Computer Control Laboratory
Multi-Filter Photometer Experiment
This experiment is an example of the computer-controlled physics experiments in PY455/456 (Electronics & Interfacing). There are many laboratory experiments that are best conducted with the presence of a person, but there are other physics experiments that can be handled better by computer. These two senior physics courses teach cadets the electronics they need to set up computer-controlled experiments and the details of interfacing these experiments to a microcomputer.
Summary of this experiment: This experiment measures the output of a tungsten lamp over a wide range of wavelengths in the visible part of the spectrum, using a photodiode as a light detector, with LabVIEW as the control software. The filter wheel is moved from one filter to another by a stepper motor driven by a microstepper controller under program control.
The raw light intensity readings are corrected for the photodiode responsivity and graphed vs. wavelength, then fitted to a theoretical Planck curve; this yields an approximate temperature of the source and allows one to judge the relative precision of the measurements.
Experiment Layout: The 10 wavelength bands are defined by interference filters with a bandpass of 100 Angstroms (10 nm). The filters are mounted in a filter wheel, which is turned by a stepper motor from one filter position to another. A computer interface card is used to control the filter wheel and record the measurements of brightness, writing the results to a disk file. This experiment involves motion (a moving filter wheel controlled by a stepper motor), the measurement of light intensity, a digital input sensor, two digital outputs, and an analog input.
The LabVIEW program: One page of the program is shown at left: LabVIEW is a graphical programming language; the diagram shown is part of the actual program.
The program first searches for an index tab on the filter wheel, designating the home position of the filter wheel, using a While...Loop. When the tab is detected by the IR LED/photodiode, a TTL signal is read by a digital input on the interface card. The program then enters a Do Loop of 11 loops, one for each filter. Inside the loop is another Do Loop to step the motor to the next filter.
In our case, with the microstepper controller set to 1600 steps/revolution and with filters spaced 30 degrees apart, we needed to step 133 steps, requiring a loop of 266, where a digital output was toggled high and then low in the next loop using a shift register. Once a given filter was reached, we took a reading of the light intensity, moved the motor one step, and repeated this cycle 10 times to look for nonuniformities as a function of position across the filter.
This cycle was repeated over the 11 filters. Finally, a disk data file was written with the resulting 110 readings.
Results: The data for each filter were averaged; the interference filters had the following wavelengths: 365, 405, 436, 492, 564, 580, 632, 766, 852, 905, 1064 nm. Two corrections are necessary. First, the sensitivity of the photodiode for various colors is known and a correction is then applied to the data. Second, the output reading of the op-amp with no incident light must be subtracted from the readings (a combination of offset voltage and results from the input bias current).
The most subtle correction resulted from the actual amount of light passed by the individual filters. After the two corrections above were made, there were still two wavelengths that showed anomalous data, for the 852 nm and 1064 nm filters. It was suspected that these two filters might have anomalous peak transmission, and they were scanned with a spectrophotometer, along with two other filters as references. Indeed, the peak transmission on the 852 nm filter transmitted 35% more light than the two reference filters, and the 1064 nm filter transmitted 86% of the two reference filters (based on the area of the scanned transmissions). Hence, corrections were made to the readings for these two filters, and these points now fall very close to the theoretical Planck curve.
Theory vs. Experimental Data: The graph shows the brightness measurements (arbitrary units) vs. wavelength (in nm), with 11 interference filters and three lamp brightnesses, with a theoretical Planck curve fitted to the points. The agreement between theory and data is best in the near infrared (right hand side of the graph), since the photodiode has a very low sensitivity in the blue part of the spectrum. The temperature for the curve shown is 3000K.