These videos remind me of the days I learned "electronics" by reading Popular Electronics. The magazine always had very interesting and fun projects one could build but, more importantly, they explained in easy terms how the components and circuits functioned. Good job.
That answers my unasked question. It works because it's only on half the time, but persistence of vision fills in the blanks. I have a small jar of toroids that I haven't known what to do with and this looks like a fun project. And just to brag a bit; today I managed to get a crystal radio to light up an LED. It was flickery and weak but, hey, I figured out how to grab electrical energy out of the air and make light light with it. Thanks for the uploads.
Sir, you explained the concept beautifully, I had no idea how it worked, it seemed very confusing to understand when I read it, but now it feels so easy, thank you.
Super vid :D This is the best one I've seen on how a JT works & is right.I've seen in other places where they give a much harder/wrong explanation.And just for anyone else who reads this know that a joule thief/ringer is one of the simple(best to me)circuits there are to turn on ANY TRANSFORMER.So hook this circuit up to Any transformer that has a center tap (or just rap one on the core if possible)& use the secondary for what ever you like(minus the LED of course to give energy to secondary) ;p
*best explanation of basic electronics by far, electricity can be very hard to understand, at least for me, i need analogies, and every single video on youtube does not use analogies, and it doesnt matter how good you are at teaching, to me explaining electricity without analogies is like trying to tell me how witchcraft works.*
Thank you for letting me know that it was clear. I know I sometimes speak too fast for some to understand. I'm happy to hear you had no problem. Cheers from Canada.
Thanks! And welcome! The time frame depends on a few things and is usually controllable by replacing the fixed value resistor with a variable one (a potentiometer). In my "Make a Joule Thief for Zombie Batteries" video at 0:36 I show on the oscilloscope a cycle time of around 40 microseconds, so a frequency of around 25 kilohertz. In my "Fun with Joule Thief Powering a Compact Fluorescent Light" video i show 60 microseconds. So somewhere in the tens of microseconds of tens of kilohertz.
I'm glad to hear you like the explanation. I'm actually enabled for longer than 15 minutes per video but I'm always wary of going on too long for fear of only a few watching. I have done 10 to 15 minute videos though. I do regret not talking about some of the other reasons a joule thief does the reversal step - bad decision on my part :(. Oh well. Thanks for the feedback!
Typically a resistance of roughly around 1000 to 2000 ohms is used for this circuit. I happened to have a 820 ohm resistor so I used that. The resistor is to protect the transistor from too much current but it also affects the timing of the cycles. A good idea is to use a potentiometer (variable resistor) instead so you can easily try different resistances.
my husband studies this , and since I watch your videos , I understand so much more about led and stuff , he also built a cnc machine that cuts stuff out , I love your videos, thank you :) Amelia
this explanation helped me a lot. thanks ..... this is the first time i understood a joule thief circuit completely..... i seen lot of videos, but only this one could be understood
This very video got me into electronics... Now i make great money fixing all sorts of electronics.. THANK YOU SO VERY MUCH FOR THIS EXPLANATION VIDEO.. IT HAS CHANGED MY LIFE SO MUCH.. ACTUALLY IT HAS SAVED MY LIFE AS NO DOUBT HEROIN WOULD PROBABLY OF KILLED ME OR LEFT MY LIFE SHORTER.. YOU HAVE SAVED ME. THANK YOU THANK YOU THANK YOU THANK YOU. I CAN'T SAY ENOUGH TO LET YOU KNOW HOW UNBELIEVABLY GREAT FULL I AM TO OF FOUND THIS VIDEO WHEN I DID BECAUSE WHEN I SAY MY LIFE WAS AT ITS END IM MEAN THAT 100% 😢😊👍🇮🇪🙏🙏🙏🙏🙏🙏🙏🙏🙏🙏🙏
Thank you for this video. It is well made and understandable. It has opened my eyes to new ways of thinking. I've probably heard dozens if not hundreds of engineers speak about circuits, and none showed such a neat concept. Thanks.
This is one of your simpler videos but normally I have to speed up videos because they bore me. You sir, almost 100% of the time I have to slow down your videos and I definitely learn a lot from your channel. Thank you very much sir! Godspeed.
Thanks! I don't have any garden lights to play with and the circuits I just saw online either had more parts than a joule thief or used a chip as one of the parts, meaning the parts were combined into a chip. Interesting that one of the descriptions I read showed that they were doing pretty much what a joule thief does, though with slightly different circuits.
Thank you sir Of all other videos I have seen everyone gave a vague explanation about the switching off of the transistor but your explanation seems quiet convincing .
I had the same problem. I was still trying to process and understand what he was saying, and then visualize it, but before I could, he was already on to the next step so I was trying to listen to what he said and not miss anything. For anyone else that couldn't keep up on the first play through, try playing the video at 75% speed. It slows it down enough that you can work through it. It also helps to break it into sections, let him explain one part, pause, think it through, make sure you understand what he's saying and why it works that way, then move onto the next section. Once you do that, it becomes a pretty simple concept to understand.
Yup. I agree they're pretty much the same. Build up energy in coil, then dump it in a burst. One added point about the joule thief is that the way the primary coil interacts with the secondary coil (green in this video) opens up and shuts down the transistor faster than it would without the secondary coil.
+Vignobles Lac Saint-Jean I've used it for a number of things, which I've shown in other videos. The most frequent use is to power an LED using a battery whose voltage would otherwise be too low to meet the LED's minimum required voltage. I show that one in my "Make a Joule Thief for Zombie Batteries" video ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-B61DU7yEsPM.html Then I found you can power a CFL using 2 AA batteries by modifying the coils ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-FkLET8MhRbU.html. And then I learned how to modify the coils again to use this to transmit electricity wirelessly ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-31Rxi8JMIys.html. Don't be misled by the "thief" in the name. This is just what's also called a blocking oscillator circuit. It provides higher voltage and current by repeatedly building up energy in a magnetic field and then releasing it in a short burst. The higher voltage and current exist only during the brief burst. The total energy out is smaller than the energy in. 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 the power to charge the batteries comes from the solar panels. It does make sense. Get a small low voltage panel from a calculator and try your joule thief on it with an led that shouldn't be powered.
adolthitler Ah, my bad. It was a case of skimming through comments at a time when I was going through a ton of them. I remember thinking he was talking about doing it at night when there was nothing from the solar panels, which clearly he wasn't saying at all! Yes, it could probably extend the useful time of solar panels to when the clouds some out. That'd be an interesting demonstration in fact.
Rohan Zener If you're referring to just the coils and the toroid core then it's one-to-one, no transforming would take place since both coils have the same number of turns. If you're talking about he whole joule thief circuit being x2, then no, it's much more than that. Also, the output is neither DC nor AC. You can see the output waveform in my video about how to make it at 0:38 ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-B61DU7yEsPM.html. Here a 1.5 volt battery is on the input and the output is the waveform shown with a peak-to-peak voltage of around 24 volts and a frequency of around 22 kilohertz. Both the voltage and frequency are affected by a number of things, including the resistance of the resistor used, as I demonstrate in this other video here ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-yz_99oVMbSI.html. I don't have any way of calculating what you should get as output.
Thanks and you're welcome! I don't know how efficient the joule thief is. It's basically just a way of stepping up voltage. It's definitely making good use of batteries that are normally considered dead, so in that way it's efficient. I looked at the store bought solar light circuit and IIRC the one I examined was basically a joule thief circuit.
OMG ! Sorry to say that but the term « joule thief » is misleading ! Nothing gained here. This is a very old concept called a « multivibrator circuit » or « oscillator circuit » that is fed through a toroïdal transformer to step up voltage (not current). You could use a regular transformer 110v:6v reversed and you'd get about 20-25 volts. You just made an inverter ! Actually you could do the same using a single SPST relay and a 9v battery hooked up as a vibrator; in that case, you might get thousands of volts on the coil as Back EMF. Usually, to counter that effect. we put a diode across the coil because back EMF can be damaging to electronic components. In this case, we don't need the diode because WE WANT the effect to create EMF. The « Noise » you are hearing in the coil is normal ! It's the vibrator circuit creating resonance !
Yes, it is misleading. But that's one of the names very commonly used for this circuit. Blocking oscillator is another. But nothing in my description points to any energy gain. Just accumulated energy being released in a burst and repeated.
RimstarOrg Blocking oscillator..makes more sense... so does this circut drain the battery faster due to it needing to build up a charge to light the LED? thus no real extra power or "thieving" is achieved? it just spends the energy in a more lump some effect?
Ohm Zen If you have a battery with sufficient voltage and current to light the LED then I would think you'd be better off using the battery directly. And yes, to your other two questions.
The Transistor in a joule thief, just acts as an automatic on/off switch, while the 2 coils act sort of like a rechargeable battery. So the magic isn't really the Transistor, the MAGIC actually happens with those 2 coiled wires. So, it starts out with a (street Traffic) Red Light. The wire coils, act like traffic getting backed up at a red light (storing the extra energy in a magnetic field). The Transistor is the traffic light, which waits for enough cars waiting in line, before turning to a green light, which releases the cars, while also blocking traffic from other directions. The Transistor does not really amplify energy, instead, the wire coils amplify the energy and the transistors only job is to be forced open from excessive electricity built up, which quickly drains, which forces it back closed. Without the transistor, the Gate/Switch/Traffic-Light would have to be MANUALLY toggled by a human (like jiggling the second wire on/off the negative). I made one without a transistor, and successfully lit an LED with 1 AAA battery. This is a horrible picture, but I didn't expect it to work, so, forgive me for the bad image quality and having nothing labeled. i.imgur.com/NOknJv1.jpg
Yep, this is the fundamental part, and basic principle with any switching converter. But that's, however, quite easy to understand, but the hardest part to understand with blocking oscillator is how it actually makes the oscillation. This video got me closer to get a grasp about that, still have to process it in my brain though.
The resistor is there to protect the transistor along the base to emitter path from too much current. Resistors reduce current flowing through them. These coils don't really serve the purpose of a normal step-up or step-down transformer but the primary is considered to be the red coil in the video, the one going through the collector and emitter path and to the LED. The secondary is the green coil, the one going through the base and emitter path.
+Kennynva T. If you're reading the same voltage then it's possible your circuit isn't right and you're just powering the LED off of the battery. Does you battery light the LED without the circuit? If the battery alone doesn't power the LED then you wired the circuit correctly and the joule thief is doing it's thing. Check the wiring going to your coil to see if it's the same as the circuit diagram. It's not wired the way you'd expect.
Yes that is whats weird..the battery will not light the led by itself..but it does in the circuit..So how does the capacitor make the led brighter...no matter which way you turn the electrolytic capacitor?
+Kennynva T. That's not weird. That's what's supposed to be the case. It means your circuit is working properly. This video is supposed to explain how, but basically it works by increasing both the voltage and the current at the LED by simply building up energy in the coil's magnetic field with the LED off, and then dumping that energy to briefly turn on the LED, and then repeating it. I haven't given much thought about the electrolytic capacitor, but it's also an energy storage medium so maybe the answer's in that. Regarding the polarity not mattering, that would be the case if it's in the circuit where there's AC, as with AC it would function in at least one orientation, and luckily not be damaged.
It's hard to say if the light flashes. The frequency is too high for a human to see. We really need a high speed camera to tell or a photo diode circuit. It's pure speculation on my part but there might also be some delay in an LED turning off that causes it to still be on when the current flows again.
the joule thief does two things almost simultaneously 1) it creates an electromagnet 2) it creates voltage with the electromagnet. normally when you create voltage with magnet and wire, you are moving the magnet by the wire, or moving the wire by the magnet. The rate of change has a significant impact on how much voltage is produced (how fast the movement happens) - the force that moves the magnet, is the same pressure that moves the electrons (and that is the voltage). now imagine instead of moving the magnet or wire, you could simply make the magnet appear. Thats a big change - no magnet, then lots of magnet. If appears very slowly or casually, you might not get much voltage, cause it's not a big rate of change, sort of like blowing up a balloon puff by puff. things move baby step by baby step. lots of tiny bits of force that add up to something. But, what if you could also make the magnet disappear? that's also a big change - lots of magnet to no magnet. If it disappeared suddenly - like if you popped the balloon (!) that would be a BIG change in a small amount of time, with a lot of force behind it (all the little bits of force from blowing up the balloon are all let out suddenly in a single instant). things would REALLY move, fast (namely electrons). unfortunately magnets can't disappear and appear. BUT electromagnets can turn on and off, which makes a magnetic field that appears and disappears. Luckily, when it shuts off, the magnetic field collapses much more suddenly than it took to inflate with electricity. When it collapses there is an explosion of electrons that shoot through the wire - imagine all the air from an exploding balloon shooting through a drinking straw. things move with the same amount of energy that was put into the balloon, but it is more forceful because its all happening at once (greater acceleration, covering the same distance shorter amount of time) so the transitor allows the joule thief to switch between too phases. phase one creates an electromagnet using the voltage of the battery. when the electromagnet appears it also creates a tiny bit of extra voltage which completes phase one and switches the transitor to phase two. in phase two the electromagnetic balloon POPS condensing all the force put into it into a small point in time creating a tremendous pressure which is strong enough to light the LED. ultimately the joule thief uses time as its currency to buy more power. the joule thief relies on self inductance. here is a simple video on self inductance m.ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-pKKsco9EgBQ.html the only difference between the device in the video and a joule thief, is the transistor replaces the person that has to repeat the step of pressing and releasing the button.
Yes, at a conceptual level in that it steps up. Though it steps up both the voltage and the current, sort of, by accumulating energy in the magnetic field over a period of time, and then dumping it all in a brief period of time, and then repeating. So functionally it's different than a transformer.
The transistor is off when the voltage is sufficient to run current through the LED. The opposite wiring of the coils makes sure of that. During the collapse of the magnetic field, the red coil induces current in the green one in a direction that further aids in shutting off the transistor.
Oh, sorry to hear you had difficulty. The anode/positive/long lead is connected to the collector of the transistor which is also connected to the red wire going to the red coil. The cathode/negative/short lead is connected to the emitter and battery negative. Maybe your confusion is because I drew the long lead connected directly to the transistor's collector without any intermediate wire. Maybe that made you think those weren't the leads and that they were missing?
I was using closed and open in the context of the water flow gate analogy I use in the video. When the water flow gate is closed, the water doesn't flow. When open, it flows. I know it's the opposite usage when talking about an electrical switch but it's only misleading when you read the quotes in the comment. When hearing the usage while watching the video, there should be no confusion. In fact, I hadn't even thought about the conflict until you brought it up, so it's not an issue (I hope :)).
With some changes, like a bigger battery and more turns on the primary and a higher power transistor, you can power the tubes of a compact fluorescent lightbulb.
The first thing is go back and check all you connections to see if they're good and if they're right. That's especially the case for the coils since they have to be connected backwards from what you'd expect. Another thing is to try an LED with a higher mcd rating. The higher the mcd, the brighter it lights up. Of course it's a trade off with the LED's voltage requirements. I find 1,400 mcd works great. Also, LEDs don't always emit light in all directions so look at it from various directions.
Nice play on words, but I love these videos,,electrical engineering is a bit beyond me, but I love the projects and lessons, understandable , get a better understanding
If I recall correctly, yes, lower voltage gives lower frequency. If the induction coil is continuously feeding power to the LED then it won't be the same. The joule thief builds up energy over time in the coils/core and then releases it in a burst at high voltage and current. Then it builds up again and releases in a burst again. With continuous feeding to the LED, the energy will be too low at all times.
I don't know what the maximum voltage would be but a lot of people run compact fluorescent light bulbs off of them and they require high voltage. Their ferrite cores are usually larger and they use higher power transistors.
Yes, it's the same type of toroid, just different coils/wires wrapped around it. But it's easy to remove the wires and wrap your own around it. I removed my toroid from an old power supply, removed the wires and then wrapped my own. You can see me wrapping them in my other joule thief video "Make a Joule Thief for Zombie Batteries" which you can find a link to in the description below this video. And I agree, it is amazing and beautiful and delightful when you see a circuit you made working.
@ RimstarOrg: Sir, I have now viewed two of your electronics presentations (Crystal Radio & now this one). Technically, you’re a talented man who has put together presentations for the “lay person.” Also, your very pleasant speaking voice brings everything together. Thank you.
It's not cycling based on a resonant frequency. In the video I talk about the cycle reversing because the core becomes saturated. But I mention that as one of a few possible reasons depending on things like the battery voltage. At these low voltages, most likely the cycle reverses when the current between base and emitter is too low for the increasing current between the collector and emitter. So the timing is due to the transistor specifications rather than a resonant frequency.
Very nice explanation, I can only imagine how good Your explanations would be if You could do them "at Your own pace" rather than having to adhere to RU-vids time limits. Still great vids, it's ppl like You that makes the W3 worth while Keep up the good work Cudos.
I'm with you! I have the added problem that since I make explanatory videos I have to choose to explain using conventional flow or electron flow. I think in electron flow. So I'd rather everyone switch to electron flow. But sadly, that's not the world I live in. I just double-checked and I didn't say anything about electrons flowing in the video. I also made clear at 2:27 that I was talking about conventional flow. So the video's right in that sense.
i wired a little speaker to the circuit just a minute ago and it gives of the same tone only a bit louder so i assume the noise is correlating to the frequency of the joule thief :) thanks for the quick response!
Connecting the LED between collector and + terminal makes the circuit even more efficient! Only the 'spikes' generated in the coil are used to power the LED. Works 24/7 for about a month on a single battery! Brightness is also ratherr consistent since oscillating frequence increases when battery voltage drops.
I recall having the battery voltage at the LED at one time too. I don't recall what it turned out to be but it wasn't a bad transistor. It might have been that had the connections to the coils wrong. If you look at them closely you'll see that they're supposed to be crossed in a way. Maybe check that.
No, the voltage isn't higher because there are different number of windings. There are the same number of windings for primary and secondary (though some joule thief circuits do have different number of windings for other reasons.) The voltage is higher because the energy built up in the magnet field is all dumped in a short time through the LED. Also, the frequency varies depending on the battery voltage (or charge?). The 40k in the wiki is just an example.
Yeah, there are three things that I know of that end the first step in what goes on, and leads to the transistor shutting off and the whole process reversing - if that's what you mean by the current being limited. Due to time restrictions and not wanting to bore people to death by going over all three, I decided to go with just the saturation reason. That's why I said "one a few things that can cause this process to reverse." I'm glad you liked the video despite it's limitations.
It works only with things that can take power in the form of periodic DC spikes. A TV needs a AC in the form of a sine wave. A HHO cell needs DC and the joule thief spikes are DC, so the HHO cell would be turning on and off at a high frequency. Provided it supplies enough current for the HHO cell then it should work. You'd have to try it to see how much HHO it produces.
I'm not sure if it's using the power more efficiently. It's hard to say if the total power it's using by flashing on and off is less than the total power it would use without the circuit by staying on all the time. It's still possible that during those 'on' times it's using double or triple the normal amount of power. The peak voltage is way above what the LED requires. I measured 24 volts in my other video. But it is making use of otherwise dead batteries, and that's good.
And re the transistor using current flow, I see what you mean. I was referring to the forward bias voltage to turn the transistor on, but didn't mention the current needed to keep it on and that the base current controls the collector current. But again, time restrictions. I wanted to give the minimum needed about transistors to understand the Joule Thief circuit. Not having a formal electronics education I'm not always familiar with the usual explanations. I learn, like I did just now. Thanks!
Thanks. Glad you like it. One thing I should have added was more of the possible ways in which the transformer could shut down besides a saturated core. Oh well.
I love your videos. You explain them very understandably. Completely without unnecessary details. That was the best explanation of a transistor that I've seen yet.. How about a video on a PNP transistor, I'm a beginner in electronics. I would really enjoy that and I'm sure many others will too.
+ThomasTheSailor Chubby Thanks. A PNP transistor just has the reverse direction of flow from the collector to the emitter and the base has to be negative with respect to the emitter for it to flow. But conceptually it's the same.
The noise is normal. If you watch a lot of joule thief videos you'll hear it in many, sometimes very loud. My understanding is that it's the coils and/or core vibrating that's causing it. I don't know offhand how to change it to get rid of it.
Ah, I misunderstood what you were trying to say in your first comment. If you apply 1.5V DC to a 1.85V diode, the diode won't light, not in my experience anyway. If by 'flash' you mean to build up the energy and discharge it quickly then that's what this circuit does.
Mine was only 26 AWG, so that's fairly thin. But I didn't think about what you're saying and my red wire was even thinner, 30 AWG. Keep in mind, when the red coil dumps into the LED, the collapsing magnetic field also induces current into the green coil, which shuts down the gate faster.
It powers the LED with batteries whose voltage is so low that they'd normally be considered dead and thrown out. I guess another is that you can power the tubes of a compact fluorescent light bulb when its circuitry dies by replacing that circuitry with a joule thief, and run it on batteries. And for those so inclined, they're fun circuits to toy with.
Yes, a transformer can raise the voltage enough, but it won't help. A transformer raises the voltage while lowering the current at the same time. The joule thief raises but the voltage and the current at the same time. That's what's needed. The joule thief does this by providing around 0 volts and 0 amps while it builds up energy in a magnetic field, then it dumps that energy at around the same time at high voltage and current. So it's not always on.
With a joule thief, the LED isn't on all the time. So you'd have to compare the brightness for the time it's on. It simple depends on the voltage and current. Depending on the battery it can be brighter or not. If you're using the same battery with the joule thief and then again without, it should be brighter with the joule thief for the times the LED is on.
A good analogy for voltage is pressure but for current it's better to think of the number of electrons passing a point for a given amount of time. I don't see how you can compare the two numerically. You can easily have 12 volts and 100 amps. Volts can be thought of as energy per electron (really joules per coulomb) and amps as electrons per second (really coulombs per second.) So 12 joules per coulomb and 100 coulombs per second.
Not my terminology. The joule thief is a very popular circuit. There's even a wikipedia page for it called Joule Thief. But I agree entirely, it doesn't steal joules (energy) at all.
I'm just going by what my scope showed me. If you look at my Make a Joule Thief for Zombie Batteries video at 0:37 you'll see the waveform for the voltage. There may just not be much current in the reverse direction. I don't see why there would be.
The one in my Make a Joule Thief for Zombie Batteries video was oscillating at 46 microseconds per cycle or 21,739 times a second. If you watch that video you can see this in the oscilloscope clip at 0:36 into the video. Each horizontal division on the scope screen is set to 10 microseconds.
Yes, but you're better off just connecting the capacitor directly to the battery and have the battery charge teh capacitor directly. That way you avoid the losses due to the additional, unnecessary circuitry.
Its the low impedance coil & high inductance ferrite toroid. Low impedance windings inducts high current and generates large flux in the ferite core, so when the field collapses the high flux translates to voltage potential. It just adds a series voltage pulse on top to the battery voltage. The physics gets more complicated, but this is a good start. BTW the opposing polarity of the 2 coils, the larger coil chokes the smaller coil as soon as the NPN opens, and in turn shuts the NPN.
The discharge is in the same direction as the charging current flow. The difference is that during charging, the path is through the coil, the collector and emitter and to the battery. During discharging, the transistor is off, so the path is through the coil, the LED light and to the battery. Putting the LED in the location you suggest and connected the opposite way won't work since the current flow is the wrong way. Put it the existing way and it won't get the benefit of the battery voltage.
I don't know the lowest voltage. It's hard to say since the power going into the LED during the part of the cycle when coil dumps its energy really depends on the coil (how much energy is stored, how fast it can dump it), transistor (how fast it can be shut down) and the resistor (how much it restricts the current going into the transistor to shut it off.) I played with tree power back in 2006 when a company called magcap was experimenting with it for a possible product (nothing came of it.)
Many appliances can run off a square wave. I know that because in the solar power world a lot of cheaper inverters put out what's called a modified sine wave which is really a square wave. If you want an idea of which appliances search for "modified sine wave".
Thanks. It's around 20,000 to 30,000 hertz. You can control it by replacing the fixed resistor with a variable resistor or potentiometer. See my latest video "Fun with Joule Thief Powering a CFL" where I show the waveforms on the oscilloscope and also using the potentiometer and its effect. There's a link to the video in the description below this one. It's also the most recent video on my channel page.
I agree, in this case it's not because the core is saturated. In the video I said there were a few possible ways the cycle could reverse. I should have explained them all, and wish I hadn't chosen not to. Too late. With this small a battery it's more likely it reverses for the reason you say. If it were a larger battery and a ferrite core then core saturation is a possible reason.
30 AWG and 26 AWG. I used two different sizes because that was the only way I could get two different colors. But you could use even up to 18 AWG. There are 13 turns of each but I chose that number at random, it's what fit on the core. See my "Make a Joule Thief for Zombie Batteries" video where I show the parts and me making and testing it. There's a link to it in the description below this video or you can find it on my channel page.
No, I'm not saying that. I just watched the video again to verify it. I don't show the current running in the red wire and through the transistor until the transistor is turned on.
I've seen a few who use it, though it's not clear what, if any changes they made to the circuit. You may have to go with more than 1.5 volts. The smallest I saw was 5 volts.
The joule thief does suck more power out of the battery than it puts out at the LED. If you have a 1.5 volt battery and a 1.4 volt LED then don't use a joule thief. Running the LED directly from the battery will involve less losses. If you have a 1.5 volt battery and a 1.85 volt LED then you need some circuitry to raise the voltage to 1.85 volts for the LED. That's what a joule thief is for. But it pulses the output. The LED isn't being powered continuously, it just looks that way.
