top of page

Design of Experiment

Our experiment did not go as we anticipated. Our initial plan was to use a simple half-wave rectifier to determine the time constant of the circuit and use that to calculate the capacitance. Figure 1 represents the circuit we used to test the value of the capacitors. We believed this to be a simple easy way for us to determine the capacitance because it did not require many calculations on our part which reduced our potential locations for error. We quickly realized that this was not the case as we began to test our hypothesis on known capacitors.

DOE Rectifier.JPG

Figure 1: Halfwave rectifier circuit we used to measure the time constant

Due to the simplicity of our testing circuit we were able to try multiple resistor combinations for each capacitor. We wanted to test each capacitor with multiple resistors to improve the precision of the measurements. We tested with up to five different resistor values: 100Ω, 1kΩ, 10kΩ, 100kΩ, and 1MΩ. Testing over this range provided us with multiple measurements of our time constant with the same capacitance. With each resistor, we measured either the time it took to drop from the peak to zero volts or the time between when it stopped and started charging again. While taking these measurements, we noticed that for the larger capacitors we had no need for the larger resistors because the time constant was so large the voltage drop while discharging was near zero volts and was difficult for us to measure. The difficulty measuring the voltage drop forced us to have a limited number of trials for that capacitor and therefore reduce the precision of our calculations. Below is a table of the measurements we found when using a 1 µF capacitor.

DOE table 1.JPG

Table 1: Measurements of half-wave circuit using a 1 µF capacitor

As we moved through our tests, we took many measurements of the rectifier circuit. Our outputs all appeared as we expected them, we could clearly see the discharging capacitor and measure the time it took it to discharge and the time it took to do so.

tek00029.png

Figure 2: Discharging capacitor of the rectifier circuit with 1 µF capacitor and 100kΩ resistor

The figure above shows an example of the clear output we were able to achieve which allowed us to make easy measurements and determine the voltage drop and time difference. However, we often had problems with outputs like figure 3 that had so little voltage drop they were hard to measure to help determine the time constant.

tek00030.png

Figure 3: Discharging capacitor of the rectifier circuit with 1 µF capacitor and 1MΩ resistor

Since we did multiple trials of each capacitor it kept trials, like the one pictured in figure 3, from derailing our measurements and allowed us to average and find a more precise time constant. We were also plagued with a similarly implausible measurement when the time constant was too small. When this occurred, our measurements struggled to tell when the capacitor had finished discharging as it was very similar to the normal curve of the input signal.

tek00027.png

Figure 4: Discharging capacitor of the rectifier circuit with 1 µF capacitor and 100Ω resistor

We intentionally tested multiple capacitors along with the 1 µF because we wanted to make sure that we were able to correctly determine the capacitance for any capacitor. We tested two other capacitors: 10 µF and 47 µF which resulted in the following measurements in table 2 and 3 respectively.

DOE table 2.JPG

Table 2: Measurements of half-wave circuit using a 10 µF capacitor

DOE table 3.JPG

Table 3: Measurements of half-wave circuit using a 47 µF capacitor

After collecting this data, we ran into our biggest problem: we couldn’t determine the time constant. We had all the data we thought necessary and knew that once we determined the time constant, we could easily determine the capacitance. However, we could not determine the relationship between voltage drop and time that gave us the time constant. As our testing time closed and we were required to determine some mystery capacitor we were at a loss. We tried to find the relationship but, in the end, we were only able to determine that the capacitor was very small. We determined that the capacitor had to be significantly smaller than anything we had previously measured so we were forced to guess. We used the collection of all the data we had and the trends of change in discharge time related to change in capacitance and guessed .5 µF. Our estimate was very far off from the capacitor that measured in the pF range.

Upon completion of our experiment we concluded that most of our procedure was tactful and lead us in the proper direction but, our inability to relate the discharge time to the time constant caused us to be unable to determine the capacitor value within our expected 5%. Afterwards, we decided that it would have been easier for us to determine the time constant if we had fed a square-wave into our rectifier because it would have been easier for us to measure the discharge time and determine the time constant. The time constant is how quickly the capacitor can charge and discharge, and if the input was a square wave then we would have been able to see the crucial milestones in the charging and discharging periods. Using length of time, it took to discharge half of the voltage we could have determined the time constant. However, using a sinusoidal input we had a difficult time recognizing when the output reached its half charger state and therefore struggled to measure the time constant. We also determined that if we removed the diode and made our circuit a simple RC circuit, our results may have also been clearer. The rectifier would have been useless once we switch to the square input because the diode would just be absorbing voltage for no reason. After completing our DoE lab, we decided that our initial approach could have been successful, but we did not choose the best approach to accurately determine the capacitance.

bottom of page