AVR-Chat

The problem is the "negate" step. You cannot negate the two bytes 
separately. Negation is the same as a ones-complement followed by 
adding one. This is covered in Atmel's application note #202. They 
do it like this:

;***** Register Variables
.def ng1l = r16
.def ng1h = r17

;***** Code
ng16:	
 com    ng1l            ;Invert low byte       ;Calculated by 
 com    ng1h		;Invert high byte      ;inverting all 
 subi   ng1l,low(-1)    ;Add 0x0001, low byte  ;bits then adding
 sbci   ng1h,high(-1)   ;Add high byte         ;one (0x0001)
 ;Expected result is 0xCBEE

Jim


--- In AVR-Chat@yahoogroups.com, "Leon Heller" <leon.heller@d...> 
wrote:
> ----- Original Message ----- 
> From: Geert De Pecker
> To: AVR-Chat@yahoogroups.com
> Sent: Sunday, January 02, 2005 10:21 PM
> Subject: [AVR-Chat] newbie absolute value
> 
> 
> Hi,
> 
> I have a question about abtaining the absolute value. I want to
> substract a 1 16bit values from eachother, but when the result is
> negative, set a bit and get the absolute value of the difference.
> 
> I thought of the following, but that doesn't seem to work:
> 
> ...
>    ldi direction, 0
> ; Substract values
>    sub moveL, curPosL
>    sbc moveH, curPosH
> ; If result positive, ok
>    brcc PM1
> ; Result negative, take absolute value and set flag
>    neg moveL
>    neg moveH
>    ldi direction, 1
> PM1:
> 
> Any hints?
> 
> I'd compare them first, to find which is the bigger of the two, 
then 
> subtract them appropriately with two separate bits of code, 
setting the flag 
> as necessary.
> 
> Leon 
> 
> 
> 
> -- 
> No virus found in this outgoing message.
> Checked by AVG Anti-Virus.
> Version: 7.0.298 / Virus Database: 265.6.7 - Release Date: 
30/12/2004