Physics 128A: Senior lab, Fall 2012

MW 1:00-5:50 or TR 1:00-5:50 in Broida 3223 http://hep.ucsb.edu/phys128/

Instructor: Ben Monreal, bmonreal@physics.ucsb.edu, Broida 5123.
Teaching assistants:
Brooks Campbell
Scott Haselschwardt
Alexander Walter

Lab managers: Bob Pizzi (rpizzi@physics.ucsb.edu) and Dan Bridges (bridges@physics.ucsb.edu)

Course structure

In Physics 128L, you will be working through a series of modern physics experiments---you'll have to understand the equipment you're given, figure out what procedures to follow and what data to collect, actually collect data, analyze the data, and present it in a scientific manner. The focus of the class is NOT pencil-and-paper physics; you will not be asked to prove or derive anything as an intellectual exercise. Rather, the focus is:

General lab conduct and skills; how do you walk up to an unknown piece of equipment and make it do something useful? How do you get a table of numbers out of your notebook and into a publishable graph or figure?
Understanding real-world data; what's the relationship between a "physics number" (an index of refraction, or the speed of light, or the mass of an electron), our error bar on such a number, and the procedure you're using to measure it? This is sometimes called "statistics", but actual statistics is only part of the story.
General science-world conduct. You have to work with a partner, follow lab safety practices, keep an honest lab notebook, write concise and sensible reports, and give short slide presentations.

The actual collection of experiments on offer can be found here. You will do 4 of these.

Course Materials

A lab notebook. This must be a bound (not looseleaf, not perforated) notebook, containing quad-ruled paper (aka graph paper). You will need TWO of it, since the TAs will be grading one lab while you need to be writing up the next one. If you have a carbon-paper notebook, that's probably acceptable since the TAs can grade the carbons.
Access to the lab writeups. The bookstore sells a bound copy, but the same information is available here.
Access to Wolfram's Mathematica software. It is installed on all lab computers, many UCSB computer-cluster computers, and in the PSR. If you want a copy on a personal machine, you can buy a ($45) one-semester or ($74) one-year temporary license, or a ($140) regular copy, via http://store.wolfram.com/catalog/. There is no physics department or UCSB license with additional discounts.
A place to keep your computing work-in-progress. I recommend a USB thumb drive, but some students may prefer uploading data to Dropbox, email, or some such. (Data will *usually* survive week-to-week on the computers in lab, but they're multi-user computers so do not trust this.)
A sourcebook for statistics and data analysis. I recommend An Introduction to Error Analysis: The Study of Uncertainties in Physical Measurements, by John Taylor. Last year some people got along with Internet resources.

Course schedule

You are in one of two sections, meeting either MW or TR. We divide the quarter into five two-week blocks; you spend two weeks on each lab. The first block, rather than doing a modern physics lab, you'll design your own simple classical measurement.

First meeting: September 27, 2012 (Thursday, short meeting) or October 1 (Monday)
"Pre-lab lecture week": Oct 1-4. We will meet in Broida 3223.
"Design your Own Experiment" week: Oct. 8-11. We will meet in Broida 3223.
Experiment Block #1. Oct 15-25.
Experiment Block #2. Oct 29-Nov 8.
Experiment Block #3. Nov 12-22. (One Monday off for Veterans Day, one Thurs off for Thanksgiving.)
Experiment Block #4. Nov 26-Dec 6.
Final exam week; no activity.

Each "block" will have the same format.

On Day 1, you and your partners will show up (having read the lab and met with your lab partners) prepared to give a short, 2 or 3 slide talk introducing it. Every group must prepare such a talk; a random subset will be picked to actually deliver it. These talks are worth 15% of your grade.
On days 1-4, you will work independently on the labs, keeping all of your data/procedures/observations/comments in a professional-style lab notebook. The TAs, staff, and professor will circulate to help you. The notebook is worth 60% of your grade. Attendance is required, firstly because your labmates expect you to work as a team, and secondly because we're paying attention.
On "day 5" (i.e. the first day of the next block) you must hand in the previous block's notebook, plus a 1-2 page synopsis or mini-writeup. The writeup is worth 25% of your grade.

Course Goals

Learn how to perform careful, organized, documented and systematic measurements.
Learn proper methods of data analysis.
Learn how to draw specific and meaningful conclusions.
Develop a tough-minded and skeptical scrutiny of results.
Understand issues of precision and uncertainty.
Become familiar with operation of standard equipment.
Appreciate the limitations of apparatus.
Learn to clearly and concisely present your methods and results, both verbally and in writing.

You will also learn some modern physics, by doing it.

Some guidance

Learn Mathematica (STRONGLY PREFERRED) or another reasonably-professional data analysis system. (If the Mathematica learning curve is too steep, the "next best thing" is Igor Pro, a spreadsheet-like program also installed on the lab computers. If you need to retreat temporarily to a spreadsheet like Microsoft Excel or Google Spreadsheets, it's better than nothing, but a spreadsheet-dependent semester will score no higher than B.)
Five useful Mathematica notebooks:
- physics128_histogrammer.nb
- physics128_plotter.nb
- physics128_fitter.nb
- physics128_nonlinear_fit.nb
- candy_example.nb, a pseudoexperiment that walks you through the difference between "standard error on the mean" and "standard deviation".
Have you given a public talk before? Most people haven't at your level, but in a few years you will. Consider these talks as practice for your first APS meeting. Some examples of slides from my own undergrad days: the Frank Hertz experiment and Positron Annihilation. For reference, these were probably 10-minute talks.
The 1-2 page paper is a new feature of the course; it replaces a long "final paper" we used to have. It is a short (1-2 page) synopsis of the experiment with an abstract, an introduction, experimental method, discussion of results with ~ 2 relevant figures, and conclusions. The idea here is that you learn how to selectively and concisely present relevant data and present a coherent, conclusive result. A sample is shown here: example lab pdf (sources: latex, figure, mathematica notebook.)

Safety

There are important safety issues in any lab work, of which you must be aware. Some examples of safety hazards are intense light sources (lasers and gas discharge tubes), electrical hazards (high voltage or current), radiation sources (radioactive substances or X-ray machines), extreme temperatures. I require that you start each experiment by doing an assessment of the safety issues. You will need to take steps to carry out your experiments safely; this is part of acting like a professional experimentalist. For example:

Look for safety-related information in the lab manual.
Look for safety-related information in the equipment manuals.
Ask an instructor about anything that concerns you.
Include safety in the planning discussions with your lab partner.
Communicate clearly and unambiguously with your lab partner while working.

Important safety rules that everyone must follow are:

You must never work alone. There must always be at least two people in the room.
Never leave activated equipment unattended without approval from an instructor.
If any accident occurs, you must immediately report it to an instructor.

How to do well

This course is quite different from most of your past classes. Skills that you have honed in past coursework may not apply, and you will need a completely different skill set. For some of you, these may come as naturally as algebra. For others, they might take a while to develop. Learning those skills is a primary goal of this class. Here is my advice on how to approach things.

Accept that you are confused. Identify what you don't understand.
This part never changes. If you know what you are doing, it is just an exercise, not research.
Formulate a set of specific questions.
Pick some questions for which you think you know the answer, and confirm that. Pick some questions for which you can find an answer.
Plan a method for answering at least one of the questions.
Carry out your plan.
Go back to 1.

This approach works to attack problems of any scale. You can use it to understand a completely new physical phenomenon or to debug why your car won't start. Note that this is just a rephrasing of the scientific method. It works for things that are not always recognized as science, such as figuring out the behavior of financial markets or people--including yourself.

The methods you use to answer questions will not just be experimental. You will need to do substantial reading on your own to understand the theoretical issues underlying your experiments and how your equipment works. We have equipment manuals and a small library of books that you may check out; they must be returned by your next lab period. You should do such reading before you begin your labs. From it, you should develop a plan of action before you begin. Careful preparation is important because you won't have much time to work with the equipment. (This is true in most experimental work, e.g., when using shared equipment like a telescope.)

An important part of being prepared is understanding how your equipment works. Before using it, you should plan to fiddle around to learn about it. I encourage fiddling, but be careful not to break things. The best way to avoid that is to read the manuals beforehand and plan your fiddling. If the equipment is not working, or if you break it, notify an instructor promptly. You will not get in trouble for breaking something, but you will be in trouble if you don't tell us about it.

Since time with the equipment is limited, you should carefully document everything that you do. You might not expect something to depend on the temperature or the time of day, but you should still record both. It could be helpful after you have collected all your data. For example, you might find that one particular set of data is completely inconsistent with the rest. By knowing when that data was recorded you can correlate it with other effects; perhaps someone powered on some other equipment that caused interference.

It is also useful to analyse the data as you take it. This can help you quickly identify problems. For example, if you are measuring voltage vs current, you should make a table in your logbook where each row is one measurement and has columns for: the time of the measurement, the voltage, the current, and any relevant comments. Then, as you are filling in the table you should also plot the data, e.g., on the opposite page of your logbook. Your plot might reveal a trend that can inform your subsequent work, such as helping you decide to use larger or smaller voltage steps to go faster or more carefully.

It will help your experiments succeed if you prepare well, clearly write down your plan, and fully document your work as you go. This is also what we will look for while grading your logbooks.