Hello Friends:

Thanks to Jerson who responded earlier.
I am able to get the floating point routine to work. I however, find the results to not be accurate or most times incorrect.

For example, dividing 148,000/12,700. The result should equal 11.6535.

The function returns the following:
Float Result = 3a74e9
Integer Result = 00000b :What happend to the .6535 ?

Examining the Float Result, I viewed a DECIMAL to ASCII chart.
It reveals the following:

"3a" = 58 ;"74"= 116; e = "14"; 9 = "9"

If I manipulate the order, this "could" mean:
"11658149" or
"11.658149"

Then if I included a decimal.It looks "almost" correct. (I'm guessing here...)

In my test code below, I can check the results of each function call to verify it returns correctly, by sending the results to the Serial Port for observation.

All the calls to int2float and float2int are correct. The results from the fpdiv32, not so correct.

Has anyone "successfully" used this routine?

A little note about my project:

My P89C668 is connected to a readout encoder. This encoder outputs a 24 bit binary code. If I divide this count by 12,700, I get a value in "inches". I now want to display this "value in inches" to an LCD display.

My guess is I needed this floating point routine first. Next I will need a routine to convert the "float to ascii" for sending to the LCD. Is my solution correct? Or am I going about this incorrectly?

Thanks in advance!

  CMD1:
	mov		Aarg,#9CH	     ;Denominator
	mov		Aarg+1,#31H
	mov		Aarg+2,#00H	
	acall	int2float	   
	
	;Debugging - Let's check result from above int2float() - 
	;
	MOV     ANSWER,Aarg+2      ;MSB	
        MOV     ANSWER+1,Aarg+1  
	MOV     ANSWER+2,Aarg+0    ;LSB
	MOV     ANSWER+3,Aexp
	;
	;Send value to Serial Display[1]
	;
	MOV     R0,#ANSWER        
	LCALL	PRNT_4H
	LCALL   CRLF
	;
	acall	copy_ab
	;
	mov		Aarg,#20H		  ;Numerator
	mov		Aarg+1,#42H
	mov		Aarg+2,#02H
	acall	int2float
	;
	;Debugging - Let's check result from above int2float() - 
	;
	MOV     ANSWER,Aarg+2      ;MSB	
        MOV     ANSWER+1,Aarg+1  
	MOV     ANSWER+2,Aarg+0    ;LSB
	MOV     ANSWER+3,Aexp
	;
	;Send value to Serial Display[2]
	;
	MOV     R0,#ANSWER        
	LCALL	PRNT_4H
	LCALL   CRLF
	;
	;Sanity Check - Let's convert Integer to Float to       if it returns
	;our orignally value of 0x3E8, before we multiply.
	;
	;lcall   float2int
	;Debugging - Checking float2Int() - Expecting 0x3E8
	;
	;MOV     ANSWER,Aarg+2      ;MSB	
        ;MOV     ANSWER+1,Aarg+1  
	;MOV     ANSWER+2,Aarg+0    ;LSB
	;
	;Send value to Serial Display[3]
	;
	;MOV     R0,#ANSWER        
	;LCALL	PRNT_6H
	;LCALL   CRLF
	;
	;
	;Floating point multiplication
        ;Input	- Aarg,Aexp
        ;        - Barg,Bexp
	;
	;Let's see if the Inputs (Aarg,Aexp ; Barg,Bexp) are correct.
	;
	;MOV     ANSWER,Aarg+2      ;MSB	
        ;MOV     ANSWER+1,Aarg+1  
	;MOV     ANSWER+2,Aarg+0    ;LSB
	;MOV     ANSWER+3,Aexp
	;MOV     ANSWER+4,Barg+2
	;MOV     ANSWER+5,Barg+1
	;MOV     ANSWER+6,Barg+0
	;MOV     ANSWER+7,Bexp
	;Send value to Serial Display[4]
	;
	;MOV     R0,#ANSWER        
	;LCALL	PRNT_8H
	;LCALL   CRLF
	;
	acall	fpdiv32	    
	;
	;Lets check the FLOAT result of the fpmul32() BEFORE we convert the result to Integer.
	;
	MOV     ANSWER,Aarg+2      ;MSB	
        MOV     ANSWER+1,Aarg+1  
	MOV     ANSWER+2,Aarg+0    ;LSB
	;
	;Send value to Serial Display[5]
	;
	MOV     R0,#ANSWER        
	LCALL	  PRNT_3H
	LCALL   CRLF
	;
	;
	acall	float2int
	;
	;Debugging - Let's check result from multiplication - Result is INTEGER-
	;
	MOV     ANSWER,Aarg+2      ;MSB	
        MOV     ANSWER+1,Aarg+1  
	MOV     ANSWER+2,Aarg+0    ;LSB
	;
	;Send value to Serial Display[6]
	;
	MOV     R0,#ANSWER        
	LCALL	  PRNT_3H
	LCALL   CRLF
	;
	RET