Weird problem with alternator type design

[(unknown)]

Hello to everyone,

I just build a small alternator using 1/2" Neo's and 26 guage coils wrapped 200 turns. I have 12 magnets equally spaved in a circle around a DC motor. The motor is spinning a circular plastic disc which has 6 coils attached that are about 1/8" away from the magnets. They are connect in single phase series and is rotating around 2000 RPM's and is producing 12volts AC and around 600 milliamps. I convert this to DC and use it to power a circuit that is connected to the plastic disk. The circuit draws 5volts DC and aroudn 300 ma. Everything works great for about 5 minutes and then the lights on my circuit start to go dim and then everntually turn off. Its like its losing power slowly over time. Can anyone explain this strange behavior to me??? If its producing more than enough voltage and current normally then why would it lose power over time? Thanks!

[TomW]

Bob;

Are you sure its the alternator?

Could it be the circuit quits for some reason?

Does the circuit work from another DC source properly?

Can the alternator power something like a light longer than the circuit?

What does your voltage from the alternator read when the circuit fails?

Anyway you don't give much info. "power a circuit that is connected to the plastic disk" is pretty vague.

Very difficult to troubleshoot from remote with no real information.

Sounds like one of those over unity type projects gone awry from here.

T

[drdongle]

Correct me if I am wrong but are you trying to make a "perpetual motion" or "over unity" device? If so it won't work. there is no free lunch.

Carpa Vigor

Dr.D

[(unknown)]

I don't understand what you mean by unity circuit? Could you explain?

The circuit is a Microcontroller that control LED's that spin and create a clock. Since the circuit board is spinning there is no real way to get power to it other than using this type of design. I got the idea from reading this site and the way windmills generate power.

I can't measure the voltage while its spinning since there is no "safe" way to connect leads to it. Spins too fast.

I am thinking that there is a problem with "eddy currents" or am I off base with that thought?

The circuit checks out ok because if I use batteries it will last for about 48 hours nonstop. That is too short and would require an endless supply of batteries so that is why I am using this method. If you draw a load off this type of design for a long period would it work reliably? I see that most of the people here seem to be charging batties and not sure if that is different that using the power right off the alternator...

Thanks again for you help..

[TomW]

Bob;

You may be seeing issues from an unstable dc voltage if the output from the rectifiers is not filtered. Adding capacitors will smooth that ripple.

T

[(unknown)]

I have a 470uf electro cap placed right after the rectifier. I just was running it again without powering the main circuit and just have it light 1 LED which shouldn't draw much at all. The light stayed on but I do notice an odd smell from it after about 5 minutes running time. There are no parts that are getting hot, not even warm, so I am not sure where that smell is comming from. Its not a burning smell but a electronics type odor. I am wondering if the motor which is a brushless DC type is having problems with the magnetic field being generated right outside its main body. The motor is using a magnet inside as well. The motor isn't losing any speed though... This is strange..

[TomW]

Bob;

I would swap out the capacitor they are sometimes sensitive to voltage spikes that you may not realize are there. if your caps are rated 12 volts or 15 volts i can almost guarantee they are seeing more than that from the alternator. I would go for 35 volt rated caps just for a buffer.

Of course from here its a crap shoot but caps do smell kinda weird when they fail especially electrolytics. A leaky cap could kill efficiency on the circuit too as it leaks voltage across its plates.

T

[jacquesm]

you might be getting too much rather than too little voltage, try stabilizing your voltage using a 78xx voltage regulator !

An led needs a resistor in series with it to limit the current to something < 5 mA if you want it to survive.

Your circuit may be failing because it gets too much voltage and heats up somewhere.

success !

Neat Idea though, I was talking something like this over with my friend Johannes a while ago and we figured a windmill would make a really cool clock ! We hit on your solution to power it, and I'm pretty sure that it can be made to work.

Your motor should not have any problem at all with a magnetic field outside its case since the case effectively shields it.

regards,

  Jacques.

[Flux]

I am inclined to agree with the others that you may have too many volts.

You could get an idea of the volts on your power supply by trying various zener diodes in series with a LED and a limit resistor.  The highest zener that will light the LED is about your supply voltage.

If you can alter the motor speed you could slow it down and see if that helps the problem.

Flux

[drdongle]

Certain parts of your original description were vague so I thought perhaps you were trying to make a "over unity device" , a device that make more power than it uses.

Obviously that is not the case

Carpa Vigor

Dr.D

[(unknown)]

I hooked it up backwards in the test phase so I could measure the voltage. I was spinning the magnets instead of the coils. I measured 12volts max when the motor was at full speed. Then I did a current check and it slowed the motor a little bit which I figured was a build up of magnetic field and causing a braking effect. I measured 500ma on that. All that aside, The electrolyte is rated for 25v and I guess could be the problem. The way I hooked the LED up was to put it through a reisitor to limit current to 10mas. I do use a regualtor of 6v to stabilize it before feeding the main circuit. I will try swaping out the cap with a higher voltage one...Does it matter if it electrolyte or not? Which type would you recomend? What value?

Regards,

Bob

[jacquesm]

Ok, so you have coil, rectifier, large capacitor, small capacitor, voltage regulator, small capacitor, larger capacitor and then your load.

The values are all very much dependent on the requirements of the load and the input voltage of your coil.

If the measurement you made was an AC measurement then you'll have to triple that for the working voltage of your Capacitor, those are MAXIMUM rationgs, not operating conditions. So if you have a 12V MAX AC voltage I'd say you'd need 50 V caps, at 500 mA use 2200 uF on both sides of the regulator (and 10 nF to quell any oscillations)

Voltage regulation needs about 3V more than the 'output' voltage at a minimum to work.

Reverse breakdown voltage of your rectifier should be at least 3x input voltage as well.

Maybe you could post a schematic of the way you have things arranged and/or an image ?

[RatOmeter]

Yeah, he should have been a bit more clear about the nature of the circuit at the start.  

I've seen others like this, even commercially produced units.  It's been a while, but at least one that I read about used slip rings to transmit the power to the spinning circuits.

Here's an example (I didn't read up on this one, so don't know how they solved the power problem):

http://pages.sbcglobal.net/ken_staton/electronics/prop_clock/p_clk.html

[ghurd]

My LED experience shows LEDs don't smell before failure.

Run it and see if the 7806 is warm.

The 7806 is disapating about 2.5w. Most of them have a thermal shutdown. Is the thing heatsinked or potted? If the heat can't get out it could be shutting down.

Just a thought. Could be I'm thinking about it wrong.

G-

[(unknown)]

No nothing is even close to being warm. That the strange part. So to at least rule out the magnets/coils, there is no strange phenomena that happens when they produce power over a period of time? Nothing that would cause it to start to generate less as its working? Is that is not the problem then I would lean towards the cap failing. If that is shorting out on the inside, it would rob the circuit of power correct? Would a cap failure cause immediate shutdown or do it gradually?

[ghurd]

I don't know. Probably depends on how it failed.

But I have got surplus electrolitic caps that didn't work when I got them.

I later read that they will 'dry up' if they are not used regularly, but it takes many years. That was new to me!

G-

[(unknown)]

I just heard from someone who saud that elctrolytes have a gel inside them that could be migrating due to the spinning and then shorting out the cap. Makes sense to me so I will swap it out. Many thanks to those who've helped.

[commanda]

Put a heatsink on your regulator. They automatically shut down if they get too hot.

Amanda

[drdongle]

If that is the case try a Tantalum cap, their solid.

Carpa Vigor

Dr.D

[ghurd]

I used tant. caps for some small stuff.

The problem I had for everything else is I could't find Tant. caps big enough.

[finnsawyer]

If you're spinning the circuit board, you could be having a mechanical separation of some kind due to the centrifugal force.  This could cause an increase in resistance in a lead or solder joint.

[(unknown)]

I can see that happening too but if I hook it up using 4 AAA cells it works until they die out which is around 2 days or so. So I would have to rule that out. I replaced the cap last night with a different type but smaller size and it is not getting the right amount of power to start now. So I think I have a problem with the magnets and coils not being as close as they should be. On their own without a cap to store a charge they will not deliver all the power needed. When I did the testing part and had the magnets spinning instead of the coils I was getting plenty of power so the only difference between the test setup and now is that the gap between the magnets and coils increased a little. I guess that plays a major part in generating power.