AVR Freaks

As such I'm going to lock this thread. If anyone has anything to add to the original article PM js, plons or clawson and ask us to temporarily unlock this.

Moderator

Thanks buddy....it helped me a lot!

Hello,

 

Sorry for bringing back this old thread. My main question is more about C/C++ programming.

 

Original AVR Butterfly LCD has 16 bits for one digit, but my LCD has 8 bits.

 

Original LCD_WriteChar method:

 

/*
 NAME:      | LCD_WriteChar (static, inline)
 PURPOSE:   | Routine to write a character to the correct LCD registers for display
 ARGUMENTS: | Character to display, LCD character number to display character on
 RETURNS:   | None
*/
static inline void LCD_WriteChar(const uint8_t Byte, const uint8_t Digit)
{
	uint8_t* BuffPtr = (uint8_t*)(LCD_LCDREGS_START + (Digit >> 1));
	uint16_t SegData = 0x0000;

	if (Byte != LCD_SPACE_OR_INVALID_CHAR)              // Null indicates invalid character or space
	  SegData = pgm_read_word(&LCD_SegTable[Byte]);	

	for (uint8_t BNib = 0; BNib < 4; BNib++)
	{
		uint8_t MaskedSegData = (SegData & 0x0000F);

		if (Digit & 0x01)
		  *BuffPtr = ((*BuffPtr & 0x0F) | (MaskedSegData << 4));
		else
		  *BuffPtr = ((*BuffPtr & 0xF0) | MaskedSegData);

		BuffPtr += 5;
		SegData >>= 4;
	}	
}

 

If I change this line:

*BuffPtr = ((*BuffPtr & 0x0F) | (MaskedSegData << 4));

to this line:

*BuffPtr = ((*BuffPtr & 0x0F) | (MaskedSegData << 2));

I can "move cursor" to the second digit fine, but it doesn't move to the third digit.

 

So my main questions are about this particular LCD_WriteChar method from C/C++ perspective:

1) what means SegData & 0x0000F in this line? Why 0x0000F instead of 0x0F?

uint8_t MaskedSegData = (SegData & 0x0000F);

2) what means Digit & 0x01 in this line?

if (Digit & 0x01)

3) what means this line?

*BuffPtr = ((*BuffPtr & 0x0F) | (MaskedSegData << 4));

 

Regards,

Furieux

Hi Clawson,

Yes, sorry, I’m actually talking about 329.

Ok, could you please help me to analyze this function deeper?

 

uint8_t* BuffPtr = (uint8_t*)(LCD_LCDREGS_START + (Digit >> 1));

Here BuffPtr is defined which corresponds to LCD registers starting with LCDDR0. Next would be LCDDR5, then LCDDR10, and then LCDDR15 (according to COM0, COM1, COM2, and COM3). This is clear for me. But again - what Digit >> 1 means here?

 

Then,

uint16_t SegData = 0x0000;

Here SegData is declared with initial value.

 

Then,

if (Byte != LCD_SPACE_OR_INVALID_CHAR)
    SegData = pgm_read_word(&LCD_SegTable[Byte]);

Here it's checking if Byte is not invalid character and if it is not, then avr function pgm_read_word() is parsing LCD_SegTable[] array for the Byte value and then store it to the SegData variable.

For example, let's take number 0 to be shown on my segment LCD which in my case in array LCD_SegTable[] is defined as 0x3132

 

Then,

for (uint8_t BNib = 0; BNib < 4; BNib++) {

Here it is looping through nibbles (4 bits). I think this is because AVR Butterfly LCD has 16 bits for the digit. But in my case I have 8 bits, so I thought I need to loop only through two nibbles? I tried to change here BNib < 2, but then wrong segments are lit.

 

Then,

uint8_t MaskedSegData = (SegData & 0x0000F);

Here MaskedSegData is defined and it takes SegData value which was read from LCD_SegTable[] array (in this example it is 0x3132) and then it is AND-ed with bottom 4 bits. Right?

If I convert 0x3132 from hex to bin, i get

0000 0000 0011 0001 0011 0010

AND-ed with 1111 it is then

0000 0000 0011 0001 0011 0010 right? The same!

Why it is AND-ed in the first place? You wrote that it is checking for even/odd numbers. Why it is needed here?

 

Then,

if (Digit & 0x01)
    *BuffPtr = ((*BuffPtr & 0x0F) | (MaskedSegData << 4));

This I cannot understand at all. It's AND-ing Digit with 1 (if the last bit is 1?) and if true, then BuffPtr is AND-ed with bottom 4 bits and then OR-ed... what that means? It's definitely related to LCDDRx, but I can't get the point.

 

Then,

else
    *BuffPtr = ((*BuffPtr & 0xF0) | MaskedSegData);

So if AND-ing Digit with 1 is false (the last bit is 0?), then BuffPtr is AND-ed with top 4 bits and then again OR-ed... what that means?

 

And finally,

BuffPtr += 5;
SegData >>= 4;

BuffPtr now is incremented by 5 (this probably is related to those LCDDR0, LCDDR5, LCDDR10 and so on, or LCDDR1, LCDDR6, LCDDR11 and so on). This probably is needed because next for loop iteration will be on the next register. Right?

And SegData? This I don't understand at all. What SegData = SegData >> 4 means in the first place? I know that shifting to the right is basically dividing, but I'm not sure if it's that case here.

+10

 

Exactly what I was going to say. You can't really hope to use a code like this until you understand it and to understand it you need a grasp of some of the basic concepts in C like shifting and masking.