Human powered generators
2004-06-17 17:50:07.409167+00 by
Dan Lyke
7 comments
I promised Mars that I'd put this up here, and then it slipped my mind (and this week has been madness, anyway).
Mars has an idea for a Burning Man project, and he'd like input and assistance from folks who know something about electricity. There's a large component of this which involves human powered generators, and all of the usual power supply questions come up. Can anyone help him?
[ related topics:
Burning Man Cool Technology
]
comments in ascending chronological order (reverse):
#Comment Re: made: 2004-06-17 23:35:35.068228+00 by:
meuon
[edit history]
Musing after a long and hard couple of days:
Two LONG rollers..10-20ft-??... with a bike rack arrangement in front on it..
Park your bike fron in the rack, put your back wheel on the rollers,
hop on your bike and pedal with everyone else, helping power the rollers, which powers the generator/alternator which powers the ????
Why duplicate the bike, which everyone (almost) already has? Simple is good.
#Comment Re: made: 2004-06-18 01:37:17.499552+00 by:
topspin
meuon, I like the idea, but folks are gonna pedal at different rates and that's gonna create some issues on a single long roller.
I stumbled on this device which seems pretty compact and simple. Perhaps it'll help fuel the discussion.
#Comment Re: made: 2004-06-18 17:02:02.41817+00 by:
mvandewettering
Hmmm. How many twinkies do you think it will take to get a kilowatt/hour?
#Comment Re: made: 2004-06-18 18:18:49.849701+00 by:
Dan Lyke
This should be easy. The assorted web pages on generators say someone in good shape should be able to hold 100 watts for a 20 minutes, we'll back of the envelope and say that's equivalent to holding over 20MPH on a road bike. This table claims a 175 lb person will burn just under 1400 calories/hr at 20MPH. So if a Twinkie is 150 calories, that's 10 people at 9.3 Twinkies each, or roughly 93 Twinkies per kW/hr. If you deep fry the Twinkie, you can get down to 33 Twinkies per kW/hr.
I haven't looked at the retail price of such things in a while, but I seem to be getting a $1.00 each vibe on up-front costs. Going with the original 93 Twinkie estimate, that means you're paying between $93 and $930 per kW/hr.
#Comment Re: made: 2004-06-23 06:01:38.962657+00 by:
Mars Saxman
So my basic worry here is that some super-fit person is going to come along and crank their guts
out and burn out the light bulbs. I could just add more bulbs, but then the resistance increases
and it gets harder to pedal the bike. So, how hard is it to build a circuit that limits current to some
specific amount?
Another idea I had was to hook a 12V motorcycle battery into the circuit, so that the bike charges
the battery and the battery powers the light. But the same problem occurs: how do you level out
the power the user produces, so that the battery charges steadily and people don't try to pedal too
fast? Also, how do you keep the battery from driving the motor - is that just a question of
mounting diodes between the generator and the battery?
finally: if I am running the lights off a battery, I would like to not waste power during daytime. I
imagine I could do this with a photocell. how complex a circuit is that? how would a clueless
person such as myself learn how to make this work?
#Comment Re: made: 2004-06-23 16:37:04.020905+00 by:
Dan Lyke
I am not an electronics geek, but my first naive pass at a voltage regulator would look something like:
The Zener Diode D1 has a breakdown voltage of about the top that your lighting circuit can handle, which means that it'll start conducting away from the arrow after the input voltage gets higher than that, that'll draw current away from the PNP transistor Q1 which will shut off flow through that, shutting off flow through the NPN driver transistor Q2.
There are two transistors there because I think you'll need more of a current flow step-up than you want to drive through the Zener diode, so this builds an array like a "Darlington" transistor, so that Q1's input can be relatively small and Q2 can be some big honkin' thing. Both are protected by D2, which I originally wasn't going to put inline with the power, but I got to thinking about the potential for backward leakage and realized that since you're building a small number of these it's easier to just get a big honkin' diode rather than trying to use reverse diodes in parallel with the transistors for bleeding off surges from your load.
R1 and R2 are appropriately sized for your zener diode and your two transistors, I'm too lazy to go through and guess at voltages and current draws for whatever else you might be using, and you can figure out v=i/r as easily as I can.
The net effect of this should be that at some point when you pedal fast enough it gets easier to pedal. You could reverse this idea to maybe turn on an extra set of lights when the voltage gets high enough, which would make it harder to pedal (ie: "fire up the Kleig lights...").
And I'm suggesting discrete components on this puppy because I'm figuring that you'll go to Digi-Key and spec out transistors that'd be way too expensive to use in quantity (ie: maybe a whopping 5 bucks for the main driver transistor) but have huge margins for error in terms of voltages (ie: build your system for 12v and get 125v or 250v capable transistors) rather than using something integrated that'll work flawlessly to 35v but turn to so much sand when someone over-revs your generator.
The simplest idea I could come up with for a regulator for daylight is something like:
Which will draw a little bit of power during the day, but if you size R4 appropriately that can be a milli-Amp or so, and off of batteries like you'll want to drive big lights won't be noticeable. You'll probably also want to run this through some sort of step-up transistor array as in the other circuit depending on what Darlington PNP transistors you can lay your hands on for Q5, you want to draw only as little current through Q4 as you have to (by making R4 as large as possible) so you don't drain your batteries during the day.
You'll probably also need to experiment with shielding and filters to make Q4 less sensitive, depending on what sort of phototransistors you come up with.
#Comment Re: made: 2004-06-23 17:53:58.913355+00 by:
Dan Lyke
Oh, it occurs to me that if you do this whole thing as a 12v circuit then automobile headlights aren't terribly efficient but might make the perfect lamp for what you're trying to do.
And have you considered perhaps a fan to blow a bit of playa dust up the column so that the light has something to reflect off of? That way you don't have to be putting out enough energy to make the air glow...