Franklin W. Olin College of Engineering Introduction to Modeling and Control: Engineering of Compartment Systems Lab 10: Building a Hot Wire Anemometer Part 2 of 2 In this lab you will construct, test, and model a working hot wire anemometer. This lab will be the final lab in thermal systems and will require you to use your new knowledge of heat transfer and circuits. The final product of this lab will be a working scientific instrument which can measure wind speed with very good temporal resolution. Create the circuit shown below. NOTE THAT PIN 4 OF THE OP-AMP GOES TO -12V. U1 1 6 OUT OPA551 OS1 OS2 5 +12 7 R1 Lamp 4 -12 R3 V- V+ 2 5K 3 + - R2 50 We will discuss how this circuit works in class, but by now you should be able to get some idea by thinking about how this circuit works at steady state. HINT: the voltage dividers on the left and right of the op-amp are fed identical voltages at their tops from the op-amp output. If their mid-point taps are at the same voltage, then their resistance ratios must also be equal. From this description, and imagining what happens if the lamp resistance goes up or down, you should be able to deduce what this circuit is trying to do. The circuit has stable equilibria at two output voltages, one positive – the other negative. It has an unstable equilibrium at zero volts. Leave the glass light bulb intact. Check that the circuit is working by turning the potentiometer. You should see the light bulb brighten and darken as you turn the potentiometer up and down. Finally, turn the potentiometer all the way down so that the light turns off. Take the light bulb out of the circuit. Wrap the bulb in a sheet of paper. Slowly crush the bulb in the vice that is in the lab. Carefully discard the broken glass and carefully remove the bulb from the paper. Carefully place the bulb back into the circuit (the glass is sharp and the filament can get hot). Place the oscilloscope probe on the op-amp output. Add a measurement of the signal’s mean to the display. Slowly turn up the potentiometer until the mean is about positive or negative 5 or 6 volts. It will be very sensitive and hard to get an exact voltage, but somewhere around positive or negative 5 or 6 volts will work just fine. You simply need the filament to get hot but not saturate the op-amp. Flutter your hand across the light bulb, causing air to move around the filament. If the circuit is working you should see the op-amp output also flutter. Set the scope such that the vertical position is zero and adjust the vertical offset so that the signal is near the middle of the screen. Set the vertical scale to 2 volts and the horizontal scale to 400 milliseconds. Take a few screen shots as you rapidly flutter a piece of paper above the filament. Build a Simulink model of the entire system. Start with the op-amp model from previous labs. You should create an op-amp subsystem that takes two voltages as the + and – input and provides the op-amp output. Consult previous labs if you forgot the op-amp model. Create a separate subsystem for the thermal behavior of the wire. This subsystem should take the voltage on both sides of the light bulb as the two inputs and provides the light bulb resistance as the single output. To construct the wire subsystem you can start with the basic thermal model we discussed in class which takes power as the input and provides temperature as the output. You will need to adjust this model such that you are now converting the voltage inputs to power. An additional complication is that the resistance changes with temperature and the power supplied also depends upon the resistance. Use the approximate linear relationship you used in the previous lab. Once the two subsystems are complete, all you need is ohms law applied to the bridge circuit to connect your simulated op-amp and wire subsystem. You should use the gain and slew rate from your old work to use in your op-amp model. For the wire model, take hA = 0.002 W/K and mC = 6.6e-4 J/K. These values were obtained from a set of experiments similar to what you did last week, but with the filament in open air. Your model may show some instability in the system when you first turn the simulation “on”, but it should stabilize after an initial transient. Also, you will need to provide the op-amp a “kick” to get it started out of the unstable equilibrium at zero volts. Try setting the initial state of the op-amp’s integrator to be 0.1 V and that should be sufficient to get the system going. Compare the basic results you obtain from your model and your experiment. Of course you cannot test them exactly as you cannot reproduce the wind signal. But you can test that the simulated response is behaving qualitatively as you observe, and get some quantitative validation. Setting the convection to be sinusoidal in your simulation, what is the highest frequency fluctuation in the wind that your model predicts your instrument could measure? Does this correspond to what you observed on the actual hardware? Finally, what kind of controller (P, I, D, etc …) is the op-amp acting as? Explain.