Questions about PASD

Mazzini

Hello again,

I'm trying to write my driver for LCD based on this guide
http://www.8051projects.net/lcd-interfacing/lcd-4-bit-programming.php
my code:
http://pastebin.com/PuyMSDXc


I'am debugging my driver by PASD
Thank to Stefan and Frank I modified my code in LcdDat
but i dont uderstand what pasd is doing, you can see the shot here

These shot describes how call LCDdat ,load 40 DEC on cmd var, copy cmd var to t2
but it doesn't work



any suggestions ?:)
Thanks

kuroneko

msb                     byte
lsb                     byte
cmd                     res 1

msb and lsb don't have a value (just a qualifier) which means they end up in the same place as cmd which will cause confusion (it certainly confuses PASD). Just change them to byte 0 or better long 0 to keep them separate.

kuroneko

Also note that or is cumulative, i.e. you keep or'ing stuff to the data bus (outa[10..13], or outa, t2) which means that once it reaches %1111 you're stuck there.

Mazzini

kuroneko wrote: »

msb                     byte
lsb                     byte
cmd                     res 1

msb and lsb don't have a value (just a qualifier) which means they end up in the same place as cmd which will cause confusion (it certainly confuses PASD). Just change them to byte 0 or better long 0 to keep them separate.

it works thanks

[QUOTE=kuroneko;
Also note that or is cumulative, i.e. you keep or'ing stuff to the data bus (outa[10..13], or outa, t2) which means that once it reaches %1111 you're stuck there.
[/QUOTE]

I'am sure you are right because your experience,
my purpose is update the port without overwrite all bits

OUTA ----> 101100001111 or
t2   ----> 000011110000
          ________________
           101111111111

kuroneko

my purpose is update the port without overwrite all bits

Understood. But look what happens next:

OUTA ----> 1011[COLOR="red"]0000[/COLOR]1111 or
t2   ----> 0000[COLOR="blue"]1111[/COLOR]0000
          ________________
           1011[COLOR="red"]1111[/COLOR]1111

Unless you clear the data portion of outa the next cycle will look like this:

OUTA ----> 1011[COLOR="red"]1111[/COLOR]1111 or
t2   ----> 0000[COLOR="blue"]1100[/COLOR]0000
          ________________
           1011[COLOR="red"]1111[/COLOR]1111

Do you see the problem here? Easiest solution is to make a hole just before you add t2 to outa. For example

andn    outa, CLR
        or      outa, t2
        ...
CLR     long    %1111 << 10

Mazzini

aaaa I feel a bit silly

I understood ...Thanks kuroneko, thank for your time

StefanL38

Hi Kuroneko,

if I do a andn directly with outa as destination am I right that this will result in the involved bits turn to zero and then with the or some of them turn on again?

If so this should only be done if the logic levels of the connected device are not "taken in" immediatly f.e because there is a "data-valid" or latch-input wich must have a certain state
to "take in" the logic levels "sended" from the prop-pins.

In case of logic levels are taken in immediatly into the other device I would have to go the way
- copying actual state of outa to a temporary long
- clearing the involved bits
- oring the new bits to the temporary long
- writing (not "OR"ing) the new 32bit value to outa

best regards

Stefan

kuroneko

@StefanL38: Yes, obviously. But given that talking to the LCD involved an ENable signal I'd argue that you can get away with in-place modification

frank freedman

Mazzini wrote: »

it works thanks



I'am sure you are right because your experience,
my purpose is update the port without overwrite all bits

OUTA ----> 101100001111 or
t2   ----> 000011110000
          ________________
           101111111111

That is why the use of AND or XOR are good to know. But I'm starting to think that the MUXs could become my main bit handlers.

As to the source for code, have you checked the object exchange OBEX? There are a number of display drivers including those for common LCD panels that may be easier to alter to meet your needs. They may only require minor modifications of say pin selections etc. Easier than starting from a different device.