Generator for LEDs

For a few weeks I have been pondering a scheme to rig up a lot of LED running lights on my bicycle frame for night time illumination.

At first I was going to use a 9v and then sets of 3-4 in series with a resistor, but as the voltage of the battery drops, the current seems to drop so much that I don't know what effect it has on luminosity.

I have no formal electronics education, it's all been from doing and reading.

I recently purchased a cheap shake-up LED flashlight and I've seen advertised the crank kind.

It seems I should be able to fashion a generator out of the wheel using magnets on the spokes, or else pull a small motor out of something and use that. As I understand from the archives, I need to find a motor with a magnet rather than an electric field. Would a hard drive magnet work for the spokes?

If I cause a DC motor to turn, does it output DC power, or AC power? I thought since a magnet moves first toward, and then away from the windings (or vise-versa) then this creates first a negative and then positive voltage; a sine wave rather. Is this correct?

Can the magnets pass next to those windings or do they have to pass through the middle of it?

Do those flashlights use a special capacitor? I notice it still has power even several days later. I would expect a capacitor to be drained?

So, I envsion a battery-free setup with a dozen or so LEDs fashioned onto the frame (very minimal hardware and very low budget) and one or more capacitors to keep supplying the current when I have to stop.

If the magnet idea isn't feasible, then I want to make a generator, but not the impossible to pedal tire eaters I remember from the 1980s. I used to spin motors from capsela toys and watch them make power.

If a DC motor consumes X watts at its RPM, does it produce approximately that many watts if made to spin at that same RPM? I presume I need to determine my current demand which I think will be 20mA times the number of LEDs and then find a motor to accomodate this. Is a motor's output voltage dependent on its RPM or is it steady? Would a floppy drive, CD-ROM, or hard drive have a suitable motor? What about trimming off the blades of a computer case fan and using that as a motor, or is there something in an old VCR I could use?

It seems so difficult to design this circuit because voltage can be so dynamic and from the LED calculators I've used, this has a dramatic effect on the current the LED uses. I don't want to burn them up, but I want them to be bright.

I don't know much about regulation, but if that is the answer, should I regulate voltage, or current?

As I understand, making them flash necessitates a transistor or two, which is probably beyond me and probably beyond the scope of this project.

I appreciate any pointers you can offer.

-Ryan

The more powerful the magnet, the larger the output. The (open circuit) voltage is proportional to he rate of change of flux passing through the coil, and also proportional to the number of turns.

As I said, the coil should be wound around a good core material, like laminations recovered from a small transformer. Wire can also be recovered from a transformer (or wall wart power supply or relay). In fact, a relay with the coil on a bobbin may be all you need.

The idea is to have the magnet go past the core so that the core almost touches both poles of the magnet, so it intercepts a large fraction of the magnets flux. If you have access to old lawn mowers, you may be able to adapt the magneto coil that is mounted near the fly wheel.

If you can think of a way to have this core bridge the passing poles of the magnets, it would work.

No. A transformer has the coils wound around the center lag of a closed magnetic path.

If you can slide the E and I laminations out of the coil, and then reassemble all the E's through the coil from one side, you end up with a magnetic structure with 3 poles (a big center one with the coils around it, and a smaller pole on each side). This structure can be used to intercept the flux from a passing magnet as it slides past the three poles.

The impedance of the coil limits the current. Whether this keeps the current low enough to not damage the LEDs depends on the number of turns and the strength of the magnets. Some experimentation is called for. With a direct connection, you will get a flash each time a magnet goes past. if you rectify the pulses and store them in a capacitor, then you can dump power continuously through a string of LEDs and a series resistor. If you use the DC to charge a small rechargeable battery, you can have a glow, even when you stop.

Yes, same concept. But more voltage implys less current, so you can limit the current to a safe value (and drive more LEDs in series) by raising the voltage.

The distance is related to the strength of the field produced by the magnet, the distance between the poles, and the permeability of the material being attracted.

Larger magnets (bigger distance between the poles) produce larger flux patterns. Really good magnets would be neodymium iron boron types. These are sold in great variety on Ebay. But they are powerful enough that larger ones will collect scrap metal as you ride.

Thanks John, you are always very helpful in this group.

Can this be any old magnet, or does it need to be made of anything in particular? How close must it pass to the coil? I presume the magnetic intensity influences voltage?

Do I want to use tiny wire for the coil so that I can make more windings? Could I rip thin wire like what is found on the back of a speaker cone and make a coil from that? Could I wrap it around an ink pen body or what? Does the radius of this coil matter?

When you say a core of ferrite, is this the stuff you find on the end of a monitor data cord? Should I wind around one of those or perhaps through one of those?

If I pull a transformer out of something, does this serve as two pickup coils already?

If I power them with AC current, does the resistor become less important since it is effectively pulsed width duty cycle? I presume I could also add two capacitors in the same fashion to store energy for each "direction" of current?

I presume more turns equal more voltage, not unlike the rule of a transformer?

Is the magnetic field "falloff" curve for any object the same? For example, can one type of metal attract from a farther distance than another type of metal given that they have the same measured amount of magnetism?

-Ryan

A current regulator isn't that hard to build - see my earlier thread on luxeon LEDs for various circuits you might be able to adapt.

For LEDs, you want to regulate current. Using a resistor in series is a trick people use when they know they have a stable fixed voltage source.

they used to make generators for that purpose that were driven by a friction wheel of the tyre.

DC, kind of noisy DC, but DC.

yes, but the contacts in the motor reverse that sine wave where it crosses 0V

through the middle is best. next to will work to a lesser extent.

possibly a rechargable battery?

no, at best about 60% of the amount it consumes. probably much less.

I presume I need to

it depends on the RPM.

All those motors have electronic switching instead of the comutator that the toy motors you experimented with had. you'd need to remove the switching circuit and replace it with a rectifier bridge.

use the generator to charge the capacitor and use a regulator to control the putput.

ideally you should hook the diodes up in series and reguulate current this may require boosting the voltage..

voltage boosting regulators are more complex still.

A good motor to use could be the head motor from a 5.25" floppy drive they're nice and big (so potentially have a good output) and give 40 pulses per rotation. you could probably drive one fast enough to light LEDs with a using a cog that engages the spokes near the hub...

FWIW the turntable motor from a microwave oven can produce hundereds of volts when turned by hand... if you could somehow drive one of them from an axle you could have CCFL lighting on your bike :)

Bye. Jasen

Around, could work through won't

the idea is to make a path from the north to the south pole of the magnet out of metal (or ferrite etc) and have the path pass through the turns of copper then move the magnet round...

If you can cut one of then ends off it it would work, some tranformers come apart quite easily.

no, you can still burn them out with AC if you feed them enough...

yes.

Bye. Jasen

You might like to have a look at Silicon Chip Feb 2004 - they describe a few human powered torches using small stepper motors to power led torches - might find some useful ideas there

It would seem not too hard to rig a motor so it is driven off a bike wheel instead of being hand cranked

David

I've been pondering the same thing. That's what brought me to this group. I initially planed on using a bunch of hard drive magnets glued around the rim but then remembered the brakes. I also decided that it would be best to have the rig near the hub so it doesn't interfere with removing the wheel as much.

I took one apart and it's not a stepper, it's just a regular DC motor. This is what gave me the idea. I actually plan to cannibalize the circuitry of the wind up flashlight and just turn my wheel into a motor/generator. The cool thing is that my flashlight has a rectifier already in it.

Actually it is. They make generators that rub against the tire but drag a lot. They also make hubs with generators in them but they cost a lot.

You don't want to drive anything with the spokes, that will bend them eventually. There is no room between the hub and the fork ends to attach anything.

Wind a couple "pancake" coils to mount on the forks, then clip magnets to the spokes, lined up with the coils. Rectify the coil output. Presto! :)

I'm looking for something that will have the coils in it that I can cannibalize. It's just more satisfying that way.

Why? The flashlight circuitry already has a rectifier.

My bad - didn't notice that part.

Actually I was hoping someone would know what I could cannibalize the coil from.

Old AM radio loopsticks with the ferrites removed, perhaps? Then again, those might be too fragile to be useful for any significant current without cooking themselves.