a 30 year cranklight

zamboniman

Newly Enlightened
Joined
Jun 9, 2007
Messages
31
I've been thinking lately about how to make a decent, high quality cranklight that could be reasonably expected to have a 30 year lifetime. here's what I came up with. comments/improvements? I chose a linear regulator because its simplicity should give the light a longer life, fewer parts to break...
flashlightplan009.jpg
 
Looks cool but why not go with an array of high output Nichia 5mm that you can select 1 or more for multiple outputs/runtimes? If you want higher output maybe swap the Luxeon for a more effecient Cree or SSC and lower the amount of juice it receives for longer runtime?

-LT
 
Great idea but I too would go with Nichia LED's. So the generator and gear section will be flooded or just the gear section?
 
the gear/generator section would be flooded with light oil, and the rest of the device would be resin potted, allowing for the whole thing to be waterproof to almost any pressure.
 
in addition, the way the switch system is set up:
Low = 1 LED
Med. = 4 LEDs
High = 1w LED
in terms of comparative brightness
 
Yes this looks good :)
Have you found a generator power plant "off the shelf" you can use?
Being completely sealed like that is great! I really hope this can become a reality. I'm really into "no batteries" generator powered devices.

I made two interesting generator type lights also...
My Crank Generator MagLite Modification [With pictures][Now with YouTube Video!]

Keep us up to date on what you do next and Don't Give Up!
The potential for a good generator crank light like you have in mind is awesome. (And there could always be an international market for such a product, in regions where long term emergency preparedness planning and institutional survival kits are more common. )
 
I intend to use an AC motor with gearbox as the generator, and the prototype would be plastic since I don't have access to aluminum anodizing and machining equipment... I'll be sure to keep everyone posted.
 
Just wondering if there is a reason you are using germanium diodes instead of shottky ones?

If the supercaps are similar to electrolytic ones (i.e. wet inside) they may not last 30 years before drying out.

Greg
 
Although schottky diodes can have a lower forward voltage drop than germanium diodes, schottky diodes range from .15 to .46 volts forward voltage drop, whereas germanium diodes are constant at about .3 volts. This means that the circuit can be properly designed without having to assume the lower forward voltage drop (to prevent overvolting the capacitors) and squandering energy and reducing runtime if the voltage drop of the diode bridge is higher.
 
Just wondering if there is a reason you are using germanium diodes instead of shottky ones?

If the supercaps are similar to electrolytic ones (i.e. wet inside) they may not last 30 years before drying out.

Greg

Yes, failure and leakage of electrolytic caps is a common problem in older electronics. A direct drive generator that can't produce enough current or volatage to burn out LED's even if you wanted to might be the way to go... keep it simple stupid :D


[As an interesting aside story, in the last 5 years lots of cap failure in small portable audio electronics like sony discman's and walkman's are showing up. (stuff that is now 20+ years old all of a sudden... but still valid and exquisite quality!) Famous top model sony discman's like the D-555 often require new caps in the amp stage.]
 
Thanks for the answers guys.

Not much electronic equipment is designed to last 30 years. There are all sorts of failure modes that you might not think of that can happen after 10, 20 or 30 years. I heard that the amplifiers in under-sea cables were designed to last 50 years. I don't know if they still do that or if they have fibre with low enough loss to reach across the ocean without (electronic) repeaters/amplifiers.

Greg
 
For that sort of reliability, you will either have to carefully match the super caps or put a voltage balancing circuit across the series pairs - not too difficult to do with an op-amp. - or add individual 2.5V shunt regulators to each cap.

Super caps have liquid inside, so they might not make 30 years anyway.

Also, I realise there are plenty of magnetic switches that can handle the current, but how about doing the main power switching by turning off the FET.
Contacts in the power line are always a source of unreliability.

You could put a normally closed magnetic contact in the 'off' position and put it in series with the gate drive resistor.
The magnet will cause that circuit to break when you move it to 'off'.
- that reduces your reed count to 3, (2 standard and one normally closed (or type C - changeover - which are easier to get)

And also - where are you going to get magnetic bearings from?
I have limited knowledge of these things, but I think the ones that can do this sort of locating job are electromagnets and have feedback electronics - they are power hungry.
A set of sealed ball bearings will last a heck of a long time and offer very little drag.
Or you can buy ceramic ball bearings.

>the gear/generator section would be flooded with light oil, and the rest of the device would be resin potted, allowing for the whole thing to be waterproof to almost any pressure.

Put plenty of turns on the coils - and even move to three phase - plus rare earth magnets and you wont need a gear box. You will get enoughpower from a crank directly on the generator spindle.

Steve
 
Last edited:
Thanks for the feedback, everybody. I wonder how hard it would be to custom wind the generator... I wonder if you could get a similar runtime with a big flywheel... that would make the thing really long lasting -- you could direct-drive an LED if you designed it properly. does anyone know of a generator that outputs a stable voltage?
 
Pure guess mode here, would any of the bike hub generators suit you requirement?, obviously it would need to be geared up. The generator should already be waterproof.
Norm
 
Top