Файл: The quintessential PIC microcontroller (S. Katzen, 2000).pdf

THE ESSENCE OF THE PIC MICROCONTROLLER 113

Program 5.2 Finding the maximum temperature using a loop structure.

;Task3: DO

;Task3A: IF Temp[i] > Maximum THEN Maximum = Temp[i]

; Task4: END

loop pass. The worst-case run time is 291 µs against 188 µs, assuming a 4 MHz crystal.

One advantage of the Indirect address mode is that the pointer address is eight bits wide. Thus it is not necessary to use the RP0 bit to switch between Bank 0 and Bank 1 of the Data store. The full 14-bit core model allows for four 128-byte banks of register files. The Status register of Fig. 5.1 shows the IRP (Indirect Register Page) bit which is used for family members with three or four register file banks. In such a device, if the temperature array of Program 5.2 were located in File 130h through File 147h then IRP would need to be set to 1 at the beginning of the routine (bsf STATUS,7) and cleared at the end. The rest of the code is unaltered. Microchip recommend that the IRP (and RP1) bits in

116 The Quintessential PIC Microcontroller

•movlw copies the specified 8-bit constant (or literal) to W. For example, movlw 80h initializes W to 10000000b. This instruction only a ects the Working register and thus cannot be used directly to set up a file register to a constant value.

•movwf is used to copy out or store the contents of W into a register file. For example, the following code fragment will initialize the contents of

File 22h to 80h:

movlw 80h ; Set contents of W to 80h movwf 22h ; and copy to File 22h

•movf can copy (or load) the contents of any register file into W. For example, movf 22h,w loads W with the contents of File 22h.

The destination of this instruction can also be the file itself, giving rise to seemingly useless instructions, such as movf 22h,f which copies the contents of File 22h back on top of itself! However, the process does activate the Z flag, which will be set if the file contents are zero, and of course this datum is not a ected by the instruction. Thus movf FILE,f is equivalent to the missing tstf FILE instruction; that is TeST File for zero, that is commonly available in other MPU/MCUs. Thus the contents of any file register can be checked for zero by this means using a single instruction. An alternative technique needs to be used to test the contents of the Working register for zero.

Given that most instructions acting on a file can specify either the same file or the Working register as the destination, then a Move operation can be considered an implicit part of such instructions. As an example, for some situations to increment the contents of a file and then move it to W could be coded either as:

or

incf 22h,w ; Copy the incremented File 22h’s contents to W

Of course the latter does not actually change the state of the file.

The final instruction swapf swaps the top and bottom 4-bit nybbles in a file. Thus, for example, if File 22h was 1001 0111b then swapf 22h,f will yield 0111 1001b. If desired, the outcome destination could be specified as W, eg. swapf 22h,w. As swapf does not a ect any flag, this latter form can be used as a transparent replacement for movf 22h,w which does alter the Z flag. Of course it does interchange the two nybbles in the process. Program 7.2 on page 183 shows this swap instruction used in this role.

Arithmetic

The PIC processors implement the normal byte-sized binary Add and Subtract instructions, as discussed on page 47, to add or subtract register

THE ESSENCE OF THE PIC MICROCONTROLLER 117

file contents to/from the Working register. In addition the W register may be added to or subtracted from an 8-bit constant.

Table 5.2: Arithmetic.

As an example that uses most of the instructions in Table 5.2 consider the problem of dividing the contents of File 24h by an 8-bit divisor in W. The simplest way of doing this is to continually subtract the divisor from the dividend, keeping a count until a borrow is generated. The residue left after this last subtraction is the remainder with one divisor subtraction to many. Thus adding the divisor once can restore the remainder if this is needed.

A possible implementation based on this approach is given in Program 5.3. Here the quotient is cleared before entering the loop, using the clrf instruction. The loop itself simply increments the quotient using the incf instruction and then subtracts W (the divisor) from File 24h – subwf 24h,f. Both Subtract instructions generate a complement borrow out, which is represented by the C flag in the Status register – labelled NB for Not Borrow in the program. Thus the loop is exited when the Carry flag is clear after the subtract, which represents a borrow out.

On leaving the loop, the contents of File 20h needs to be decremented, as the last subtract was one too many. Using the decf instruction allows this correction to be applied directly on the file register.

THE ESSENCE OF THE PIC MICROCONTROLLER 119

Example 3.2 on page 64 gives a practical implementation of this algorithm. Multiple-precision subtraction is carried out in a similar manner, using the Not-Borrow flag. Example 5.4 implements a 16-bit − 8-bit subtraction.

Data in memory can be incremented or decremented apparently in situ, although in reality it is transferred from the Data store into a temporary register, incremented or decremented using the ALU and transferred back to the Data store – a type of read-modify-write action. However, it still takes only one bus cycle to implement.

These instructions are especially useful in counting passes through a loop, as in Program 5.3 where QUOTIENT is located in the Data store at File 20h. However, incf is not quite the same as a addlf 1,20h type of instruction as it does not alter the state of the Carry flag. Thus if you wanted to increment a 32-bit number in Data memory at File 22:3:4:5h then this is how you would have to do it:

This depends on the algorithm IF when byte n is incremented it wraps around from FFh to zero THEN increment byte n +1. See Example 5.1 for a multiple-precision decrement routine.

One of the more important operations is the comparison of the magnitude of the two numbers. Mathematically this can be done by subtracting the datum (designated [f] for either a register file or a literal) from the contents of the Working register [W]. The outcome gives the actual magnitude di erence between the operands, but in most cases it is su cient to determine the relative magnitude of the quantities – eg. is W higher than the datum? This is determined by checking the state of the C and Z flags in the Status register.

Working register higher than datum . . . . . . . . . . . . . . . .No borrow, non-zero Working register equal to datum . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Zero Working register lower than datum . . . . . . . . . . . . . . . . . . . . .Borrow, non-zero

In terms of our processor, the C flag represents the complement of the borrow after subtraction and the Z flag is set on a zero outcome. This gives: