# Making a bicycle powered Christmas Tree prop

#### toozie21

Joined Oct 4, 2012
59
So I've been kicking around making a bicycle powered Christmas tree prop for the front yard. I've got a VERY old stationary bike that I am thinking of using. My plan is to take it apart and take the pedals off of the crank arm; then someone can use their hands to turn it and drive the lights. What I am curious about is if my electronics concepts make sense....

My play is to hook a dynamo up to the drive chain. Then coming off of the dynamo's wires would be a diode to make sure they can only power the lights in one direction. Following that would be a current limiting resistor. Next would four partial strings of Christmas lights in parallel with different power resistors connected to them to change the load on each string. Each string would have its own zener diode that would require different voltage levels to run each string of lights (say 6V, 12V, 18V, and 24V for instance).

My plan is that as the user turns the crank arm faster, the dynamo spins faster, causing the voltage output of the dynamo increase. Each string of lights would have a minimal voltage required to turn them on based on their associated zener diode (listed above) and would turn on in order as the user spins (and maintains the spin) of the shaft.

Does this make sense? Am I overlooking something (totally probable)? I am planning on using LED Christmas lights as they run on much lower current.

#### Alec_t

Joined Sep 17, 2013
11,504
It takes quite a lot of effort for a human being to generate more than a few Watts for a prolonged period, so don't count on powering many lights, even LEDs.
Providing you can source a dynamo able to generate at least 24V the concept should be workable, but is likely to need a bit more electronics than resistors and zener diodes.
Have you already lined up the volunteers to crank the dynamo for hours on end ?

#### toozie21

Joined Oct 4, 2012
59
It takes quite a lot of effort for a human being to generate more than a few Watts for a prolonged period, so don't count on powering many lights, even LEDs.
Providing you can source a dynamo able to generate at least 24V the concept should be workable, but is likely to need a bit more electronics than resistors and zener diodes.
Have you already lined up the volunteers to crank the dynamo for hours on end ?
Thanks for the quick feedback!!! I guess should have mentioned that my "tree" would probably be a flat cutout with the small sections of strings hanging on it.

Agreed on the wattage, which is why I was thinking of using cut down pieces of lights. Let's say about 10 lights per string, four strings total. My LED light strings draw 9.6W per 100 lights, so my mini strings would draw ~ 1W per string. So I imagine someone could turn out 1 or 2 watts pretty easily (totally a guess), but it would take a little more effort to light the third string and maybe superman to light the last string.

The lights wouldn't be running normally, just have a sign as an invite to try to light the tree and maybe a short write up on how it is working (for the little engineer's sakes).

#### wayneh

Joined Sep 9, 2010
16,399
My play is to hook a dynamo up to the drive chain. Then coming off of the dynamo's wires would be a diode to make sure they can only power the lights in one direction. Following that would be a current limiting resistor.
Those circuit ideas don't make a lot of sense until you identify the actual power source, including its voltage, amperage capacity, output waveform (AC versus DC) and so on, and the lighting you want to use.

I would definitely choose all LEDs for this project, since you can get so much more light for a given amount of power. I wouldn't plan on more than about 20W of total load. A healthy human on a bike can sustain 100W for long times, but that's using the body's largest muscles. With just arms, with cold hands and a coat on, I just don't think you can expect much more at the output. There will be losses in the regulation and so forth.

I'd be tempted to use commercial LED lighting and plan to supply it with an inverter, the type you use in your car. The challenge then would be to provide the inverter enough DC power to make it happy, perhaps using a DC-DC converter to feed it. There are many other strategies that might be more efficient, but this solution would allow you to focus on just one design issue – getting your generator to make ~14VDC. Your inverter and lights would be off-the-shelf items that could go into normal service once your project is complete.

Joined Apr 30, 2011
1,569
Consider that since the crank arms of the bicycle were designed for the geometry of adult legs, the distance between the arm ends may be too great for efficient hand cranking, especially by smaller-statured people such as children.

#### GopherT

Joined Nov 23, 2012
8,012
@toozie21

Here is an example of the effort. The bulbs are 15W each if I remember correctly. This is not my video - it came up on a search of the Henry Ford Museum - Deerborn, MI.

#### toozie21

Joined Oct 4, 2012
59
@GopherT Thank you for the link, that is pretty cool. I had previously seen a video on a similar design in a MN museum, but that was bike based, not hand turned. I did find someone's blog post that mentioned that the bulbs were 40W each (take it with a grain of salt). I would think that my 1W strings would be pretty easy to light up based on that.

#### #12

Joined Nov 30, 2010
18,210
I think a person could do 10 watts, sustained, with their arm(s).
I think you're perfectly OK with (4) 1 watt strings.

At first, I thought you meant legs, and did the math for that.
Athletic person at peak is 1/4 HP (186 watts)
Cut that by 80% for non-athletic people trying for a sustained output.
Cut that by 75% for arms instead of legs and I get 9 watts for arms without breaking a sweat.

#### atferrari

Joined Jan 6, 2004
4,069
I think a person could do 10 watts, sustained, with their arm(s).
I think you're perfectly OK with (4) 1 watt strings.

At first, I thought you meant legs, and did the math for that.
Athletic person at peak is 1/4 HP (186 watts)
Cut that by 80% for non-athletic people trying for a sustained output.
Cut that by 75% for arms instead of legs and I get 9 watts for arms without breaking a sweat.
And then, no matter what of the above, after one minute for anyone (pedaling?), two words will come to mind: boring and tiring. Short period of popularity.

#### toozie21

Joined Oct 4, 2012
59
I think a person could do 10 watts, sustained, with their arm(s).
I think you're perfectly OK with (4) 1 watt strings.

At first, I thought you meant legs, and did the math for that.
Athletic person at peak is 1/4 HP (186 watts)
Cut that by 80% for non-athletic people trying for a sustained output.
Cut that by 75% for arms instead of legs and I get 9 watts for arms without breaking a sweat.
Thank you for the math, that really helps. I wasn't sure how to calculate it until my dynamo came in and I did some playing around.

And then, no matter what of the above, after one minute for anyone (pedaling?), two words will come to mind: boring and tiring. Short period of popularity.
I am not planning on having people do it all the time, just a neat little prop for someone to full around with for a minute or two while out seeing my other displays. I figured a little write up on what is going on next to the bike and some kid might learn a little something while out to see some animated displays...

#### toozie21

Joined Oct 4, 2012
59
Another thought is to punt on the direct powering of the lights and use a micro instead. I could do a voltage divider and read that into a micro and switch on the lights at different levels. Instead of a voltage divider a voltage to frequency converter IC could maybe used through a LPF and then read that into a micro via the ADC. Some sort of IC that has adjustable voltage levels would be sweet, but I don't think that those exist.

#### djsfantasi

Joined Apr 11, 2010
6,692
toozie21 said:
Some sort of IC that has adjustable voltage levels would be sweet, but I don't think that those exist.
Wouldn't an LM3914, or two or three, work for you?

#12

#### wayneh

Joined Sep 9, 2010
16,399
Another thought is to punt on the direct powering of the lights and use a micro instead. I could do a voltage divider and read that into a micro and switch on the lights at different levels. Instead of a voltage divider a voltage to frequency converter IC could maybe used through a LPF and then read that into a micro via the ADC. Some sort of IC that has adjustable voltage levels would be sweet, but I don't think that those exist.
That's a good idea, to sense the presence of the user rather than require power from the user. That would also allow for the display to continue for a while without constant input. Maybe a slow ramp down, fade to black. A slow fade might encourage more interaction than a sudden blackout.

There are really cheap remote controllers for LED strings out there. They allow you to change the color and brightness. You might hack one of those.

#### #12

Joined Nov 30, 2010
18,210
A slow fade might encourage more interaction than a sudden blackout.
Peak detector into a capacitor, then to an LM3914

#### toozie21

Joined Oct 4, 2012
59
Peak detector into a capacitor, then to an LM3914
@#12 Since the peak detector is a diode into a cap, I assume the second cap you mention could be dropped, right?

What would you do on the output of the 3914? Drive the lights directly? Drive some SSRs?

#### #12

Joined Nov 30, 2010
18,210
I assume the second cap you mention could be dropped, right?
I didn't mention a second capacitor, so, right, don't use it.
What would you do on the output of the 3914? Drive the lights directly? Drive some SSRs?
Add transistors to get enough current for the lights.

The 3914 is a data processing (thinking) chip, like a microprocessor. Thinking chips suck at power driving so they almost always need transistors as power amplifiers.

#### toozie21

Joined Oct 4, 2012
59

You thinking something like a darlington transistor pair to power the light? Something like the ULN series?

#### #12

Joined Nov 30, 2010
18,210
You thinking something like a darlington transistor pair to power the light? Something like the ULN series?
I haven't found the current requirements of the LEDs, but a 2N7000 mosfet comes to mind.

#### Alec_t

Joined Sep 17, 2013
11,504
The LM3914 outputs are low-side constant-current sinks, so it might simplify the circuit logic to use them to control high-side (PNP/PFET) transistors rather than the NFET 2N7000.

#### #12

Joined Nov 30, 2010
18,210
The LM3914 outputs are low-side constant-current sinks, so it might simplify the circuit logic to use them to control high-side (PNP/PFET) transistors rather than the NFET 2N7000.
Sorry. I didn't bother to look up the chip for its normal outputs before I posted.
So, yeah...pfet would be the answer.