Physics307L F09:People/Callow/EdiffractionSum

Electron Diffraction Summary

SJK 22:08, 14 November 2009 (EST)
22:08, 14 November 2009 (EST)
Very nice job on this lab. I know you're doing the Millikan for your formal (I'm stoked to read what you do with that), but it would have been fun to see if you got better answers with the new bulb for e-diffraction that we now have.

Link to lab manual used

Link to lab notebook

Purpose of the lab

The purpose of this lab is to verify the de Broglie hypothesis $\displaystyle {\lambda}=\frac{h}{p}$ by diffracting electrons through a thin sheet of graphite and comparing a calculated value of the slits created by the graphite and the accepted value of these.

Summary of data

From the calculations in my lab notebook I found the size of the slits created by graphite to be .268(5)nm and .148(1)nm. The accepted values are .213nm and .123nm. So based on my sdem for each value, my experiment doesn't agree with the accepted values (the accepted values are much further than 3 sdems away). In the lab notebook near the bottom there is a list of possible reasons for my error.

SJK 22:05, 14 November 2009 (EST)
22:05, 14 November 2009 (EST)
Good comparison to the accepted value, using statistical reasoning with your SEM. I agree, huge systematic error is evident, and I commented more on your primary notebook page.

Other than what is listed I couldn't think of any reason other than an error in my calculations somewhere, but nothing from my calculations looks like it would change the answer by only 20% if I did the units incorrectly. The sdem for each individual measurement tended to be huge though, so my data just might not be very good.

What I learned

Find formulas first instead of trusting an approximation in the manual will actually make sense with my problem. Even though not making the assumption the manual does would mean my best estimates are even worse as mentioned in my lab notebook on finding the lattice spacing, because this assumption was not true I should have used the exact values.

This is the second experiment I have done involving an electron gun and an accelerating voltage in my formulas (the other being e/m). Both had large error. In the future if dealing with similar equipment it would be nice to find a better approximate of the accelerating voltage. Because of the much higher voltage in this experiment and that my data just looks bad, I don't feel nearly as comfortable naming this the cause of all my troubles though.

There was a lot of opportunity in this lab to practice error propagation and found everything involving went much quicker than previous experiments. Hopefully that means I'm getting better at it. SJK 22:07, 14 November 2009 (EST)
22:07, 14 November 2009 (EST)
That's great! Hard work is paying off!
One final thing I'm hoping I learned from this lab is that if I have multiple tools, and one isn't working how I want it to, just switch. I would have hoped I learned this from the maple trouble in previous labs but nope. Googledocs gave me a lot of trouble when trying to plot my data. It took about an two hours before I finally just put it in matlab after finding googledocs is just not very good for plotting data. (or well, compared to matlab) SJK 22:07, 14 November 2009 (EST)
22:07, 14 November 2009 (EST)
Sorry you wasted time, but a good lesson to learn.