Good stuff for the novice tinkerer. Started making these out of components scavenged from compact fluorescent bulbs, a source for ferrite cores and no cost NPN transistors. Have toyed with the idea of draining “dead” batteries to charge a lipo through a diode with a Zener to set voltage. It is fun to watch the output waveform change as the power source is depleted.
I've made a number of these and you can just about use any transistor and core for the inductor to prove salvaged work just fine. I've made one so small I ended being able to place it in the base of a screw in type flashlight bulb base and a white LED sticking out of the epoxy to seal it all, basically and LED replacement bulb for a single cell flashlight.
With just a capacitor across the resistor you can turn this into a LED flasher. R ~ 82 k, C ~ 22uF > 1s. My test setup ( started with new battery ) is running over 5 years now. Better than the old LM3909.
I always associated joule thieves as circuits providing free energy (maybe an old definition), but as long as you have free batteries you have free energy.
'Joule thief' is a modern term for something called blocking oscillator. Forgoten technology that was rediscovered by modern 'arduino engineers' and given a new name, but the same oscillator was the basis of any flyback type (buck, boost, flyback) converters for about 50yrs. Blockings, being monostable or astable like this one were great for obtainint high-power sharp shaped pulses and were essential in radar, TV, computers... And all of that was based upon high positive feedback through a transformer, and by precise designing of magnetic one can get impressive results. For example one of the best ways for driving SCRs is blocking oscillator. Even today with modern opto devices and stuff. For example it is nice to show and explain action of a transformer in your circuit? Why it should have gap? 😁 Never the less, great video! I'm glad to see great but forgotten technology be given a new life.
Of course, lighting an LED with just one battery is not very exciting, but i have a multimeter that measures volts, Hz, line phase, capacitance and has a led to illuminate. All this works with just one battery.
"Any sufficiently advanced technology is indistinguishable from magic" --Arthur C. Clarke Makes me wonder how long a joule thief would run from a potato battery.
I'd been meaning to make one of these years ago and got side tracked with other things and somehow forgot entirely about it, despite watching Big Clive regularly. Thanks for reminding me about a great circuit and what I'm sure will be a fun and informative project. Cheers!
@hardstyle905 It takes 5 minutes to build one. Not to build one properly and understand its operation on a level deeper than first year bachelor degree. And anything worth doing is worth doing right. Not to mention I can't exactly just drop everything I"m doing to start a new project. I do other things with more priority. Hence it was forgotten in the first place. So, yeah. Compared to real life, it really wasn't important.
Of course it works first time, it was designed by the lovely and well respected TH-camr Big Clive! and while it's design isn't as you say magical, It is however educational.
Clive only added the name. The circuit dates back to the days of tubes and the common transistor implementation comes via an article in Everyday Practical Electronics magazine from 1999 called "One Volt LED - A Bright Light" by Z. Kaparnik from Swindon, Wiltshire, UK.. Clive is very up front about Z. Kaparnik being the source.
If all you have is a ferrite core, some wires and a transistor, it's a good circuit to get higher voltages for biasing. Cool kids these days use fancy charge pump ICs :)
Use it as the basis for the circuit to replace the salt damaged strobe light on my longline sea fishing torpedo, replacement cost NZ$45. LDR to switch it automatically at change of light, charge pump fed from the output, to a diac which, once pumped up to its break over 30 volts, discharges through a line of clear LED's and produces a flash, every few seconds. All off a 1.5 volt cell. Whole thing cost less than a $1, and if it's salt damaged, I have the spare parts, unlike the original. Battery seems to last forever. Do you have any diacs in your 'Aladdin's Cave'? They have other uses apart from triggering triacs. Might make an interesting video for those who have never used one.
Classic simple, but effective, circuit. I'm still laughing at indestructibles. Eat your heart out, PCBWay. I have to say that I learn something new every day (waka waka).
I built one (a little tidier) and use it to charge 18650 lithium batteries from all my tired primary cells. The diode I use is a 1N60 to minimize loss.
@@IMSAIGuy I do. They are everywhere. Meters, remote controls, flashlights, kids toys etc... I currently have a box of about 40 that need to be "processed". It will take most of those to charge one or two 18650 cells depending on their state of discharge and capacity. It's (slightly) better than throwing them out with energy still in them but I'm certainly not saving the planet. More like a conversational diversion. Maybe I will do a video on it. There is still about 40% of the original energy in a alkaline cell when it's down to 1.2V and generally not useable anymore for many applications.
I've made some bike lights that ran off of a single AA battery and drove an LED or two. Instead of a joule thief, it uses a Zetex (now Diodes Inc) ZXSC310. This was a fairly simple boost converter that could run down to 0.8VDC. It was quite handy for me. It seems to be out of production now.. not sure if Diodes Inc has replaced it with an updated part or not.
A very cool little circuit. Apparently these are made in developing countries out of old compact fluorescent parts, and sold at flea markets. Probably quite useful if you don't have electric lights. You could charge a AA or AAA NiMH battery with a small solar panel out of an old calculator. In bright sunlight this could give about 10mA. Might need a diode (or half of a transistor) to prevent the battery from discharging through the solar cell when it gets cloudy. Apparently these are not very efficient because there is a significant DC component going through the coil and resistor. (There are several "energy harvesting" chips on the market to do this more efficiently.) It would be interesting to devise an experiment to measure the input power from the battery, and the output power to the LED. Probably using shunt resistors (maybe 1ohm or 10ohms) to get the current waveforms. Then multiply channel A and channel B waveforms on the scope in real time to get the power. (Need to multiply A and B point by point, add them up over several cycles, and divide by the time. Not just multiply the RMS values.) P batt = V batt * I batt wavefore P led = V led waveform * I led waveform Food for thought. :-)
What do you want to prove here? You can not violate the law of energy conservation; the output power is definitely less always than what is being consumed from the source. De VU2RZA In ABQ, NM
@@subramanianr7206 Yes. I totally agree with you. The power supplied by the battery will be equal to the power consumed by the LED plus the various bits of power consumed by the other components in the form of heat. The simple Joule Thief circuit is not very efficient because a significant portion of the battery power is wasted as heat in the transformer, the resistor and the transistor. Some "energy harvesting" chips do a similar thing but are more efficient (i.e. waste less power as heat). I guess I was not very clear about that. Sorry. Cheers :-)
@hardstyle905 A simple joule thief circuit, as shown in this video, has an efficiency of 40 to 60%. This can be improved to around 80% by making a high efficiency joule thief. Search around in google to find out more.
@hardstyle905 When I say that the joule thief is not very efficient, I am not knocking it. It is a brilliant little circuit because it is so simple to make just by soldering a few components together, and can use up the last remaining energy from a battery. It would make a great little flashlight :-)
my joule thief never worked, i tried a lot of different inductors but it seems doesnt work, maybe the problem is at the c2383 transistor, should i buy transistor like you
I was interested in making a night light with a low power high-efficiecy LED powered from a supercap. And it might be a cheap solution in order to use most of the stored energy compared to very expensive energy harverster boost controller ICs that I managed to google back then. What alternatives you may think of?
This is similar to how battery operated sound measure instrumentation gets their 200 V for condenser microphone. Electret is no good it drifts calibration
I tried to get more efficient with my video editing and made a master file with all my normal settings. each time I use this file as a starting point. many videos later, someone pointed out the 50% pan to right. Oh no.... I had made a bunch of videos with this new master file and it would be too much work to go back and re-do each file. So much for efficiency 😕
You spent so much time putting the circuit down you failed to see the important lesson that this setup teaches with a resistor, a hand wound coil and an NPN transistor. If you use the capacitor what happens if you move it, like in parellel with the resistor, what happens if you increse the resistor and capacitor size (make larger) when in parellel, what happens if you put the LED to positive rather than negative, what if you change the LED to a two pin RGB, add a diode before the LED and a capacitor to ground after the diode and before the RGB LED. Go back and learn from your errors with a little experimentation. The big question, Can you make the circuit better i.e. higher volts, lower amps with minimal extra parts.
this circuit is called a DC-DC boost converter. there are dozens if not hundreds of ways to do this. here is one: th-cam.com/video/oj7oub16XyM/w-d-xo.htmlsi=0MqplKOVoYDOOakR
One of those rare "simple" circuits you really can throw together from whatever close enough junk and get results.
Great way of making old batteries leak even more!
Good stuff for the novice tinkerer. Started making these out of components scavenged from compact fluorescent bulbs, a source for ferrite cores and no cost NPN transistors. Have toyed with the idea of draining “dead” batteries to charge a lipo through a diode with a Zener to set voltage. It is fun to watch the output waveform change as the power source is depleted.
I've made a number of these and you can just about use any transistor and core for the inductor to prove salvaged work just fine. I've made one so small I ended being able to place it in the base of a screw in type flashlight bulb base and a white LED sticking out of the epoxy to seal it all, basically and LED replacement bulb for a single cell flashlight.
Clever!
There are patents on this concept that go back to 1930 using tubes. Clever idea.
goes back even earlier " induction coil" my school teacher told me
With just a capacitor across the resistor you can turn this into a LED flasher. R ~ 82 k, C ~ 22uF > 1s. My test setup ( started with new battery ) is running over 5 years now. Better than the old LM3909.
I always associated joule thieves as circuits providing free energy (maybe an old definition), but as long as you have free batteries you have free energy.
'Joule thief' is a modern term for something called blocking oscillator. Forgoten technology that was rediscovered by modern 'arduino engineers' and given a new name, but the same oscillator was the basis of any flyback type (buck, boost, flyback) converters for about 50yrs. Blockings, being monostable or astable like this one were great for obtainint high-power sharp shaped pulses and were essential in radar, TV, computers... And all of that was based upon high positive feedback through a transformer, and by precise designing of magnetic one can get impressive results. For example one of the best ways for driving SCRs is blocking oscillator. Even today with modern opto devices and stuff.
For example it is nice to show and explain action of a transformer in your circuit? Why it should have gap? 😁 Never the less, great video! I'm glad to see great but forgotten technology be given a new life.
wakkawakka. It's simple harmonic oscillators all the way down.
I made one with two discrete inductors a while ago. With a bright white led it made a decent little flashlight. I think I got it down to about 0.3v.
Of course, lighting an LED with just one battery is not very exciting, but i have a multimeter that measures volts, Hz, line phase, capacitance and has a led to illuminate. All this works with just one battery.
"Any sufficiently advanced technology is indistinguishable from magic" --Arthur C. Clarke
Makes me wonder how long a joule thief would run from a potato battery.
I'd been meaning to make one of these years ago and got side tracked with other things and somehow forgot entirely about it, despite watching Big Clive regularly. Thanks for reminding me about a great circuit and what I'm sure will be a fun and informative project. Cheers!
@hardstyle905 It takes 5 minutes to build one. Not to build one properly and understand its operation on a level deeper than first year bachelor degree. And anything worth doing is worth doing right.
Not to mention I can't exactly just drop everything I"m doing to start a new project. I do other things with more priority. Hence it was forgotten in the first place. So, yeah. Compared to real life, it really wasn't important.
Isn't the name 'Joule thief' created by Big Clive?
Indeed!
Correct.
Yes, it was.
it's Big C. Live, not big clive
Let me check. One moment, please... :D
I beat everyone to it in 1971, I constructed an AM radio that would only operate on a dead battery
Of course it works first time, it was designed by the lovely and well respected TH-camr Big Clive! and while it's design isn't as you say magical, It is however educational.
Clive only added the name. The circuit dates back to the days of tubes and the common transistor implementation comes via an article in Everyday Practical Electronics magazine from 1999 called "One Volt LED - A Bright Light" by Z. Kaparnik from Swindon, Wiltshire, UK.. Clive is very up front about Z. Kaparnik being the source.
Your quite fluent in ' Pac Man '...cheers.
Funny I said wirewrap the same time as you did (-: And then I had my wine time.
The ultimate demo would have been to run it of a single potato which (from what I gathered from the internet provides around 0.5V).
That would be a potato thief. 🤣
If all you have is a ferrite core, some wires and a transistor, it's a good circuit to get higher voltages for biasing. Cool kids these days use fancy charge pump ICs :)
Use it as the basis for the circuit to replace the salt damaged strobe light on my longline sea fishing torpedo, replacement cost NZ$45. LDR to switch it automatically at change of light, charge pump fed from the output, to a diac which, once pumped up to its break over 30 volts, discharges through a line of clear LED's and produces a flash, every few seconds. All off a 1.5 volt cell. Whole thing cost less than a $1, and if it's salt damaged, I have the spare parts, unlike the original. Battery seems to last forever.
Do you have any diacs in your 'Aladdin's Cave'? They have other uses apart from triggering triacs. Might make an interesting video for those who have never used one.
th-cam.com/video/2WWUhsus9ic/w-d-xo.html
A stiff drink, I hope! 3 fingers.
You an find these pre-built inside practically every solar light. Single chip + inductor on a tiny PCB.
😊😊All though, Whacca Whacca ! I wouldn't hire him her it that 😂😂😂😂 73 Leo
Classic simple, but effective, circuit. I'm still laughing at indestructibles. Eat your heart out, PCBWay. I have to say that I learn something new every day (waka waka).
I built one (a little tidier) and use it to charge 18650 lithium batteries from all my tired primary cells. The diode I use is a 1N60 to minimize loss.
That could be an interesting video.
you must use a lot of primary batteries
@@IMSAIGuy I do. They are everywhere. Meters, remote controls, flashlights, kids toys etc... I currently have a box of about 40 that need to be "processed". It will take most of those to charge one or two 18650 cells depending on their state of discharge and capacity. It's (slightly) better than throwing them out with energy still in them but I'm certainly not saving the planet. More like a conversational diversion. Maybe I will do a video on it. There is still about 40% of the original energy in a alkaline cell when it's down to 1.2V and generally not useable anymore for many applications.
I was on 7.143 Whacca Whacca last night...
This solution to our energy crisis needs to be taken far more seriously.
Do the "Super joule ringer looper" (Capacitor based LED driver - SJRL V4) by lasersaber. Thats where the real magic is :D
Merci for this video.
I've made some bike lights that ran off of a single AA battery and drove an LED or two. Instead of a joule thief, it uses a Zetex (now Diodes Inc) ZXSC310. This was a fairly simple boost converter that could run down to 0.8VDC. It was quite handy for me. It seems to be out of production now.. not sure if Diodes Inc has replaced it with an updated part or not.
A very cool little circuit. Apparently these are made in developing countries out of old compact fluorescent parts, and sold at flea markets. Probably quite useful if you don't have electric lights.
You could charge a AA or AAA NiMH battery with a small solar panel out of an old calculator. In bright sunlight this could give about 10mA. Might need a diode (or half of a transistor) to prevent the battery from discharging through the solar cell when it gets cloudy.
Apparently these are not very efficient because there is a significant DC component going through the coil and resistor. (There are several "energy harvesting" chips on the market to do this more efficiently.)
It would be interesting to devise an experiment to measure the input power from the battery, and the output power to the LED. Probably using shunt resistors (maybe 1ohm or 10ohms) to get the current waveforms. Then multiply channel A and channel B waveforms on the scope in real time to get the power. (Need to multiply A and B point by point, add them up over several cycles, and divide by the time. Not just multiply the RMS values.)
P batt = V batt * I batt wavefore
P led = V led waveform * I led waveform
Food for thought.
:-)
What do you want to prove here? You can not violate the law of energy conservation; the output power is definitely less always than what is being consumed from the source.
De VU2RZA
In ABQ, NM
@@subramanianr7206
Yes. I totally agree with you. The power supplied by the battery will be equal to the power consumed by the LED plus the various bits of power consumed by the other components in the form of heat.
The simple Joule Thief circuit is not very efficient because a significant portion of the battery power is wasted as heat in the transformer, the resistor and the transistor.
Some "energy harvesting" chips do a similar thing but are more efficient (i.e. waste less power as heat).
I guess I was not very clear about that. Sorry.
Cheers :-)
@robinbrowne5419 thanks for the clarification; it is much appreciated 👍
De VU2RZA
in ABQ, NM
@hardstyle905 A simple joule thief circuit, as shown in this video, has an efficiency of 40 to 60%. This can be improved to around 80% by making a high efficiency joule thief. Search around in google to find out more.
@hardstyle905 When I say that the joule thief is not very efficient, I am not knocking it. It is a brilliant little circuit because it is so simple to make just by soldering a few components together, and can use up the last remaining energy from a battery. It would make a great little flashlight :-)
wakka wakka wakka
my joule thief never worked, i tried a lot of different inductors but it seems doesnt work, maybe the problem is at the c2383 transistor, should i buy transistor like you
I was interested in making a night light with a low power high-efficiecy LED powered from a supercap. And it might be a cheap solution in order to use most of the stored energy compared to very expensive energy harverster boost controller ICs that I managed to google back then. What alternatives you may think of?
There is a cap in the lower diagram but not in the upper. Weird it still works without one, though.
I noticed that too but at 11:08 he said, "capacitor is optional" so no cap.😀
This is similar to how battery operated sound measure instrumentation gets their 200 V for condenser microphone. Electret is no good it drifts calibration
after your microphone teardown #1592 every video just had sound on the right channel didnt you put it back together?
Sound on both channels for me - iOS
Fun
whacka-whacka
Where did you get your schematic from? I was hoping you'd share .
From the scope it appears you only have 200mV across the LED.
He's probably using 1:10 probes without telling the oscilloscope (that thinks it's 1:1). So the oscilloscope sees 2v *1/10=200mV.
Audio mostly in right channel is deliberate or not?
I tried to get more efficient with my video editing and made a master file with all my normal settings. each time I use this file as a starting point. many videos later, someone pointed out the 50% pan to right. Oh no.... I had made a bunch of videos with this new master file and it would be too much work to go back and re-do each file. So much for efficiency 😕
Breadboard would of been better.
🤣🤣🤣👍👍👍👍👍
Nice contents, but who was the Joule thief? You didn't mention him/her?
The Joule thief would either be Z. Kaparnik of Swindon, Wiltshire, UK (circuit) or C. Mitchell or Ramsey, Isle of Man (name). It's a he in both cases.
You spent so much time putting the circuit down you failed to see the important lesson that this setup teaches with a resistor, a hand wound coil and an NPN transistor. If you use the capacitor what happens if you move it, like in parellel with the resistor, what happens if you increse the resistor and capacitor size (make larger) when in parellel, what happens if you put the LED to positive rather than negative, what if you change the LED to a two pin RGB, add a diode before the LED and a capacitor to ground after the diode and before the RGB LED. Go back and learn from your errors with a little experimentation. The big question, Can you make the circuit better i.e. higher volts, lower amps with minimal extra parts.
this circuit is called a DC-DC boost converter. there are dozens if not hundreds of ways to do this. here is one: th-cam.com/video/oj7oub16XyM/w-d-xo.htmlsi=0MqplKOVoYDOOakR