[A83] Re: Quick quistion about djnz

To: "Assembly 83" <assembly-83@lists.ticalc.org>
Subject: [A83] Re: Quick quistion about djnz
From: "Ronald Teune" <rtwolf@gmx.net>
Date: Mon, 19 Nov 2001 16:39:29 +0100
Reply-To: assembly-83@lists.ticalc.org


From: Nicolas Gilles <kingofhackers2K@hotmail.com>

Hacking must be cool...

>From: "Ronald Teune" <rtwolf@gmx.net>

>
>
>>
>> >> >well in a way, all of these instructions aren't stored in ram.
>> >> >
>> >> >if it was stored in ram we would have a much larger area to work
>> >>with(well,
>> >> >83+ and the se have ram, but they wouldn't be running from ram).
they
>> >>are
>> >> >already on the chips memory.  so it's only the instruction execution
>> that
>> >> >takes any time, not loading it.
>> >
>> >hmmmm, everybody gets it the wrong way. I mean the transfer
>> >of the instruction from 9327h to the istruction register,
>> >if there is such a register. And where the instruction
>> >is decoded.
>>
>> Why do you have to transfer it? It's just a byte, and probably points to
a
>> location in the z80 or something
>>
>
>Well I'm not sure how exactly it works.. but I think it works kind of like
a
>multiplexer/demultiplexer + a bunch of State holders ....
>
>the chip sets certain 'lines'(as Ronald Teune said) high (logic 1) which
>tells what memory address to be at... the byte is just stored as open or
>closed gates (logic)

Of course...

|      _
| <=> (o)
           _
^         |\
|           \
The wall     My head banging on it


>and therefore the transfer from that depends only on
>the speed of electricity, and the time it takes the CPU to switch to the
>next memory address if it's a two byte opcode... and as far as I know it
>pretty fast.... enough to be faster that the individual transistors (that
>make up logic gate) can switch state :) which make the whole thing...
>work... :)
>
>uhm... the last part seems a little confusing but I can understand it.. so
>maybe you can too (come on ... just a little effort)...

I understand it.

>-Nicolas Gilles



From: Mike3465@aol.com <Mike3465@aol.com>

>no, the way ti works, is that it sort of has a register if you will, that
>says where it is in memory, from 0000h, to ffffh.  there isn't a loading
>register, it just executes from where it is.  the whole incrementing of it
is
>done automatically and is included in the tick count.

...but I don't understand this. :-o


>> >              SUCKER
>> >
>> >>
>> >>I think 'loading' wouldn't take time at all.
>> >>The instructions are probably somewhere on the chip with some address
>> >>lines,
>> >>and all you have to do is change the address lines, and you have it in
>the
>> >>blink of an eye, unlike a harddisk.
>> >>
>> >>Groeten,
>> >>         Ronald Teune
>> >>         Using Opera since 08/10/2001
>> >>         http://www.zeekoe.cjb.net
>> >>         http://cloud.prohosting.com/zeekoe *NEW* (English)
>> >>         iets@zeekoe.cjb.net
>> >>=================================
>> >>Anything that is good and useful is made of chocolate.
>> >>A hen is an eggs way to make a new egg.

>http://explorer.msn.com/intl.asp

Groeten,
        Ronald Teune
        Using Opera since 08/10/2001
        http://www.zeekoe.cjb.net
        http://cloud.prohosting.com/zeekoe *NEW* (English)
        iets@zeekoe.cjb.net
=================================
Anything that is good and useful is made of chocolate.
A hen is an eggs way to make a new egg.

Follow-Ups:

[A83] Re: Quick quistion about djnz

From: "Nicolas Gilles" <kingofhackers2K@hotmail.com>