The Genny


Twin rotor axial flux alternator with 50mm dia rotors
6 10mm(d)x3mm(h) N42 NdFeB disc magnets per rotor
Air Gap; approx 5.5mm
Stator: Single phase, 3 coils seriesed, 700 turns 32 SWG each coil, 4mm thick
Drop: 1470mm
Weight: 670g
Drop/run time: 47seconds
Rotor RPM: approx 200-225
Load: 3mm White LED with 660ohm resistor
Output: 7.6v @ 7.05mA = 0.05359 Watts or about 1/13920 th Horse Power

More tinkering

At the moment, there's a 100uf capacitor across the + & - on the rectifier and the LED visably flickers... prob something to do with very little amps supplied & too big a capacitor for the LED current draw. I'll try various resistors with the LED... I'm thinking around 2k to 2.5k ohms to bring the LED current draw down too 3 or 4 mA (if it can run on that little). Hopefully the cap will charge quicker and stop the flickering.

I changed the 24 tooth gear at the top of the chain link for a 16 tooth gear, bringing the overall gear ratio to 1:37.5, the weight was increased to 820g . I swapped the 100uf cap for a 47uf cap and ran again.
The cotton tape slipped throught the rollers... another weight brick on the tensioning craddle stops the slippage :)
Drop: 1470mm
Drop time: 67 seconds
Weight: 820g
Rotor RPM: approx 180-210
Output: 7.00v @ 6.29mA = 0.04403 Watts
I think the weight and gearing is about the max the Lego structure can take...
820g @ 15mm is a fair amount of torque for Lego!
I'll try changing the LED resistor and cap to see if I can get a longer runtime.

Doh... the above is true for the load!!!
How could I forget a simple thing like input torque=output power (less losses)

I changed a couple of things, 2.2k ohm resistor on the LED, and used a 680uf 35v capacitor.
Drop: 1660mm
Drop time 97 seconds :)
Weight: 410g
Rotor RPM: approx 200
Output: 7.57v @ 2.27mA = 0.01718 Watts

I was using a meter with a frequency counter to calculate the RPM above... the meter is out by alot!

The cap carries enough charge to let the LED glow for a few seconds after the alt stops.

From what I learned about the max torque, I think I can gear up some more as the weight this time is alot less (1/2!) I can go to either 1:69.444 or 1:78.125... currently the gearing is 1:37.5
I'd have to add more weight for the higher gear ratios... but the overall effect would be a longer runtime.

Another way, I think I can get a longer runtime, is to increase the capacitor size... one that would allow the LED to stay lit for enough time to reset the weight.
I have some 2.5v 10f and 5.5v 1f caps, as well as several in the 2-3000 uf range.
I've already tested the 5.5v caps, 4 in parallel will keep 2 5mm 10,000 mcd white LEDs goings at a starting draw of 5v @ 60mA for about 55mins... I don't think they're suitable for this app.
I think I'll try a parrallel/series array of some of the 2-3000uf caps.

Yet more tinkering.

I've had a couple of days tinkering and tweaking, some results are here, Weight driven generator - Yet more fun with Lego

I upped the gearing to 1:5, 3:5, 1:5 which gives an overall gear ratio of 1:46.666.
I changed the rectifier to a schottky and regular diode setup.
I ditched the resistor for the LED and used a 5k ohm trim pot... the resistance is set to limit the LED current draw to 1.5mA.
I lost the caps and used a 3.6v 300mAhr nicad pack.
And added tension weights to either end of the tape.

Drop: 1570mm
Drop time: 167 seconds
Weight: 550g
Tension weights: 180g
Output: 4.20v @ 4.11mA = 0.017262 Watts
As the weight is dropping, not only is it powering the LED (1.5mA), the battery pack is being charged as well.
The useable light from the LED lasts for another 159 seconds after the weight has grounded. The nicad pack is down to 3.05 volts and fading quick... the LED draws a mere 0.5mA
Dinges, over on The otherpower discussion board provided some efficeincy conversion formula...
The efficiency of this setup works out like this.
m = 0.55 kg
g = 9.81 N/s^2
h = 1.57m

Epot= 0.55*9.81*1.57 = 8.47 J

Epot= 8.47 J

Energy to NiCad:
Q=P*t [J]

P=U*I = 4.20 *0.00411 = 0.017262W
Q=P*t=0.0173 * 167 = 2.8891 J

2.89/8.47*100= 34.12% efficiency ??? :)

So... what's next... mmm...
Well when I started this, I had in mind something that could be used as a light source for a night time toilet visit... 6 minutes, 26 seconds of useable light covers that :)

I have an idea for something bigger & better ;)

If you try this at home, please take a sensible precaution to safety.
The NdFeB magnets are brittle.
You are responsible for your own safety.

prev page

This page is part of a frameset, Click here to view the page as it is meant to be seen