1kohm are both fine, though having a potentiometer to try different values will allow you to find an optimal one for brightness.
Wow! Congratulations! Well done.
Cool. What voltage do your leds normally require?
Good on you to keep at it until it worked! Did you light an LED with the peltier powered joule thief? You should upload a video - it'd be great to see.
I got a ferrite core from a powered landline phone, usually where the phone line socket thing is.
Thanks Tom. As long as that's the case, I'll keep making them.
I have been thinking or replacing the 1000 ohm resistor with an avalanche diode. I hope it would cause a voltage drop while making a jump in current. Votage may not matter as the led only needs 3 or 4 volts.
New RimStar Video...woot woot! Holy hell he's like freaking MacGyver, he can rig up anything! I liked the start too with the zombie batteries! Always very interesting how you can make all sorts of appliances and other electronics by jimmy rigging up some wires and batteries! I need to get my head into the books more dude, you are so smart! Liked and will share!
Just made a Joule thief myself and it works very nice :) I winded the coils with different length on the ferrite core (Longer for the LED and collector) it can light up a 3W 3,7V powerled with only one AAA battery that is as low 0,7V (0,5V under load) even if just a tiny bit of light. a battery with 1,2V (0,6-0,7V load) is enough to get a pretty strong light from that led. running 2 parallell depleted AA batteries gives even more light :)
thanks for a good video sir
Pulverrostmannen Nice. Thanks for letting me know. I'm glad it helped.
RimstarOrg Hi it's me again :) when I put in those 2 AA batteries that had 1,4 volts in them has been lit 24/7 and enough to light up the whole room in the night ever since I posted that last comment. I never thought it would run for so long before the batteries would die. they still give light but now so weak the led just glows. really decent light for 8 days and weaker the last from already weak batteries. sick
I made one version with a 10K potentiometer to set the brightness, and with half of the brightness the led lasts extremely long (I'm using a diffused blue led that shines bright enough even with half the normal brightness). With full 10K of resistance it shines a lot dimmer but since that is under your control, it can be very efficient.
PunakiviAddikti indeed, the pot I use is a 0 to 22K. and it is possible to just make the led glow or get super bright.
I actually forgot to mention that I still have a resistor connected in series with the potentiometer: A transistor needs some resistance or it can get really hot.
I'm using the peltier+joule thief to turn heat into electricity that can be collected from readily available sources like hot coffee cups, the sun heated roof of a car ect., i think they are also called seebeck generators, they are very simple devices, i wanted to make my own after seeing a camping water kettle with a peltier built in.
Your transistor is an NPN, it seems. How should the circuit be modified to work with a PNP?
Hello sir, what is the Hertz range on the little Toroid? I've seen those around, but most of the time their rated at 10-50 MHz.
Oh, hadn't thought of night lights. That would work.
Thank you for making videos like this, im happy to see that there is still a corner of the Internet thats actually usefull & worth watching in the era of reality television, jersey snore, & honey boo boo! Thank you from the bittom of my heart for giving us something worth watching! That being said i just assembled this very same joule thief circuit and im testing it with a $10 peltier cooler i bought at amazon, so far its kinda so so, i need to adjust it more.
RimstarOrg Hello Rimstar, Can you please do a video on how to build a particle accelerator?
Can you do a Bedini Magent Motor? Every design i've tried, Failed miserably.
I imagine I'd have the same experience.
So no desire to even try? Be nice if someone could show me what i've been doing wrong...
Oh, I did at least look at it and even attend a workshop a number of years ago, while attending a fair for something else. No one there, even the long time builders had succeeded so I didn't see any point in pursuing it further myself.
get all the transistors out of and google the parts number. get a general use or switching type "NPN".
I know the transistors from a camera flash circuit work. but not nearly as well as something like the 2n2222 or 2n4401, etc.
Can you try finding how to make a joule thief on a miniscule scale? Like maybe 0.5 volts to 3 volts?
Can we over lap the second wiring over the first in coiling
I think so. Just make sure to keep winding in the same direction around the core material.
I cant figure out pm's, so i guess my main question is, is the main point of the transformer to have a current flowing one way and then a current flowing the other way? I have a very simple 4 prong transformer, actually i have a few differant ones with all with 4 prongs, and also im having trouble with resistors because im not sure what one to use, i am working with 2 aa's and have 5 1kohm resistors, and 4 smallish transistors, the idea is making a very simple high voltage driver
Thanks very much on the tutorial, my slayer exciter didn't work for te upcoming science fair on 6-th of April, so i remembered on your joule thief. I suceeded to make it on the first try!Thank you veeery much!
we're do you go to get you're magnet wire and you're diods and resisters
+Andrew Kimble I get my parts from local electronics stores, from taking apart electronic devices (microwave ovens that people have thrown out are great, but so are a lot of other things), and less frequently, from searching online.
Ah, dead batteries. Now it's cool. :)
can i use a mosfet ? because i only have 2 pairs of irfp250s right now,or i have to buy a power transistor
i guess i'll try itmyself,with irfp250,flyback transformer core ,and some 22 awg copper wire
damn VGS of this mosfet is 2-4v
A 2N2222 General Purpose bi polar transistor is the one recommended Though you can look into other higher powered types like those used in power supplies. Also you do not have to use an NPN type, you can use a PNP type with slight modification to the circuit.
ElfNet Gaming If I recall correctly, for PNP you just have to reverse the + and - of the battery.
yeah pretty much. You can use this circuit to run LEDs off a lemon or potato also if you dont have a battery
Hello, i was woundering if i could get some help with a small transformer i am trying to wire for a project im working on?
24 volts huh? Question. Could you use the 1.5 v battery to start or jump a dead car battery? If not then...Could you use the car battery and the joule thief circuit to start the car?
Hello Rimstar, Can you please do a video on how to build a particle accelerator?
Adam Fattal I don't even know if it's possible for a DIYer to make a particle accelerator. :) It would be cool though. It's certainly not something I'd do in the near term. Off the top of my head, it would involve at least a vacuum, which I don't have.
+Adam Fattal Just Google diy particle accelerator. There's lots of them out there.
RimstarOrg Thanks, it is fine, I have been building one since last week and thanks a lot, your videos are one of the reasons that I got into physics (I am only 14 :p). But can you at least do a video on how to build a cathode-ray tube? Because my younger brother wants to build one for a science fair and is not understanding from all of the videos out there!
+Adam Fattal Why build a cathode-ray tube? That's a pretty tall order. You will need to have a phosphor screen, and envelope and an electron gun.
Maybe not as hard as I thought. Again google provides: www.instructables.com/id/DIY-Electron-Accelerator-A-Cathode-Ray-Tube-in-a-/
Robert Shaver I saw that one but my brother didn't understand how to do it and I can't help him because I am too busy!
Great tutorial, Just did one and happy that box of batteries collected over the year for disposal got second life. :)
I use night lamps, to see the way in corridor :)
Thank you!
hi i have a small question,
I have another circuit that produce low millivolts (200mv - 700mv) with a very low current. Is it a good idea to use this circuit to power up loads or charge a rechargable battery? Any modifications are needed
thank you very much and best regards ! :D
nice thanks,yeah it was me that made the comment your refering to lol.
im just trying to find out as much as i can before i start buying stuff,thanks again man.
Yeah it works, kewl project thanks. I swapped out the NPN with a PNP, flipped the battery, flipped the LED. Nothing. Do I have a Dud transistor?, need to flip one of the windings, or just give up? Regards Bruce
It sounds like you did just what you should. There's nothing on the windings to flip. It could be the transistor.
So i went and got a pack of resistors and a new transistor, how do i know how much resistance/current my transistor needs?
Im looking at the data sheet, but unfortunately i dont quite understand it, the 'conditions' are confusing.
Hello, can we use the same battery if it is completely drained to store very small amount of current with this joule thief circuit.
I imagine at some point a battery can be so drained that it won't even power a joule thief circuit.
so what alternate can i use , because there is a project im working on and i need a similar process there.
Are you sure your battery is completely drained? For example, a battery may not have enough power to run a small motor but it may still be able to power a remote control for a TV. If all you do is test it with the motor then you'd think it was drained. But if it can't power even the lowest power device then I guess you'd call it drained. In that case you need another battery.
Sir it's not about using a device , its just about storing even a very small amount of energy even if its negligible and store enough and give an energy burst , so my intentions are to store even small amount of energy and use is like a burst just like the joule thief circuit, so do you have any alternative suggestions that might help me in me research.
In that case maybe what you should look into is using a capacitor to store the energy.
Will it work with any other diameter of wire( around ferrite core)?
Yes, thinner or thicker wire should work too, as long as you can get enough turns of wire on it.
Thanks a lot
+RimstarOrg it work with bf869 transistor?
And also could i use differant resistors, im not sure how the resistor if effecting the circuit im thinking knowing this will help me decide, my end goal is to make a high voltage power suply out of this, but im starting with just simply making this.
hello +RimstarOrg ! I need 3.5V for my project but the source is capable of giving only 2V.......... so will this joule thief be able to solve the issue?
+Siddhant Rai Probably not. It depends on what you're trying to power. The output of the joule thief is not flat DC, it's very spikey and pulsed, as you can see from the waveform I show on the oscilloscope near the beginning of this video.
However, you can add further parts to make the output be DC like one fellow shows here on my website rimstar.org/science_electronics_projects/joule_thief_powering_clock_ls.htm. But the current will be very low, so it again depends on what you're trying to power. He's just powering a clock. Also, you'll have to play around with the part values to get your 3.5 volts.
+RimstarOrg thanks a lot !!!!
why not just use a large filter cap?
My transistors base is rated at 32v max open and emittor to base voltage is 5v and base current is 1 amp, how do i figure what resistor to use?
Can I use torus shaped magnets for the ferrite core?
where can you find a thermometric generator in stores
Mine isn't working for some reason. I used a BC547 transistor because they are more common in my area, and when I shortcut the LEDs leads with the negative wire and let go, the led flashes so I think that the rest of the circuit works, the transistor just doesn't pulse like it should. What can I do?
PunakiviAddikti You might want to check the specifications for your transistor. I think the emitter leg and the collector leg might be the reverse of what my transistor is.If that's not it then the most common problem is getting the wiring to the coils wrong. They're not connected to the circuit in the way you'd expect.Lastly, try different resistance values. The higher the resistance, the dimmer the light. You shouldn't have to go over 100 ohms.
thanks, I'll try that. Does it matter how much wire is wrapped around the ferrite core? I have less than a meter of wire around it, and the wires are the same diameter.
PunakiviAddikti I don't have any feel or formulas for how many turns there should be wrapped around the ferrite core. You will get best results if cover the whole core though; that keeps the magnetic field inside the core as much as possible. Though it's probably good to play it safe and leave a small gap of core with no turns around it where the ends of the coil meet.
I got it to work, thanks! I actually used a pre-made toroid from some old electronics and changed the 1K resistor to the base of the transistor. I actually found one 2N3904 transistor and used it and it works great!
PunakiviAddikti Glad to hear it!
does a crystal radio make ac or dc.
if it makes ac do you know of something like this that will work on ac.
and do you know any stores i can walk in and find a germanium diode.
So then a coil has has capacitance ? would not a capacitor be a better storage.
Hi I working on a joule thief but it's not working right .It works on my 9v battery but not on a AA battery ? I think I didn't do sum thing right
My guess is that you have the connections going to the coils reversed. Compare what you've done very carefully with the diagram in the video. If you don't have them wired correctly then your 9v battery is just powering the LED directly like normal.
If that's not the problem then it could be that your resistor is too big. I used a 1 kilohm resistor.
If it still doesn't work after you've checked the above then tell me which transistor you used, what the voltage rating of your LED is and how many turns you've put on the coil.
can a ferite rod be used instead of the core?
I'm using 2n3904 transistor in the same circuit but the output voltage is less than that of the input voltage. I can't seem to identify the problem what do you think?
If I use thinner wire should I increase the number of turns?
Are you sure you have it wired up correctly? The wiring going to the coils is the opposite of what you'd expect and is where most people go wrong.
If you use thinner wire then you don't have to increase the number of turns.
Yes the wiring is correct
well did a small ferrite core works or not
Does it matter if you use the same gauge wire for the coil?
No, it doesn't matter. You can use the same or different gauge wires.
does it have to be ferrite or can I use other metals?
Will I like your Joule Thief it works great. I played around with it after I made it. I put a new 1 1/2 volt battery on it and powered a Blue LED for 7 days yes it went dim but it stayed on for that time. Your layout is the only one I found that work good. I learn a lot from your videos about them. Now I am going to try it with solar for a light. I am new to this so a lot of errors but figure it out. the only thing I did different was to use a 10K adjustable resistor in place of the 1K it work. thanks for the good information.
Everyone talks about using the long lead of the LED to denote polarity. BUT ... What if the LED is scavenged and the legs are both the same length? Do you know how to tell which is the negative and which is positive?
Great video, by the way! Thanks for the information!
Thanks!
It depends on the LED but there are often other differences. The 3V white light ones I have have a flat edge to the plastic base on the side with the short leg. But since I often can't remember which way even with the long and short legs, I usually just connect it to a battery with a suitably sized resistor and try it both ways. Once way lights up and the other doesn't.
was that the smallest toroid you could find? (-;
Smaller toroids make the induction harder and thus makes the led dimmer so that is a pretty good size to use
So i made everything like in the diagram, and it didnt work, so i tried connecting the transformer differantly and one combonation worked when i diconnected the resistor and conected it again repeatedly i got the light to blink but thats the only way i got it to work, so i tried making a rind farrite core coil, and still cant make it work, iv tried 7 fidderant transistors( i have an old tv and its just full of great stuff) but iv got my circuit exactly the same as yours, tried differant l e ds
can i buy the ferrite core in the other shops
The only ferrite core i have is about an inch in diameter, will it still work?
I think it might. I'd put enough turns of the wire to cover the core.
@@RimstarOrg alright! It worked very well I managed to get a 7.5 volt output from a single double A. Anyways I used the same circut with a dead battery and for some reason my multimeter told me 0.5 volts output even though it could power two LEDs. Whats up with that?
The voltage output is neither flat DC nor sinusoidal AC, meaning that your multimeter can only show you samples of the fluctuating voltage. You really need an oscilloscope to see the output voltage. I show what it looks like near the beginning of the video. If your multimeter has a min/max function then you could use that to get an idea of the maximum value.
Regarding the dead battery, it's of course not dead and that's the point of this circuit. If you again look at the oscilloscope output which I showed, you'll see that the circuit produces voltage spikes and its just after those spikes that the LEDs are powered. They just appear to be powered all the time. I explain how the circuit works in this video th-cam.com/video/0GVLnyTdqkg/w-d-xo.html.
Excellent video! Don't know about the zombie noises though....
Thanks! I was trying to make them sound like Minecraft zombies. Beware of zombie batteries in the night :).
is it a PNP or an NPN transistor?
Roush Hawley It's NPN.
+Roush Hawley You can tell from the transistor symbol. If the arrow on the emitter is pointing toward the base then it is a Pointing iN Pointer (PNP). If the arrow is pointing away from the base then it is a Not Pointing iN (NPN) type. I've used that method to remember the difference for over 40 years ... and still have to say it to myself to remember. :)
Hi i recently made a joule theif and was wondering why a 1.27 volt battery didnt power my 3 volt LED, i used a 3 volt battery through the mechanism, and it lit the led, i used a multimeter to measure the output voltage of the entire circuit, but it was the exact same voltage of the 1.27 volt battery, did i mess something up on the torroid? Is there a specific amount of loops around the core? - Thanks
Scotlahn Mccallister The number of turns I put on the coil was randomly chosen. The most frequent error is in connecting to the coils in the wrong way. Recheck that since it's the opposite of what you'd expect. Another possibility is that the value of your resistor is too high. The lower the resistor value, the brighter the LED.
RimstarOrg Hi again, i tried what you told me to do, checking the connections, i changed the resistor value to 1 ohm and it still didn't work, i'm starting to wonder if my transistor is the correct type to use for this circuit, my transistor type is PN2222A6E.
- Thanks
Scotlahn Mccallister PN2222A should work. I can't find anything that tells me what the 6E means. My guess is it should work. Try the same number of turns on your toroidal core that I use, around 13 turns.
RimstarOrg Okay, i checked those, i think it might just be my connections on my breadboard, and i see that the voltage may be traveling directly to the LED and not through the transistor first, i would like to see a picture of how you connected everything to the breadboard so i can check if i am correct or not.
- Thanks Again
Scotlahn Mccallister
I don't have any picture of how I did it. Try figuring it out from the video and let me know if there's any placement that you can't figure out.
Cana a troide lose it magnitism if it is droped? I droped it multiple times.
i put 2 led's in series on a 9v battery, this circuit is amazing!