Clone this wiki locally
Circuit: a 0.1 microfarad capacitor was connected to a 1Mohm resistor in series. connect to a function generator. edit: changed to a 10 kohm resistor a 10 microfarad capacitor.
I wired the circuit given on a prototype board. hooked an oscilloscope across the resistor. Playing with the frequency it seems to attenuate more sharply around ~1.5Hz.
Paul pointed out to Dr. Koch that the input impedance was 1Mohm on the scope so it was changing the effective resistance value. We changed our resistor to a 10 kohm resistor and our capacitor to a 10 microfarad one.
The new capacitor seems to have a polarity, which is usually a bad idea with AC, but we're going to go ahead and just try it. A piece of paper in the drawer I got it from said "Tantalum caps". Looking up "Tantalum Capacitors" on wikipedia describes the switched polarity problem as breaking down an oxide dielectric causing the capacitor to fail.
The new circuit seems to be working, and the sharp change (by sight) is approximately at the same frequency. To account for the polarity i am DC offsetting my function generator. (in the correct direction!!) This did not start until after putting low voltage across the cap the wrong direction. Played with triangle and square waves. Triangle waves are changed less in amplitude but gain a curved edge. Square waves have exponential decay top edges instead of flat lines (since this is the V across the resistor)
Modified my code from last week (follow the example given) to add a second graph that is the previous data, and an auto save button that takes the data and saves it (and a picture) to a folder in the same place as the VI.
Returning to lab today to finish the requirements. I edited my Labview code to read the frequency and amplitude of my signal and graph it cumulatively. Also a button to save that data as a 2D array in a spreadsheet. At Dr. Koch's prompting I've changed my graph to be power (in some unknown units since I don't calculate voltage) vs. frequency.
I expect my cutoff frequency to be ~1.59 (from the equation given and nominal component values) and by sight, half maximum power appears to be at 1.6Hz. I inspected the graph by changing my frequency scaling to zoom in near where my rough estimate was zoomed out. I did this on the data I took keeping between 5 and 10 periods in my sample window so it would average the amplitude better.
[COMMENT FROM ZEKE: Nice, your calculated cutoff was close to expected. Also, is it annoying to put comments in here like this?]