CS-280 Lab 7: Interrupt-Driven Feedback Control

by Dr. Eric A. Durant
This lab was developed in conjunction with Dr. Henry L Welch.

Overview

In this lab, you will expand upon your work from lab 6 to develop a programmable "DC" power supply. Using assembly language you will develop an application that controls whether the capacitor in the RC circuit is charging or discharging in an attempt to generate a "constant" DC value in the range from 0–5 VDC. This is not the best way to implement a variable DC supply, but for our purposes it will introduce you to timing interrupts and the basic concepts of feedback control.

This is a 2-week assignment. You may work alone or with a partner and submit a group report.

Lab activity

As with lab 6 you should complete this assignment entirely in assembly language. You will reuse the getkey routine from lab 4.

Wire the circuit as shown to the right. You should recognize that it is essentially the same as the lab 6 circuit, except that the pushbutton has been replaced by one of the HC11’s digital outputs. Also, to have more agreeable time constants, the resistors have been swapped (at the cost of a higher minimum voltage – there are resistors on the Fox11 between your circuit and the HC11 pins that limit the minimum voltage, especially as RF1 becomes large).

Use fox11_portb (any one of the 8 pins) for your output. This will allow you to monitor your control pin via a built-in LED. These pins can be accessed using the large header just below the HC11 chip on your Fox11.

Modify your program from lab 6 so that in addition to monitoring the voltage using the A/D system and displaying the results (0–500, i.e. 0.00 V to 5.00 V) on the 2×16 character Microtips/Hitachi display, it also actively controls the analog voltage on the capacitor. This should work as follows:

1. Accept user input of a desired voltage using up to three keypresses (0–500). “5” should be interpreted as 0.05 V, “50” or “050” should be interpreted as 0.50 V, and “23456” should have characters after the 3rd ignored and be interpreted as 2.34 V. You are not required to do special error handling for inputs such as “740”—this is just a voltage that is higher than attainable, so your algorithm should naturally seek out the maximum voltage when given such a large value.

2. The display should show the actual voltage (using the averaging and scaling technique from lab 6) on the top line.

   The display should show either the voltage being entered or the target voltage (if the input buffer is empty) on the bottom line.

   The display should be updated about 10 times per second via a dedicated timer interrupt.

3. Develop another ISR that monitors the analog voltage and either sets or clears the control pin as follows about 100 times per second:
   • If the voltage is too low output a 1
   • If the voltage is too high output a 0
4. Continue to monitor the keypad (use a polling loop, i.e., the getkey function) with the following functionality:
   • "0–9" add a digit to the displayed entry if the 3-character input buffer is not full
   • "A" increases the desired output by 0.10 V. (do not go over limit of 9.99)
   • "B" decreases the desired output by 0.10 V. (do not go under limit of 0.00)
   • "E" enter the numeric entry in progress
   • "F" backspace—Be sure to handle the case where there is no entry in progress.
5. Under all circumstances the program should continue to control the voltage; even in the middle of a numeric entry. You may find a DMM useful in determining this as well as the general accuracy of your voltage supply.

Note: When the program starts up, the target voltage should be 2.50 V.

The component values may vary somewhat without materially affecting operation of the circuit; RF2 should be small, but not so small as to draw too much current when the control pin goes to ground. The RC time constant RF1×CF should be between roughly 0.5 and 1.0 s.

Hints and Caveats

• Setting the timer prescaler select from ASM will not work on our platform since there is no convenient way to get code to run in the first 64 cycles after reset. You will need to use the ‘I’ command in the Wytec debugger to set TMSK2 to 0x03 (make sure that you understand why this is the correct value). Once you change this, your setting will be remembered for the active S19 file, so you do not need to do this repeatedly. However, some features of the Wytec debugger are affected by using a non-default value for the prescaler (TMSK2:PR). So, you should debug your code with TMSK2:PR==0. You may use either TMSK2:PR==0 (come as close to the timing requirements as possible) or TMSK2:PR==3 for final testing and your demonstration. (An example of a debugging problem with PR!=0 is that single-step will actually step several instructions.)
• PE0-PE3 are used for the keypad, so it is recommended that you use PE4 for your analog input. Port E can be used for analog and digital input simultaneously. Digital reads are done via “porte” and analog reads are done via “adrN”.
• Set-up the A/D in continuous conversion mode. Then you do not need to poll for A/D completion. Just look in the ADR* registers for the most recent values when you need them.
• Use two separate timer interrupts—one for the display and one for the digital control output. Be sure to schedule the next interrupt during the ISR. A very common error is forgetting to clear the interrupt flag before ending the ISR. This will cause your system to continually handle interrupts and not accept key input, etc.
• Design and test the program in small chunks. For example, get the display interrupt working and make sure it shows you the proper voltage (try manually connecting the digital input to logic 0 and logic 1). Then, enable key input and setting of the target voltage. Finally, implement and enable the control interrupt. A few different implementation paths are possible; this is only a suggestion.
• Although Wookie does not simulate the A/D system or your 2×16 display, it does simulate timer interrupts. So, you can use Wookie with the A/D in continuous conversion mode and the Registers window open for manually modifying the ADR* registers to test part of your program.

Report (due by Friday of Week 10, 11 P.M.)

Note: No late reports will be accepted after 11 P.M. Wednesday of week 11.

• Prepare a report in Microsoft Word or PDF.
  • Include source files that you created or modified in your report file.
  • Summarize how much time each group member spent on the lab using the following categories: design, code, test, debug, report writing, and other.
  • Document your design.
  • Discuss how the various parts of your program shared the desired voltage. Be sure to mention all parts of the program that needed access to this value.
  • Instead of sample output, discuss the behavior of your system (accuracy, speed) for various inputs.
  • Detail any problems you encountered and any questions that remain.
• Create a Zip archive including your report, source files, final listing files (.rst), and executable file (.s19).
• Follow the report submission requirements.
• Email this file to the instructor with a subject and message body indicating that this is your CS-280 Lab 7 submission.