Hi everyone.

I am so thankful for all the help I have recieved in the past that I was hoping to gain more applicable knowledge while preparing my next project (this time, I want to get ready to make 4 copies of this.)

To keep this short and sweet, my next project is to build a faraday's flashlight (not crank but shake) for me as a prototype, then remake this for my nefews and a friend later on. I want to use LEDs to make it a very useful flashlight, but as I gather the electrical components, I am left with wondering if the design can be improved. My idea is to double up the recipe to decrease effort, but I want to check my theory here.

First: Faraday's law is applied by having a thin insolated wire tightly wrapped around the circumference of a plastic pipe, with enough of the pipe hanging out on both sides so that a neodymium magnet can pass in and out of the coil on both sides creating voltage (? I would think it's amperage but I digress) and that this process can be scaled for a handheld or regular sized flashlight (I haven't seen a big box flashlight done, but I am guessing the bang of the shaking magnet wouldn't make this worth it.)

Since the scale of the coil can vary for different sizes, and the flashlight I am using is one that takes 2 D batteries, my question here is if I would have enough room to mount 2 coils? My idea is to plug the pipe half ways and insert a magnet on both ends, with the same polarity touching to avoid sticking, so that one shake would do twice the work. I realize magnetic fields are like doughnuts, so doubling it up in sequence might give me better results.

Second: I am still trying to grasp the concept of a superconductor. If I understand it correctly, it's like a temporary rechargable battery in it that you can fill it full of a charge to use later. However, unlike the battery, the accumulated charge is flushed out immediately when used, rather than gradually, and that the charge doesn't store for too long. This is why 2 AAA batteries can be used to create a powerful camera flash, rather than a dim light, and how the shake flashlight stores a charge for a shake and go action (adding a resister to make the flow more gradual).

The reason I compared it to rechargable batteries is that I was wondering if, with the double coil or if one had a smaller superconductor, you can wire them in series (or parallel to each other, but both feeding the same circuit) like batteries to store double the charge? Since it's in series, and a resister is put to control the output, would this allowe an accumulation of volts? Or would the loss of charge outweigh this?

To conclude (image for visual refference) my grandios scheme to take over the world (narf) is to create a shake flashlight, doubling the magnets, coils and superconductors to, essentially, either double the output or half the effort. My worry is that if this is possible, why hasn't it been done before?

 

Another question that came up, while searching for insolated copper wire, is what would happen if you had two wires in the coil? Like making the coil out of copper speaker wires?

 

Are you sure that you are not mixing up superconductors with supercapacitors ?

Les.

 

And supercapacitors and LEDs are polarity sensitive but the coil/magnet system will generate AC and so a rectifier will be needed.

 

Are you sure that you are not mixing up superconductors with supercapacitors ?

Les.

You're right. Sorry, for some reason I see them in my head, but it ends up coming out wrong. Will edit.

And supercapacitors and LEDs are polarity sensitive but the coil/magnet system will generate AC and so a rectifier will be needed.

I will look into that. Don't quite know what a rectifier is/looks like. Where would you place it in the sequence?



However, is there any tips on the double system I was asking about?

 

You have many problems to solve using supercaps. As you show three LEDs in series you will need at least 10 volts to power them. This is above the voltage rating of many supercaps. The voltage available from the supercaps as they discharge varies very much as they are discharged so to to recover most of the stored energy you would need some form of switch mode regulator. As AlbertHall has told you you need to rectify the output from the coils to charge the supercaps. To use the output from one direction of movement of the magnet you can use one diode (Half wave rectifier.) in series with each coil. If you want to use the output from both directions you will need a bridge rectifier connected to each coil. (4 diodes.) Schottky diodes would have a lower forward voltage drop than silicon diodes but if you can find any germanium diodes of suitable rating they would have an even lower forward voltage drop. Your schematic is totaly wrong. (Even if it was possible to store alternating current.) My feeling is the windup torch with rechargable batteries is a much better solution to the problem.

Les.

 

Not to rain on your parade, but you cannot build one for less than what it costs to buy one. plus labor.
https://www.kjmagnetics.com/blog.asp?p=shake-flashlight
http://www.ebay.com/itm/like/222864579573

ak

 

You have many problems to solve using supercaps. As you show three LEDs in series you will need at least 10 volts to power them. This is above the voltage rating of many supercaps. The voltage available from the supercaps as they discharge varies very much as they are discharged so to to recover most of the stored energy you would need some form of switch mode regulator. As AlbertHall has told you you need to rectify the output from the coils to charge the supercaps. To use the output from one direction of movement of the magnet you can use one diode (Half wave rectifier.) in series with each coil. If you want to use the output from both directions you will need a bridge rectifier connected to each coil. (4 diodes.) Schottky diodes would have a lower forward voltage drop than silicon diodes but if you can find any germanium diodes of suitable rating they would have an even lower forward voltage drop. Your schematic is totaly wrong. (Even if it was possible to store alternating current.) My feeling is the windup torch with rechargable batteries is a much better solution to the problem.

Les.

I see I have to review the schematics a bit and add a few other components. However, the 3 LEDs are there as a thought I had, seeing someone take appart an LED lightbulb where the disk of LEDs were connected to a mini circuit board with a single capacitor and a regulator, much like the tutorials I've seen for a shake flashlight. That said, I might have been jumping the gun with my explinations and the diagram.

Not to rain on your parade, but you cannot build one for less than what it costs to buy one. plus labor.
https://www.kjmagnetics.com/blog.asp?p=shake-flashlight
http://www.ebay.com/itm/like/222864579573

ak

You're right that they are sold cheap, and that buying one is most likely simpler than wasting my time to build one.


Thank you everyone for your help. I am still unsure if I can proceed with my modifications, but I do understand that I need to do much more studies on the hardware and circuitry before I should think about going beyond tutorials.

Have a good day

 

My worry is that if this is possible, why hasn't it been done before?

Are you sure it hasn't? I mean, most shake flashlights are cheap and mostly a novelty, but that doesn't mean there isn't a really good one on the market somewhere already.

One thing that comes to mind is that most of the work required to shake these things is moving the mass of your arm, not the tiny work you are doing to move the magnet back and forth. That suggests to me that much more useful work (energy for lighting) could be extracted from the shaking without seriously increasing the work done by the user. So I definitely think the cheap ones I have used could be made more practical by using bigger, heavier, more costly magnets and perhaps more copper in the coils. Instead of the crude circuitry you usually find in these things, you could use an energy harvesting IC that has been designed to capture energy as efficiently as possible. The LED itself could be upgraded to one used in high quality flashlights instead of the piece of crap LED these usually have in them.

 

And replacing the very-lossy supercapacitor with a rechargeable cell, plus a led driver.
Or adding a solar panel, or replacing the shaking idea entirely :

----> https://www.aliexpress.com/item/50-...897074a&transAbTest=ae803_5&priceBeautifyAB=0

----> https://www.aliexpress.com/item/SMT...61aafa9&transAbTest=ae803_5&priceBeautifyAB=0

 

You have to read my article and project on Tic Tac LED Flashlight but the main fault in most of the other projects is the way the winding is laid.
A voltage (and current) is only developed in the turns when the end of the magnet is moving towards or away from the turn. When the barrel of the magnet is passing down the length of the turns, only the "end turns" are generating anything. This means 90% of the turns ore not doing anything for the majority of the time. The only way to really solve this is to bunch the turns into what is called a "pancake winding " Because the magnet is moving so slowly, it takes hundreds of turns to create a few volts.

SHAKE TIC TAC LED TORCH
In the diagram, it looks like the coils sit on the “table” while the magnet has its edge on the table. This is just a diagram to show how the parts are connected. The coils actually sit flat against the slide (against the side of the magnet) as shown in the diagram:

​

The output voltage depends on how quickly the magnet passes from one end of the slide to the other. That's why a rapid shaking produces a higher voltage. You must get the end of the magnet to fully pass though the coil so the voltage will be a maximum. That’s why the slide extends past the coils at the top and bottom of the diagram.

The circuit consists of two 600-turn coils in series, driving a voltage doubler. Each coil produces a positive and negative pulse, each time the magnet passes from one end of the slide to the other.
The positive pulse charges the top electrolytic via the top diode and the negative pulse charges the lower
electrolytic, via the lower diode.
The voltage across each electrolytic is combined to produce a voltage for the white LED. When the combined voltage is greater than 3.2v, the LED illuminates. The electrolytics help to keep the LED illuminated while the magnet starts to make another pass.

 

Are you sure it hasn't? I mean, most shake flashlights are cheap and mostly a novelty, but that doesn't mean there isn't a really good one on the market somewhere already.

One thing that comes to mind is that most of the work required to shake these things is moving the mass of your arm, not the tiny work you are doing to move the magnet back and forth. That suggests to me that much more useful work (energy for lighting) could be extracted from the shaking without seriously increasing the work done by the user. So I definitely think the cheap ones I have used could be made more practical by using bigger, heavier, more costly magnets and perhaps more copper in the coils. Instead of the crude circuitry you usually find in these things, you could use an energy harvesting IC that has been designed to capture energy as efficiently as possible. The LED itself could be upgraded to one used in high quality flashlights instead of the piece of crap LED these usually have in them.

Every project or video I have seen so far on the shake flashlights (except the one from Colin 55 on this thread) have only used one coil. Using the knowledge I gained, I wanted to see if I could, or why I could not, make more use of Farday's law and apply it to my own flashlight (which has a crappy light and takes 3 D batteries) and then reutilize it to make survival/camping flashlights for my 2 nefews. With me being as ambitious as I am though, I wanted to repurpose an LED lightbulb parts a) to gather materials from their circuit chips, but mainly b) use low-voltage/high-output to make a really cool, really bright flashlight that would survive.

As for the mass of the arm moving, by faraday's law, the current us caused by the magnet passing through the coil, and not via the size of the magnet. Ergo, the bigger your magnet, the bigger it's magnetic field thus the more distance your magnet has to travel away from your coil. And if the math is the same for the faraday coil as is in a simple crank generator, changing the size of wire, rotations, or such would give you less output per effort, or not be able to generate enough, making the larger magnet more a detriment than anything. Not to mention that it would be heavy for your arm to shake due to Newton's third law.

And replacing the very-lossy supercapacitor with a rechargeable cell, plus a led driver.
Or adding a solar panel, or replacing the shaking idea entirely :

----> https://www.aliexpress.com/item/50-...897074a&transAbTest=ae803_5&priceBeautifyAB=0

----> https://www.aliexpress.com/item/SMT...61aafa9&transAbTest=ae803_5&priceBeautifyAB=0

Yes, solar panel flashlights are nice, and really cool, but there's a paradoxal problem with them in the area where faraday flashlights outperform all other flashlights. Out of all flashlights, the first thing to go down regardless of usage is the battery. Even the best rechargable linthium ion battery has a shelf life since physical particles that hold either a positive or negative charge has an attraction to the opposing charge. Because of this, overtime the battery neutralizes itself and corrodes. So even being recharged by solar powered, those flashlights do have a shelf life shorter than you or I have. Even if you were to remove the battery aspect, rendering the flashlight only usable during the day, the best solar pannel has a shelf life of 20-30 years before the mineral starts wearing out. But given the price, I would say 10 year life is stretching it.

Meanwhile, stable magnets work for 1000+ years and the other components only get wear and tear when used, meaning that the potental for a faraday flashlight is an ability to last up to 1000 years.

Not saying that I am planning to live that long, but looking at battery vs shake, I'd rather use a renewable, sustainable energy source than one where I need to pack extra batteries and might not get the full bang for buck.

Besides, if your batteries die in the middle of the night, you're stuck. But a few good shakes and "let there be light"

You have to read my article and project on Tic Tac LED Flashlight but the main fault in most of the other projects is the way the winding is laid.
A voltage (and current) is only developed in the turns when the end of the magnet is moving towards or away from the turn. When the barrel of the magnet is passing down the length of the turns, only the "end turns" are generating anything. This means 90% of the turns ore not doing anything for the majority of the time. The only way to really solve this is to bunch the turns into what is called a "pancake winding " Because the magnet is moving so slowly, it takes hundreds of turns to create a few volts.

SHAKE TIC TAC LED TORCH
In the diagram, it looks like the coils sit on the “table” while the magnet has its edge on the table. This is just a diagram to show how the parts are connected. The coils actually sit flat against the slide (against the side of the magnet) as shown in the diagram:

​

The output voltage depends on how quickly the magnet passes from one end of the slide to the other. That's why a rapid shaking produces a higher voltage. You must get the end of the magnet to fully pass though the coil so the voltage will be a maximum. That’s why the slide extends past the coils at the top and bottom of the diagram.

The circuit consists of two 600-turn coils in series, driving a voltage doubler. Each coil produces a positive and negative pulse, each time the magnet passes from one end of the slide to the other.
The positive pulse charges the top electrolytic via the top diode and the negative pulse charges the lower
electrolytic, via the lower diode.
The voltage across each electrolytic is combined to produce a voltage for the white LED. When the combined voltage is greater than 3.2v, the LED illuminates. The electrolytics help to keep the LED illuminated while the magnet starts to make another pass.

It sounds like you're really touching on my wavelength regarding increasing the faraday's law, although it seems that you opted for a coil on either side rather than two with a magnet passing through them.

However, although I get what the aspects of your diagram shows, I am not quite following it due to my novice knowledge. I would love to read more about your project, if you would link me, and see how you built it in real life (since I learned that electronical diagrams and real life doesn't always match up graphically) I'd like to see if I can borrow on your research, if you don't mind.

 

The only time when the turns are generating a voltage is when the magnet is passing over them. With a long coil, only 10% of the coil is generating at any given time.

 

As for the mass of the arm moving, by faraday's law, the current us caused by the magnet passing through the coil, and not via the size of the magnet.

Nope. The relative motion of the conductor past a magnetic field line generates the EMF. More field lines, more EMF for a given coil. 10% larger face area means 10% more power. If you look at the specs of magnets, you'll see that larger magnets have stronger fields. Some magnetic materials offer denser fields for the same sized magnets.

A longer cylinder magnet and a longer coil might allow a greater portion of each cycle to be making power, but the total cycle distance has to be related to ergonomic considerations, to optimize shaking. I have no idea of the current designs are optimal in this regard. Maybe.

Ergo, the bigger your magnet, the bigger it's magnetic field thus the more distance your magnet has to travel away from your coil.

All that matters is the rate of change of magnetic field seen by each loop of the coil. That's maximal as the face of the magnet passes a loop. There is very little contribution to EMF anywhere else.

And if the math is the same for the faraday coil as is in a simple crank generator, changing the size of wire, rotations, or such would give you less output per effort, or not be able to generate enough, making the larger magnet more a detriment than anything.

Optimizing the coil winding has a lot more to do with matching the load than in capturing more energy. The general geometry of the coil will be related to the magnet geometry, but the choice of wire gauge depends on the load. You can have lots of loops and a higher output voltage, or less loops and a higher current. Matching the coil impedance to the load impedance will maximize power transfer to the load.

The cheap devices use a diode rectifier. The voltage drop across that is a significant power loss if the coil is wired to low voltage. But wiring for high voltage causes an increased DC resistance in the coil and much of your EMF gets burned off in the coil. In other words, these things have to be designed as a system, where the various design choices are all inter-related.

Here's a reference that might help you a little. https://www.kjmagnetics.com/blog.asp?p=shake-flashlight

 

"You can have lots of loops and a higher output voltage, or less loops and a higher current."

Simply reducing the number of turns does not increase the current.
The current is determined by the voltage produced.
Because the magnet is moving so slowly, you will never get to the condition of delivering the maximum current.
However the current in this case is severely limited by the thickness of the wire and the number of turns that are not actually delivering a voltage.

 

Here's a reference that might help you a little. https://www.kjmagnetics.com/blog.asp?p=shake-flashlight

There are a number of mistakes in the article above.

The output voltage does not reverse twice for each travel of the magnet and there is no point in adding a resistor as the impedance of the generating circuit will match the impedance of the LED.

 

A 20mm diameter magnet with the same flux density as a 10mm magnet will produce the same output.
A 20mm magnet will "feel stronger" because it has a greater surface area and thus more magnetic lines of force.
Thus is is very difficult to visually differentiate between the two.

 

I actually attempted a similar project but encountered many problems. I used a 20lbs pull force neodymium magnet and ~1500 turns of magnet wire (insulated copper wire). I added ferromagnetic (iron) washers and still got only 2.5V AC RMS max. Worse yet, there was an internal resistance of 35 ohms, meaning it took forever to charge my energy storage (5.5V 7.5F supercaps). This does not even go into the losses of my circuit. I lost 1.2/2.5V (50%!) on the bridge rectifier and the supercaps did not even get to the fV of the LEDs. I tried using single diodes, and other things, but still had so many losses I had to give up.

So either use far more powerful magnets and lower-resistance wire or go with a hand-crank motor that can generate a decent voltage. It is going to be VERY difficult to make a decent design. Not impossible, but it will not be easy.

 

A 20mm diameter magnet with the same flux density as a 10mm magnet will produce the same output.

Care to make a wager on that?

 

Sure...
https://www.kjmagnetics.com/proddetail.asp?prod=D7X0
https://www.kjmagnetics.com/proddetail.asp?prod=DEX0
It's only 4x the pull force. That's trivial!