XV. Evaluation of the instrument
There are two ways to write this section of the exercise. I could list the parameters I want you to have measured. If I do that, I am sure you will measure all those properties, but you won't ask yourself "what characteristics matter?" One goal of having you build this system is to work out for yourself what parameters are important, to figure out how to measure them, and then to decide how to improve them. So here is what I will do: when the class comes up with at least 6 parameters to evaluate, I will be willing to suggest an additional parameter or suggest what you may wish to consider in figuring out an additional parameter. I'll even start you with a hint: did everyone use the same camera? If not, did all the numbers for intensity at a particular wavelength come out to be the same? If not, why not?
XVI. Evaluation of the experiment
There are many ways to evaluate what you have done. Here are some of them.
- Did you succeed in building a working spectrophotometer? If so, how well did it work? If not, what did you see about components, instructions, software, or other circumstances that prevented it from working?
- Did you learn anything about designing and building instruments from working on this device? If so, what did you learn? If not, can you identify why you learned little or nothing? Would better instructions have helped? Better software? If you can see ways to improve the software, describe how it should function.
- There are only 30 contact hours for American faculty when teaching at HUS. This spectrometer consumed at least 2 hours of that time. Was this a good use of that time? If not, what would you suggest would be a better use of that time?
XVII. If you want to do more
There are many questions we haven't answered. Here are some of them, in no particular order.
- What is the theoretical resolution of this instrument, and what sets that resolution?
- What is the ideal distance from the LED to the grating? How did you decide this number?
- What limits photometric precision – shot noise, digitization noise, stray light, detector saturation, too much adaptability in the camera (making photometry impossible because of ever-adjusting exposure times), or something else?
- Is there some other component arrangement that would be just as inexpensive but work better?
- Could these components be used to develop a fluorimeter?
- Given how much of a problem stray light and digitization noise are, is this spectrometer worth using for anything but teaching purposes?
In addition, it is obvious that having the data processed the way it is now is quite awkward. Future plans for the software include:
- Recast the current software so it runs in a browser. That way, the processing can occur on any computer, not just a Windows machine.
- Rewrite it a second time so that it runs natively in a cell phone. That way, the whole data processing cycle can be done in the palm of the hand of anyone with a camera phone.