Yes, that was a brute-force quick-n-dirty
method. I'm sure there are "better" methods
with fewer instructions using loops.

B.t.w, have you verified your code below ?
Should be easy using MPSIM in MPLAB.

*I* am not going to verify it...

Jan-Erik.

Harry H. Arends wrote:
> Hello Jam-Erik,
> 
> Following your recomendations i came up with this. There are a lot of bits
> moved forward and back :-)
> 
> SetData
> 	movlw		b'00001111'	;*set preamble
> 	movwf		DATA0
> 	movlw		b'10000001'	;*bit<7:6> alsways 10
> 	movwf		DATA1		;*set low byte address
> 	movlw		b'11111001'	;*set command
> 	movwf		DATA2
> 	xorwf		DATA1,0
> 	movwf		DATA3		;*save controlbyte
> 
> 	bsf		STATUS, C	;*set Carybit to be insert at
> SHIFT3<0>
> 	rlf		DATA3		;*create endbit
> 	rlf		SHIFT4	;*SHIFT4<0> holds DATA3<7>
> 
> 	bcf		STATUS, C	;*reset Carybit to be insert at
> DATA2<0>
> 	rlf		DATA2		;*and place it in DATA2
> 	rrf		SHIFT5	;*SHIFT5<0> holds DATA2<7>
> 	rrf		SHIFT4	;*retrieve DATA3<7>
> 	rlf		DATA2		;*move it to DATA2<0>
> 	rrf		SHIFT5	;*SHIFT5<0:1> holds now DATA2<7:6>
> 
> 	bcf		STATUS, C	;*reset Carybit to be insert at
> DATA1<0>
> 	rlf		DATA1		;*and place it in DATA1		
> 	rrf		SHIFT3	;*SHIFT3<0> holds DATA1<7>
> 	rlf		SHIFT5	;*get DATA2<6>
> 	rlf		SHIFT5	;*get DATA2<6>
> 	rlf		DATA1		;*move it to DATA1<0>
> 	rrf		SHIFT3	;*SHIFT3<0:1> holds now DATA1<7:6>
> 	rrf		SHIFT5	;*get saved DATA2<6>
> 	rlf		DATA1		;*and insert it
> 	rrf		SHIFT3	;*SHIFT3<0:1:2> holds now DATA1<7:6:5>
> 
> 	bcf		STATUS, C	;*Create Carybit to be inserted at
> DATA0<0>
> 	rlf		DATA0		;*and place it in DATA0
> 	rlf		SHIFT3	;*Create DATA0 using DATA1<7:6:5:4> from
> SHIFT3
> 	rlf		SHIFT3
> 	rlf		SHIFT3
> 	rlf		DATA0
> 	rrf		SHIFT3
> 	rlf		DATA0
> 	rrf		SHIFT3
> 	rlf		DATA0
> 
> 	bsf		NEW_PACKET	;*new packet received
> 	bcf		DCC4BYTE	;*if set there is a 4 byte packet
> 	return 
> 
> If it can be done easyer i am happy to hear that.
> 
> Harry
> 
> -----Oorspronkelijk bericht-----
> Van: piclist-bounces@mit.edu [mailto:piclist-bounces@mit.edu] Namens
> Jan-Erik Soderholm
> Verzonden: dinsdag 12 augustus 2008 0:12
> Aan: Microcontroller discussion list - Public.
> Onderwerp: Re: [PIC] Packets problem to be solved
> 
> Argh !
> Got a few DATA/BYTE's and some
> bit numbers mangled.
> Corrected version below.
> Jan-Erik.
> 
> 
> 
> Jan-Erik Soderholm wrote:
>> Harry H. Arends wrote:
>>
>>>> OK by now i am aware to be more specifik. The language is ASM.
>>>> The data to be sendout...
>> Can you better define what "sendout" is ?
>> How will it be sent ?
>> Using what interface ?
>>
>>>> is in four 8 bit variables:
>>>> DATA0 contains all ones
>>>> DATA1 contains a 8 bit address (AAAAAAAA)
>>>> DATA2 contains a 8 bit command (DDDDDDDD)
>>>> DATA3 contains a bitwise exclusif OR from the DATA1 and DATA2
>>>>
>>>> The result should become
>>>> BYTE0 11110AAA
>> Let's number the bit positions 0-7.
>> Let's also assume that the bits are 76543210.
>>
>> So BYTE0.2 = DATA1.7, BYTE0.1 = DATA1.6 and BYTE0.0 = DATA1.5, right ?
>>
>> That could be done with a few bit-test and bit set/clear instructions.
>>
>>>> BYTE1 AAAAA0DD
>> OK, so rotate DATA1 3 positions to the left and move to BYTE1.
>> Then set BYTE1.1 = DATA2.7 and BYTE1.0 = DATA2.6.
>>
>>>> BYTE2 DDDDDD0E
>> Rotate DATA2 2 positions left and move to BYTE2.
>> Then set BYTE2.0 = DATA3.7.
>>
>>>> BYTE3 EEEEEEE1
>> Rotate DATA3 1 position to the left and move to BYTE3. Set BYTE3.0 = 
>> "1".
>>
>> Then just do that in assembler.
>>
>> *THEN* send BYTE0 to BYTE3 using whatever method you'd like.
>>
>> Jan-Erik.
>>
> --
> http://www.piclist.com PIC/SX FAQ & list archive View/change your membership
> options at http://mailman.mit.edu/mailman/listinfo/piclist
> 
> 
-- 
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist