[time-nuts] science projects

Jim Lux jimlux at earthlink.net
Fri Feb 10 01:12:46 UTC 2012

Yes, a fun project, and a fine physics lab exercise.  But not a good 
science fair project because it doesn't meet the "originality" bar.

This is something that I confess I had a hard time figuring out what 
that meant when I was entering science fairs... as it happened, my 
projects *were* original (reviewed in retrospect), but I couldn't figure 
out how you'd evaluate it at the time.

In general, if the general idea for a project came from a published list 
or book, that would be ok. But if the method of attacking the problem 
also came from a book or the web, that wouldn't.

So a good topic to prompt projects would be "Measure the speed of light 
in a novel way".

If you were to do Roemer's technique from the 17th century, but do it in 
a clever way (webcams, telescopes, etc.) I think that would do ok in the 
junior division at least.  (i.e. removing the human measurement element 
is a good experimental refinement on the basic technique).

Or if you were to use some other extra terrestrial event as the 
predictable time hack at a varying distance.

Here's a wild one.. Set up  detectors some distance apart where an 
Iridium Flash will be visible (http://www.heavens-above.com/)  The flash 
comes from the satellite, and the path length to the two detectors will 
be different, so you will see the pulse at a different time.  By knowing 
the distance between the detectors and the angular displacement of the 
image with reference to something, you could figure out the distance to 
the satellite, and therefore, the difference in propagation distance 
between satellite and each sensor.

But these wouldn't hack it in Senior division.

On 2/9/12 7:19 AM, Brooke Clarke wrote:
> Hi Jim:
> Check out:
> http://www.ted.com/talks/lang/eng/clifford_stoll_on_everything.html
> Here's part of an email from Clifford:
> "Oh, the speed of light?
> For detectors, I use two fast-response photodiodes and feed their
> outputs into opamps.
> The experiment uses one cheap laserpointer (they cost a buck or two on
> ebay - I buy 25 at once, and then select the ones with the fastest
> switching rate.)
> The laser pointer is fed by a signal generator (I use an HP 3312A); the
> square wave output connects to the laser pointer with a pair of
> clipleads. I start with the laser switching about once a second, to show
> the kids that it really is turning on and off. THen I boost the speed to
> a few hertz, then tens, and hundreds of hertz. Eventually, I get it
> switching on/off at 2 to 5 MHz.
> I aim this switched laser beam at a distant mirror (across the room -
> maybe 10 or 15 meters away). Coming back along almost the same path,
> this long-path beam then hits a lens which focuses the beam onto one
> photodiode. The photodiode feeds its pulses into an opamp, and then into
> the bottom trace of a dual trace 100MHz oscilloscope.
> I then create a short-path beam using a beamsplitter that's right next
> to the laserpointer. I use a piece of microscope slide cover-slide for
> the beamsplitter. A few inches from the beamsplitter I set the
> short-path photodiode, which goes through the 2nd opamp and into the
> upper beam of the squigglescope.
> There'll be a time delay in the arrival of the long-path beam; I get the
> students to measure this time difference. Then we measure the path
> lenght difference with a tape measure, do a division, and out pops the
> local speed of light.
> Lots of gotcha's ... for instance, I make sure that I trigger the
> oscilloscope on the output of the function generator. Also, alignment is
> very difficult, and requires rock-solid furniture (seldom found in
> classrooms).
> It's a fun project with high school students, because we spend a long
> time afterwards discussing errors and problems in the system ...
> typically, we measure things about 20 or 30 percent off. The main source
> of error turns out to be mismatched responses in the two photodiode/opamps.

More information about the time-nuts mailing list