Description

This set of experiments has the following objectives:

• Familiarization with advanced instructions of 8051 family.
• Familiarization of breaking down a problem into subproblems and using subroutines to solve a problem.

You should browse through the manual for Keil software and the data sheet for 89C5131A uploaded on moodle site. Specifically, refer to these to understand the memory-map.

89C5131A has 256 bytes of RAM, of which the top 128 bytes can only be accessed using indirect addressing. Since the stack is always accessed through indirect addressing using the stack pointer, we shall locate the stack in this part of the memory. In our assembly programs, we shall initialize the stack pointer to 0CFH, which provides 32 bytes for the stack. It is common to place arrays also in this part of this memory, because arrays are also easily accessed using a pointer. The address range from 80H to CFH is available for this (since memory beyond this is reserved for the stack).

Refer to the textbook by Mazidi and Mazidi for syntax of instructions.

1           Homework

1. Understand the sample (sample led.asm) program provided (in zip file) before getting started on these exercises. Specifically, understand how the code is written down as subroutines to make the code readable. Also, the subroutines use PUSH and POP instructions for storing and restoring the callers’ registers. (Viva 5 points)

You will be using a board in the later weeks which uses a 24MHz crystal. In order for your simulation to match design running on the board, you have to setup the crystal frequency in the simulator to be 24MHz.

2. Write an assembly program to blink an LED at port P1.7 at specific intervals. At location 4FH a user specified integer D is stored. You should write a subroutine called delay. When it is called it should read the value of D and insert a delay of D/2 seconds. Then write a main program which will call delay in a loop and blink an LED by turning it ON for D/2 seconds and OFF for D/2 seconds. D will satisfy the following constraint:

1 ≤D≤ 10.

You can use the code sequence below to introduce a delay of ∼50ms.

MOV R2,#200 BACK1: MOV R1,#0FFH BACK :

DJNZ R1, BACK

DJNZ R2, BACK1

The counters R1 and R2 are setup in such a way that the code above would take roughly 50ms. See the Appendix to know how the delay is calculated.

You have to nest this inside another loop with appropriate counters to get the required delay.(5 points)

You should assemble and debug these programs using Keil software on a PC or laptop. You should check that the program is operating correctly. If necessary, use single stepping and breakpoints provided by Keil Software.

2           Lab work

2.1         Problem1:Clear Memory ZeroOut

Write a subroutine zeroOut which will read a number N from location 50H and a pointer P from 51H. The subroutine should zero out the contents of memory in N consecutive locations starting at P. N will satisfy the following constraint: 1 ≤N ≤ 16. This routine can be used to initialize memory before usage.(5 points)

2.2         Problem2:Nibble display on Port P1

Write a subroutine display which will read a number N from location 50H and a pointer P from 51H. The subroutine must read the last four bits of the values at locations P to P +N−1 and display on the LEDs, one at a time. There should be a delay of 1s between each such display. Use the ports P1.4 to P1.7 for the 4 LEDs. Note that the simulator supports display of 8 output ports but only 4 LEDs are present in the board. Later, you will be able to use this subroutine as is when we try designs on the board.(5 points)

2.3         Problem3:ASCII Conversion

Write a subroutine bin2ascii checksumbyte that will read a number N from location 50H, a read pointer P1 from 51H, write pointer P2 from 52H. The subroutine should compute the checksum byte (as done in lab-1). The checksum byte should be at the next location after the N bytes. The lower and higher nibbles of the checksum byte should then be converted into a pair of ascii (8-bit) format numbers with higher nibble in first and lower nibble in next memory location specified by pointer P2.(10 points)

For example: if checksum byte is 1FH then the two ascii bytes should be 31H and 46H representing characters 1 and F respectively. If the write pointer stored at 52H is 60H, then these two bytes 31H and 46H should respectively be stored in locations 60H and 61H.

2.4         Problem4:Memory Block Copy

There are many circumstances where a series of memory locations are copied to another location. For example, a network packet captured in a certain memory location by a network device may get copied by the OS to another location where an application processes the packet.

Write a subroutine memcpy which will read a number N from location 50H, a pointer A from 51H and another pointer B from 52H. It should then copy N locations from address A to address B. Assume that the N is smaller than the difference between the addresses A and

B so that overlap issues do not arise.(5 points)

3           Learning Checkpoints

Following tasks have to be shown to TA to get full credit for this experiment.

1. Initialize an array to any number other than zero using the subroutine zeroOut and clear it using the same subroutine.
2. Explain the purpose of initializing stack pointer to 0CFH. What is the content of SP when 8051 is powered up?

4           Appendix

The code segment below can be used to get a delay of ∼50ms:

MOV R2,#200 BACK1: MOV R1,#0FFH BACK :

DJNZ R1, BACK

DJNZ R2, BACK1

The delay calculations are as follows. MOV and DJNZ instructions take 1 and 2 machine cycles respectively. Therefore, the number of machine cycles required for the for loop is 1+255*2 = 511 machine cycles. Since this is inside a loop where R2 = 200, the outer loop at BACK1 takes 200*511 + 200*2 = 102,600 machine cycles. Each machine cycle is 12 clock cycles making the total number of clock cycles 1, 231, 200 clock cycles. Dividing this by the crystal frequency (24 MHz), we get a delay of ∼50ms.