PDA

View Full Version : Internal character ROM

Cobalt
10-23-2006, 06:36 PM
I have a graphic LCD that I've been able to interface to, but it doesn't have a built-in character table. No problem since the Propeller does! I want to be able to write from the charater map in the Prop's ROM to the LCD, but I have one little problem:
I don't know how to get the data from the ROM http://forums.parallax.com/images/smilies/sad.gif
I'm sure it is just very simple code, but I would like some help on how to access the ROM to send it to my LCD. Thanks.

Graham Stabler
10-23-2006, 07:12 PM
Check out page 236 of the manual, the long function can be used to get a long from a given address, the addresses are on page 32

Graham

Cobalt
10-23-2006, 09:58 PM
Thanks for that bit of into, I didn't know that 'long' would work that way in a PUB/PRI
Let me see if I am reading this right:

VAR
long temp

PUB Main
temp := long[\$8000]

This would put the long from \$8000 (the start of the character ROM) into the long var "charlong" I take it? (I ask because I'm at school still, not in front of my stuff) I also assume that when I get a long from the ROM, it is a row of data from left to right... am I correct at this too?

Also, to clairify the character ROM design, the characters are kind of 'meshed' together into a 32x32 bit square (4 longs x 32)?
I read the section at the start of the manual (p. 32) with the character map, but it really didn't make 100% sence.

I need a "Propeller for Dummies" book!

Graham Stabler
10-23-2006, 10:19 PM
yes long is a variable type and a function.

What you have written looks right to me but it would provide you with the first row of the first two characters, each character pair is made up of 32 longs (which are 32bit each).

The long you read is just a number so its 01010101_01010101_01010101_01010101 or something

It says the left most pixel is the lowest bit so I assume that means that the lsb is the top left pixel of the first character (which is why the last diagram shows it flipped). The next bit is the top left of the other character and so it goes on. Considering the letters shown it goes:

(top right pixel) BABABABABABABABABABABABABABABABA(top left pixel,lsb)
Next long, next row
BABABABABABABABABABABABABABABABA
Next long, next row
BABABABABABABABABABABABABABABABA
.
.
.
32 longs later
(bottom right) BABABABABABABABABABABABABABABABA (bottom left, lsb)

I think the wording of the manual tends to be correct rather than helpful.

Graham

Beau Schwabe
10-23-2006, 10:25 PM
Cobalt (http://forums.parallax.com/member.php?u=47565),
·
There is probably an easier way to do this, but here is·the formula·I used to extract the 'meshed' ROM data in·the VGA graphics driver.
·
The Character table is located at \$8000
·
Y ranges from 0 to 31
X·ranges from 0 to 15 and is represented within a word but is interdigitated as a long (see below)
chr is the ASCII character value
·
·
The formula that I used is as follows....
·
X = long[ \$8000 + y*4·+ chr*\$40 - (chr & 1)*\$40 ]
···
convert·interdigitated long to non-interdigitated word:
···

X_Data := 0
MaskBit := |< (chr & 1)+30
repeat 15 '(<--This may need to be 16)
X_Data := X_Data << 1 | X_Bit

X_Data returnes a non-interdigitated word value
·
·
·
·

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Beau Schwabe (mailto:bschwabe@parallax.com)

IC Layout Engineer
Parallax, Inc.

Post Edited (Beau Schwabe (Parallax)) : 10/23/2006 7:01:53 PM GMT

Cobalt
10-24-2006, 01:22 AM
Great stuff, I'm going to throw it at my Propeller as soon as I get home, thanks a million!

Cobalt
10-24-2006, 09:28 PM
Beau Schwabe said...

Y ranges from 0 to 31
X ranges from 0 to 15 and is represented within a word but is interdigitated as a long (see below)
chr is the ASCII character value

The formula that I used is as follows....
X = long[ \$8000 + y*4 + chr*\$40 - (chr & 1)*\$40 ]
convert interdigitated long to non-interdigitated word:

X_Data := 0
MaskBit := |< (chr & 1)+30
repeat 15 '(<--This may need to be 16)
X_Data := X_Data << 1 | X_Bit

X_Data returnes a non-interdigitated word value

Wow, ok, for starters I thank you for taking time to help me out on this, and I am really glad that you are trying. I have no doubt in my mind that this code works, but due to my lack of coding syntax abilities, I can honestly say that I understand about 70% of that.
I can tell that there are multiple variables that need to be defined somehow, but I'm not sure which ones.
X_Data, X_bit, MaskBit, X, and chr all ovbiously need to be delt with, and I know that 'chr' is the ASCII character I want, and that X_Data is the word value from the ROM. 'X' seems easy enough to get from the niffty little equation you provided. However, the X_bit and MaskBit both seem to just be temporary spots for the method. I want to be able to use this in its own PUB, such that I can call it as needed. I tried this, but it didn't seem to work out too well:

VAR
word X_Data
word chr
word X

PUB GetChr(chr, X) | MaskBit, X_Bit
X_Data := 0
MaskBit := |< (chr & 1)+30
repeat 15 '(<--This may need to be 16)
X_Data := X_Data << 1 | X_Bit

This way, I would just use 'chr' and 'X', and the GetChr method would modify X_Data to what I wanted. I'm not sure if this is a proper way of doing this, I'm quite certain that my code has wonderful errors!

Also, if I wanted to read only the first or second byte of the X_Data word, what would be the proper command, would "X_Data.byte[x]" work? (x = the byte)
Thank you!

Beau Schwabe
10-25-2006, 12:45 AM
Cobalt,

You don't need to define any variables outside of the PUB.· By stating the variables within the PUB in this way...

PUB GetChr(chr, Y) | MaskBit, X_Bit, X

Here is·"tested" code that is a bit more compact.

PUB GetChr(Chr, Y) | MaskBit, X_Bit,X
X := (long[\$8000 + Y*4 + Chr*64 - (Chr & 1)*64])<<(1-(Chr & 1))
Result~
repeat 16
Result >>= 1

When you call GetChr, you only need to specify the Y location (0-31) and the Character.·
To get the upper BYTE...

ByteVar := (GetChr(Chr,Y)>>8) & \$FF

To get the lower BYTE...

ByteVar := GetChr(Chr,Y) & \$FF

..In both cases, you only need to define ByteVar.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Beau Schwabe (mailto:bschwabe@parallax.com)

IC Layout Engineer
Parallax, Inc.

Post Edited (Beau Schwabe (Parallax)) : 10/24/2006 5:55:26 PM GMT

Cobalt
10-25-2006, 03:29 AM
Excellent, that makes things a lot more clear!
I'll try and post some results later tonight.
Again, thank you very much!