ELEC 243 Lab

Experiment 4.2

Improved Resistance Measurement

Let's apply our new Labview skills to the task of improving the resistance/temperature measurement system we built last week.

Part 1: Measure Vs

One shortcoming of last week's system is that we have to be careful to set the power supply voltage to the correct value. Since $V_S$ appears in the formula for temperature, if this voltage is incorrect, the temperature reading will be incorrect. A much better design would be to have Labview measure the actual value of $V_S$ and use that value in the formula.


Step 1:

 Restore the connections you had in Experiment 3.2 last week with the 0-6 V output of the power supply (set to 5.00 V) connected to J1-3. Check that the circuit on your breadboard is still wired correctly. Load and run the VI you saved last week and verify that it still produces the correct reading for temperature. Vary the power supply voltage and observe that the temperature reading changes in response.

Step 2:

 Stop the VI. On your breadboard, add a wire to connect $V_S$ to ach4, so that your circuit looks like this:
\includegraphics[scale=0.500000]{ckt3.3.5.ps}


Step 3:

 We are now asking the A/D converter to read two different voltages simultaneously. This is not a problem for the A/D converter, but it does require us to figure out how Labview represents multiple simultaneous samples.

The obvious way to handle this would be to simply add another A/D block and set it to A/D Channel 4. If you try this, you will get a cryptic error message to the effect that the device is already in use. This is because all of the channels are on the same card, and the first channel which is opened has exclusive use of the device. Labview handles this by combing multiple samples into a vector, with one entry for each channel.

Double click the A/D block or select Properties from the menu. This will bring up the DAQ Assistant dialog. Click on the Show Details button near the top of the panel



Step 4:

 Click on the Add Channels button.
Select Voltage from the popup menu, then select ai4 from the list of Supported Physical Channels. A new entry for ai4 should appear in the channels list. Note the Scan Order column. This tells us that channel 5 will be the first element of the vector of samples and channel 4 will be the second. Although this seems untidy, it won't cause problems as long as we remember the order. Click OK and wait until everything has settled down.

Step 5:

 The block diagram looks the same as before, but now the wire is carrying two separate values. Perhaps surprisingly, this fact does not change the original operation of the VI. Subsequent blocks, expecting a single value, simply take the first element of the vector. Since we left the original signal ($V_T$ ) as the first element, everything to the right of the A/D converter block works as it did before. Start the VI and verify that this is the case.

Step 6:

 If we know how to look, we can see that the wire coming out of the A/D converter block is in fact carying two values. With the VI still runing, bring the block diagram window to the top. Place the cursor over the wire between the DAQ Assistant block and the numeric indicator labeled VT. The wire should start flashing and the cursor will turn into a circle with the letter "P" in it. This is the symbol for a probe. Left click to place the probe.

A small rectangle labeled 1 will appear at the output of the A/D converter block

and a window labeled [1] data will pop up.
The probe display window shows the value of the data on the probed wire. The numeric control in the upper left corner (currently zero) determines which of the wire's signals is displayed. Type "1" into this field or click on the upper arrow to the left of the box. The value displayed for Y should change to a value close to 5.

Step 7:

 Stop the VI. Widen the block diagram window, or scroll it to create some empty space to the left of the while loop. Place the cursor on the border of the while loop. Resize handles (small black squares) should appear. Place the cursor over the one in the center of the left edge and widen the while loop by about an inch. move the DAQ Assistant near the left edge of the loop.

Step 8:

 Click on the wire segment coming out of the data terminal of the DAQ Assistant block and press Delete to remove it. The remainder of the wire will become broken, since there is now no input. From the Functions palette, select Sig Manip, then Split Signals. Place the resulting icon between the A/D converter output and the leftmost portion of the broken wire, and click to release.

The split signals block is expandable to accomodate the number of signals in a bundle. Place the cursor over the middle of the bottom edge and move it around until it turns into a resize arrow. Drag the bottom edge down by one increment. The resulting icon will look like a wishbone in a box.



Step 9:

 Wire the left side of the split signals block to the A/D output. The upper output is $V_T$ (ach5). Connect it to the broken $V_T$ wire.

Step 10:

 The lower output is $V_S$ , which will replace the constant value of 5 currently connected to the upper input of the subtract block. Delete this numeric constant block and wire the lower output of the split signal block to its previous destination.

Step 11:

 Start the VI and verify that the displayed temperature value is still correct. Vary the value of $V_S$ and verify that it remains correct.

Part 2: Generate Vs in Labview

Our temperature measurement system is now more accurate, but it's still very bulky since it requires the power supply in order to function. In the previous Experiment we were able to use Labview to generate constant voltages. If we were to generate such a voltage and connect it in place of the power supply, we would have a self-contained measurement system.


Step 1:

 Stop the VI. On your breadboard, remove the wire connecting J1-3 to $V_S$ and replace it with a wire connecting dac0 to $V_S$ .
\includegraphics[scale=0.500000]{ckt4.2.2.ps}


Step 2:

 On the block diagram, move the bottom edge of the while loop down to create room for additional components. In the resulting space add a DAQ Assist output block, selecting Analog Output, Voltage, ao0, and 1 Sample (On Demand) as before. To the left of this block, place a numeric constant, set its value to 5, and wire it to the D/A block input.

Step 3:

 Start the VI and verify that all is well.

Part 3: Smoothing

You have probably noticed that the displayed temperature is not very steady. This is due to undesired signals, or noise, finding their way into our system. Noise is like death and taxes: it will always be with us and the best we can hope for is to minimize its impact on what we are trying to do. (In fact, some types of noise can be eliminated by freezing the circuit to absolute zero. This treatment is also purported to eliminate death, but has no effect on taxes.)

Much of what we will do in subsequent Labs, especially Labs 7 and 8, will be devoted to minimizing the effect of noise. For now we will content ourselves with the most common response to unwanted variation in data: taking the average.


Step 1:

 Before we start to improve the situation, let's try to get a quantitative idea of how bad it is. At the moment what we know is that the displayed value jumps around a lot. Let's get a picture of those jumps.

Stop the VI and go to the front panel. Right click to bring up the Controls palette, go to Graph Inds, and select Chart from the Graph Indicators palette. Place the resulting chart in a convenient location on the front panel.

Step 2:

 Use Find Terminal or your own navigational skills to find the icon for the Waveform Chart on the block diagram. Position it directly above the icon for the VT numeric indicator. Connect its input to the T output of the formula node.

Step 3:

 Return to the front panel and start the VI. The chart will provide a graphic record of the variations in the temperature reading. Accumulate about a minute's worth of readings, then stop the VI and make a printout of the front panel.

Step 4:

 Go to the block diagram and widen the left-hand of the while loop by about 1.5 inch. Move the DAQ Assistant block to the left edge of the while loop. Disconnect the data output of the DAQ Assistant block from the input of the split signals block.

Step 5:

 Right click to bring up the Functions palette, select Analysis, then Statistics. Place the resulting block between the DAQ Assistant block and the split signals block. When the Configure Statistics dialog appears, select Arithmetic mean, then click OK. Connect the output of the DAQ Assistant block to the Signals input of the Statistics block. Connect the Arithmetic Mean output to the input of the split signals block.

Step 6:

 If we were to run the VI at this point, we would get the same behavior as before. This shouldn't be surprising, since the average of a single value (the 1 Sample from the A/D converter) is just the original value. What we need to do is take the average of a large number of input values to produce a single output.

Double click on the DAQ Assistant block to bring up the configuration dialog. Under Acquisition Mode, change 1 Sample (On Demand) to N Samples. Under Clock Settings, set Samples To Read to 1000. Since the Rate is 1000 samples per second, this will give us 1000 values to average. Click OK.

Step 7:

 Since the process of gathering the 1000 samples now consumes one second, the Time Delay block is no longer needed. Either delete it, or edit it and set the delay to zero.

Step 8:

 Return to the front panel and start the VI. The signal on the Waveform Chart should be much smoother. In fact we might be willing to believe that the variations that remain correspond to actual changes in temperature. Save a minute of new data and print a copy to compare with the unsmoothed plot.

Step 9:

 Stop the VI and save it in a persistent location.

Part 4: Measure Illumination

Having spent considerable effort developing a system for accurately measuring temperature, we would like to leverage this technology into new market areas. Since this technology is based on the measurement of resistance, we can use it to measure any physical phenomenon for which we have a resistance based transducer. A quick search of our parts kit turns up an ideal candidate for this new venture: the CdS photocell.


Step 1:

 In the circuit from the previous Part, replace the thermistor with your CdS photocell.

Step 2:

 Make a copy of the VI you used in the previous part and give it an appropriate name. Load this VI and go to the block diagram.

Step 3:

 Based on the information in the CdS photocell data sheet, derive a formula which gives the illumination level (in Lux) in terms of the resistance.

Question 1:

 Summarize your derivation in a form that will convince both you and your labbie of its correctness.

Step 4:

 Replace the formula in the formula node of the block diagram with the one you derived in the previous step. On the front panel, change the label of the temperature display from T to Illumination.

Step 5:

 Since light can vary much more rapidly than temperature, the Waveform Chart display would be more useful with a faster update rate. Edit the A/D converter block and change the Samples To Read from 1000 to 100.

Step 6:

 Start the VI and verify that it works correctly.

Step 7:

 Determine the illumination level under various conditions: under-shelf flourescent lamp on or off, incandescent lamp on or off, photocell shaded by lab notebook, etc. If necessary, clear the plot by right-clicking over the display window and selecting "Clear Chart".

Step 8:

 Turn off the incandescent lamp and under-shelf flourescent lamp. Clear the plot. Observe changes in the illumination seen by the photocell as people move around in the vicintity of the lab station.

Question 2:

 Suggest how the information in the illumination vs. time plot could be used in an intrusion detector.

Step 9:

 Set the (old-fashioned, analog) function generator to produce a 1 Hz, 8 V p-p square wave. Use BNC clip leads to connect a red LED to the 50 Ω Output (polarity is not important). The LED should be flashing at a rate of once per second.

Step 10:

 Hold the LED over the photocell, pointing downward. Observe the resulting waveform on the Waveform Chart display. Increase the distance between the LED and the photocell and note the maximum distance at which the signal from the LED is still discernable in the displayed waveform.

Remark:

 The last step provides an example of an optical communication system where the signal delivered to the LED is transfered over an optical channel to emerge (somewhat the worse for wear) some distance away as the output of the photocell. This is essentially the same arrangement we used in Part 3 of Experiment 2.3, except we're using the photocell instead of the photodiode.

The response of the photodiode is much faster than that of the photocell, but we found that its output voltage was a distorted version of the optical input. The I-V plot we made for the photodiode in Experiment 4.3 suggests the reason for this: viewed as a current source the photodiode's output is linear in the input irradiance, while viewed as a voltage source its output is logarithmic. We will deal with this next week by building a cicruit which converts this output current to a voltage.