2008QuarkNetGWinters

From MariachiWiki

This is Gillian's blog from the 2008 QuarkNet workshop, held at Stony Brook University.
Return to Gillian's main page.

1 Day 1: Monday July 14, 2008
2 Day 2: Tuesday July 15, 2008
3 Day 3: Wednesday July 16, 2008
4 Day 4: Thursday July 17, 2008
5 Day 5: Friday July 18, 2008

Day 1: Monday July 14, 2008

My understanding is that QuarkNet is a Fermi Lab program, designed to provide enrichment to high school physics teachers. The emphasis is on particle physics, but this week we will build and test a handful of experiments, finishing with cosmic ray scintillation detectors.

We started with donuts

Helio started by talking briefly about modern physics, and its uses. Applications of note:

General theory of relativity: GPS
Maxwell's equations: cell phone, radio
Quantum methanics: semiconductors
Photoelectric effect: CCDs
Particle physics: world wide web was initially created mainly for particle physicists.

We had a very short introduction to writing wikis. Those of us who have participated in a Helio-workshop have already posted material on the MARIACHI wiki, while others had not. The next session was an introduction to Excel by Harry Stuckey. He provided data from a Helmholtz coil, from which he asked us to calculate the mass of an electron. Really, we were supposed to find the mass/charge ratio, but we took the known value of the charge on an electron to calculate the mass. Instructions are here, with links to an Excel file with data and specific instructions. Effectively, we are given a set of magnetic fields and the resulting radius (diameter, but divide to find the radius) of the electron beam. With the electron beam in a circular path, the magnetic force is what provides the centripetal force, so equating the magnetic force with the centripetal force and doing a little algebra leads to $B = \sqrt{{2Vm} \over e } \left( \frac{1}{r} \right)$ . Graphing B vs. r leads to a nicely curving inverse curve. Graphing B vs. (1/r) leads to a linear curve, and the slope can be used to find the mass of the electron (all other quantities are known). I calculated $9.18 \times 10^{-31}kg$ , a difference of 0.7% from the accepted value of $9.11 \times 10^{-31}kg$ . That's not bad. I think the original data in the Excel file that Harry gave us may have come from actual measurements taken on the lab Helmholtz coil. If so, the results we got would lead me to want a Helmholtz coil for my classroom. I wonder where you can get funding for that ...

In the afternoon we started assembling our apparatus to calculate the speed of sound. More on that tomorrow.

[edit]

Day 2: Tuesday July 15, 2008

We continued assembling our apparatus to calculate the speed of sound. It's a little more high-tech than having a student pop a paper bag from a distance of one football field away and timing the time it takes the sound to reach the observer.

Here we have a PVC pipe, put a speaker at one end, and pick up the sound using two microphones at either end of the pipe. We use a digital oscilloscope to find the difference in time between the two microphone pick-ups. First we drilled holes at either end of the PVC pipe for the microphones. The holes in mine were 137cm apart, and a tight fit for the microphones. 5-minute epoxy holds them in place. Apparently the leads to the microphones are quite delicate, so I should eventually put little protectors (bottle caps?) over them for protection. Then we drilled a hole for the speaker wires in the center of the cap on one end of the PVC pipe and glued the speaker in place, again using 5-minute epoxy. Note: The epoxy has peeled off the PVC on mine, so I will need to reattach the speaker. Finally, we soldered parts onto a little circuit board: resistors, capacitors, and leads for a 9V battery, the two microphones, and output line to the external sound card. The last steps were to tape the parts in place and then test it.

To run the speed of sound experiment, first test the signal generator, then plug things in, then check settings on the Control Panel, and finally run the Scope program.

Test the signal generator:
- Run Audacity
- Load Helio's speed of sound waveform (download Media:speedofsound.doc then rename to speedofsound.wav)
- Play. You should hear a regular ticking sound from the computer, or if it's hooked up to the speaker, from the speaker. I usually have the sound on my computer turned down (or off); it needs to be on
Plug things in:
- Line from circuit goes to external stereo sound card, Line IN
- Plug 9V battery in to circuit
- Speaker (line from end of PVC pipe) goes to headphone plug on computer
Check settings on Control Panel:
- Audio: Sound playback: for my computer it needed to be set to Realtek HD Audio output
  - Volume: Mic volume should be on (default for mine is muted, which then doesn't work)
- Audio: Sound recording: USB Audio
  - Volume: select Line, make volume nonzero
Run the Scope program
- Trigger: normal
- Use cursors to determine the time difference between the two signals.

I measured a distance of 137cm between microphones, with a time difference of 3.951ms between signals. That corresponds to a speed of sound of 346m/s, close to the expected 345m/s at $22^\circ$ C. Not bad! Certainly better than 10% error!

Wiring diagram for speed of sound circuit	Measuring distance between microphones	Reaming out holes for the microphones
Microphone and wires from end speaker		Line for testing the speed of sound equipment	My circuit board
Time from 1 microphone to the other: 3.951ms

Pendulum with red bob and black pick-up

Graph to determine g

When we finished playing with the speed of sound, we used the same Scope program to measure the acceleration due to gravity from the period of a pendulum. We taped a small neodymium magnet to the bottom of a heavy pendulum suspended over (close to but not touching) a telephone pick-up coil. The pick-up coil sent a sharp signal that we caught on the same Scope program, from which we measured the time for 1 period of the pendulum. We graphed the period squared $\left( T^2 \right)$ vs. length of the pendulum. The relationship between period (T) and length (l) is $T = 2 \pi \sqrt {l \over g }$ , so the slope of $T 2$ vs. l represents ${(2 \pi )^2 } \over g$ . From my graph, and I wasn't particularly careful with the length measurements, I determined g = 9.801m/s^2, not far from the accepted value of 9.81m/s^2; in fact I think someone said around here the value should be 9.803 m/s^2. This is waaaay more accurate than the standard hand-timed method to determine g. Nice!

[edit]

Day 3: Wednesday July 16, 2008

Tom's LEDs: "Raising the Bar", or "Can you hear me now?"

Today we soldered and tested a circuit to measure Planck's constant. The circuit was a 9V battery connected via a $2 k Ω$ resistor to any one of 6 LEDs. A switch determines which LED, and the LEDs varied in wavelength from IR to UV as shown in the table below.

**LEDs for Planck's Constant**
LED color	Wavelength (nm)
IR	945
red	630
yellow	592
green	525
blue	470
UV	400

Planck's Constant determined using LEDs

We switched on one LED and then varied the potentiometer until we could just start to see the light, and recorded the potential (voltage) across the LED. IR was visible using our cell phones. Helio had also put together an IR camera from an inexpensive video camera. He opened it up and replaced the IR filter with a piece of exposed 35mm color film (to filter visible light). I should try that; it was fun to test what was visible in the IR range.

There was some discussion about the most appropriate way to determine the voltage. We agreed that turning the potentiometer until we could see some light was somewhat arbitrary. More objective alternatives would be to 1) measure the potential (voltage) at maximum current, or 2) take several readings of both current and potential for each LED and graph them to determine the onset voltage for each LED. One of us should try the three different methods to decide which gives the best measurement of Planck's constant h. Using the onset of light method I calculated $7.07 \times 10^{-34} J \cdot s$ , a 7% difference from the expected value of $6.63 \times 10^{-34} J \cdot s$ . Again, not bad!

Sometime during the day we also talked about semiconductors and band gaps. Helio did a demo where he dunked a red LED into liquid nitrogen, with two effects: it got colder and the light turned orange -> yellow. He also dunked a yellow LED in LN2, and it turned green. At least those are the color changes I think I remember. I really liked seeing how a change in temperature could change the band gap energy, directly changing the emitted color. Very cool.

Sean McC and early computing

In the afternoon Sean McCorkle of the BNL Biology Department (but originally an astrophysicist) talked about the origins of computers. He gave a really nice talk. He took us all the way from slide rules and number systems to the design of modern computers, CPUs, and even FORTRAN, all in 1 hour.

Light oil

After the talk Helio did a demo where he shone a green laser through different bottles of olive oil. The different grades of olive oil fluoresce differently, apparently because of the different amounts of chlorophyll in them. It was fun to look at, but I didn't spend enough time with it to get a sense of what the color changes were (sometimes appeared yellow, other times green) or which color was "good". At a guess I would say that high quality olive oil probably has low chlorophyll content, so fluoresces less than low quality olive oil (assuming it's the chlorophyll that fluoresces). Fluorescing would occur at a lower frequency than the incident green laser, so I would expect high quality olive oil to show a green light (the original beam) through the oil, while low quality olive oil would show a yellow fluorescent beam through the oil. I should get a green laser pointer and play with this.

[edit]

Day 4: Thursday July 17, 2008

Buzzword bingo

In honor of today's theme of the day: Particle Physics, I decided to play buzzword bingo. I started first thing in the morning, but it took me until late afternoon to get bingo. I even asked pointed questions fishing for specific buzzwords; that got me strange looks but no buzzwords.

Our first demo was of Brownian Motion (thanks Helio!) (good for Brownian motion, particle, Avogadro, milk fat). Helio suspended some 1-micron latex spheres in water, put a drop on a slide, and projected it via a microscope onto the screen. A small amount of milk (source of milk fat) with blue food coloring for contrast, suspended in water also works. We could see the latex particles bouncing around in their random walk.

Tom with Cloud Chamber base

The second demo was a Cloud Chamber (thanks to Tom, with assistance from Rich L.). When it got going we could see the trails in the cloud left by particles whizzing by. Since heavier particles decay into lighter particles, by sea level (where we are) any pions have decayed into muons and then possibly into electrons. The straightish lines are mostly from muons, while the lines with several changes in direction are most likely the lighter electrons. Charged particles have curved paths because of the strong magnet at the base of the Cloud Chamber, and occasionally we see electron-positron pairs: starting from the same point, 1 trail curves to the right while the other curves to the left. Some people really enjoy the Cloud Chamber; I could watch it for hours.

Ice cream for dessert!

We had the best Picnic Lunch, organized by Rich G., complete with wiki sign-up sheet. Now this is what I call a good use of the wiki! It was so good, we had ice cream for dessert and left-overs on Friday.

In the afternoon we had presentations by Tom Hemmick and Helio. Tom's talk was The Perfect Fluid. He talked about particles, plasmas, the evolution of the universe, and how to define a prefect fluid. If viscosity/entropy density defines how perfect a fluid is, then Tom says that the most perfect fluid is plasma at 10^12K. Helio's talk was about the ATLAS experiment and the search for the Higgs. So we had lots of particles to think about.

And, as I mentioned above, I got buzzword bingo towards the end of the afternoon.

[edit]