EXPERIMENT #4: Arduino as Oscilloscope

Transcription

1 Name/NetID: Teammate/NetID: EXPERIMENT #4: Arduino as Oscilloscope Laboratory Outline This module adds the ability to visualize time-varying signals to your portable bench equipment. If you have completed the module that shows the different ways you can use the Arduino/RedBoard as a voltmeter you know that the Arduino USB interface with the computer works both ways. The incoming data are displayed in a window called the Serial Monitor that is part of the Arduino IDE. If the voltage actually varies more quickly and you want to capture and visualize the signal like an oscilloscope you can use a couple of methods to capture the data scrolling in the Serial Monitor. The simplest method is to cut and paste the data from the Serial Monitor into either Excel or MATLAB. Other methods use a programming environment like MATLAB that has the capability to access the serial interface at the command level providing a more elegant method of data capture a module coming soon.

2 The Analog Inputs on the Arduino This module uses the same Analog In pins as the Arduino as Voltmeter Module. These pins accept an analog voltage that can be time-varying or not and convert the continuous voltage into a number from every 100 microseconds. Let s see if we can plot this data so that we can visualize the signal in MATLAB or Excel providing you with a makeshift oscilloscope. Figure 1: Physical layout of the RedBoard.

3 Retrieving Serial Data Start with exactly the same set up as the module that uses the Arduino/RedBoard as a voltmeter. Set the power supply to +25V mode. Set the power supply to +5V. Hook-up the power supply to the analog pin labelled A0 as shown in the schematic below. The arrow shapes indicate which pin on your Arduino/RedBoard to connect to the positive and negative terminals of the power supply. Enter the code that outputs either the 10-bit integer from the A/D or the computed voltage to the Serial Monitor. You can use the same program used in the voltmeter module. For the observant yes there are examples of both under File > Examples > 01.Basics. Save the code in a file and save often during the lab so you do not lose your edits. After checking under the Tools menu that the software knows which board you are using (the RedBoard is a clone of the Arduino Uno) and which COM port you are using. When you plug the USB cable into the lab computer the associated COM port is usually the highest numbered port. For Mac users the USB communication ports are the device file names. Upload the code to the board by clicking the icon at the top of the window. Open the Serial Monitor by clicking on the icon at the top right of the window.

4 The problem now is how to get the data that is streaming from the Arduino/Redboard to the computer in a format that can be plotted. The Brute Force Method This method uses the very useful cut and paste feature included in the Arduino IDE (slight sarcasm). The data streaming to the Serial Monitor scrolls by at a set rate. Without a method to control either the sampling rate or the number of samples it is nigh impossible to select the data sample you want. Luckily, you can tailor the acquisition to your specifications by: i) slowing down the sampling rate, and ii) specifying the number of samples to print. Slowing Down the Display Rate The program that you uploaded to the Arduino may include a statement delay(int); where delay is a statement that uses the parameter int to suspend the program for int milliseconds where int is an unsigned long integer value. Unsigned means that the value must be positive so that the extra bit for the sign is not needed in the binary representation of the integer. Long means that the integer is stored using 32 bits rather than 16 bits providing a range of 4,294,967,295 (2 32-1). Change the delay statement to delay(1000) or add the statement at the end of the code in the loop() section; You should see that the display rate should have decreased. Using a timer (you can find one on the internet or your phone), check that the samples are showing up every 1000 milliseconds. Are they? How many hours of delay can you specify?

5 Limiting the Total Number of Points Measured Limiting the number of data points sent to the Serial Monitor allows you to capture time variations in the signal at a specified time. To do this you need to add a couple of statements to your code that execute ONLY when certain specified conditions arise. This class of statements referred to as Conditional statements or Control Structures will be used to program the board to take 100 data points and stop. In the code: 1. Insert the statement int length=100; just after the Serial.begin(9600); in the setup () portion of the code. This statement declares that an integer variable named length is created and can be used in the setup{} portion of the code ONLY. Length is initialized to Copy all of the statements in the loop section into the setup section after the int length=100; 3. In the setup section insert a while(condition){code segment} statement. The condition inside the parentheses tells the Arduino when to do all of the statements that are between the curly brackets {}. Put while(length>0){ just after the int length=100; statement. 4. Close the curly brackets so that the while statement includes ALL of the statements that need to run while length is greater than Leave the loop section empty. At this point, nothing about the program seems to have changed. Upload it, run it, and open the serial monitor to convince yourself. It does not do anything different because the variable length never changes inside the while loop. 6. Insert the statement length=length-1; inside the while loop somewhere. For those of you unaccustomed to programming this statement means that the current value in the variable length is replaced by the value length-1; After adding this statement length will decrease by one each time through the while loop. The C-like statement - length--; is also accepted. section? What is the purpose of taking the statements out of the loop section and putting them in the setup

6 Reset the board so that it runs the program again by either pushing reset button on the board or uploading the program again. While it is running, vary the power supply voltage so that the values change before the program stops. Be careful to stay below 5V. Open the Serial Monitor and select the data by hitting Cntrl-a, and copy the data to the clipboard by hitting Cntrl-c. Paste them into Excel 1. Open Excel and bring up an empty spreadsheet. 2. Click the first cell of the second row of the spreadsheet and Cntrl + v to paste the data into column B. 3. In column A you can construct a set of numbers corresponding to the time. Starting at 0 you can compute the relative time associated with the time each data point was taken since you know that there is a 1 s delay before the while() loop repeats. Paste them into MATLAB 1. Open MATLAB. 2. Find the workspace section mine looks like this 3. Click on the down arrow in the upper right corner and choose Paste to bring up the MATLAB import tool. 4. A window pops up the data already parsed into a column. Unless you change the name at the top of the column it will automatically name the variable something generic. Change the name to V1 or StormagedanDarkLordOfAll anything that is meaningful to you.

7 5. In the command window type time=0:timeincrement:(numpoints-1)*timeincrement where timeincrement is the time between sample in this case 1000 ms or 1 s and NumPoints is the total number of data points which you instructed to specify as 100. Another method is to type the statement time=linspace(0,(numpoints- 1)* timeincrement,numpoints). As with most complex environments like MATLAB there are many ways of doing the same thing. NOTE: for this module, to be sure that you are not going to run into trouble because you re-used variable names you can run the command clear at the MATLAB prompt when starting a new task. If you want to save all your results be sure to name variables and arrays using unique names. Plot the voltage signal using either Excel or MATLAB. Now you have data in either Excel or MATLAB. As with all devices that sample time-varying signals, there is a limitation on how fast the measuring device can sample. From the data sheets and forums the claim is a sample every 100 microseconds. In the Arduino code remove the delay function and set the number of points to any value you like as long as it is greater than 30. Make certain that the signal generator is in HIGH Z mode ask your TA for help checking this. Set the signal generator to output a sinewave with a frequency of 10Hz, peak-to-peak amplitude of 4V and an offset of 2V. IT is VERY IMPORTANT that the signal does not go negative. Actually the board is surprisingly robust as I just put 8V in and it is fine this is why the previous instruction about the HIGH Z mode was added. Using a BNC-to-BNC cable, connect channel 1 of the oscilloscope to the signal generator, and be certain that the sine wave stays positive just to be certain. Disconnect the oscilloscope and connect the signal generator to the Arduino using a BNC-to-banana cable and the special termination wires. The black connector is connected to GND and the red is connected to pin A0. Uploading the code again. Get the data into either Excel or MATLAB and make a plot of the data you do not know the time increment so the x-axes value corresponds to the numbers from for now. Plot the waveform.

8 Count the number of data points in one period. You can do this for several periods and average. Use the number of data points per period and the fact that the period of the 10Hz sinewave is.1 sec to determine the unadulterated sampling rate for a single analog input pin using this method. (A faster way would be to save all the data before printing it to the Serial Monitor.) The number you got is not very close to 100µs but for slowly varying signals like most of the signals you will encounter when connecting the Arduino s analog input pins to the sensors in your kits it should be fine. The slow speed comes from several sources including the fact that there is only one A/D on the processor chip so all of the analog pins need to share. There are ways to increase the speed if needed that are not too difficult if you are comfortable with assembly and interfacing with Python or MATLAB or writing to memory using pointers. Adding a Time-Stamp A number corresponding to an onboard time reference can be obtained and printed out to the serial monitor so that you know, approximately, when each sample was taken. Add the statement String toprint; in the statements before the setup section. This sets up a variable that can store text containing many characters in a row these types of variables are commonly called Strings in most programming languages. Add the following statements just after the statement that reads the analog value so that the while loop looks like the code below. These statements write the information in the correct format directly to the string before writing it out to the Serial Monitor.

9 Upload and run the code, then open the serial window. Copy and Paste the data into Excel or MATLAB o Into Excel choose the Home menu and click on the Paste icon all the way on the left. Choose the option Use Text Import Wizard. This will parse the string for you if you indicate that each line contains numbers separated by a comma. o Into MATLAB Follow the same steps and now you will see the data parsed into two columns. Rename the column data if you wish. The data at each time interval should include two numbers. The first is a time stamp or a number provided by the processor indicating the number of milliseconds from an arbitrary starting point. Plot the data using the time stamp as the x-axis. Is the interval between samples the same as the value you computed in Question 4? It should be close. Without some additional programming it seems that the Arduino analog pins and sampling is not very useful for high frequency signals. For our purposes observing how a sensor responds to a stimulus or how the signal you will use to control the turning on and off the motor varies it is fine. This method is a simple procedure to capture what you observe in the serial monitor. You will find this useful when you are trying out different ideas for your final project. If you want to improve the performance the Arduino can be programmed to be a decent oscilloscope. You can control it with a computer through MATLAB or Python to automate the process. The sampling rate can be improved by a factor of at least 100 by delving into the workings of the processor in a variety of ways. Look for a future module.