Post by thread:[EE] HD44780 - 2 line/1 line effects?

I recently bought this LCD from sparkfun electronics: http://www.sparkfun.com/commerce/product_info.php?products_id=709# The driver chip is compatible with the HD44780. If you set the function set for this chip to 1 line, instead of 2 line, even though it is driving a 2 line display, would it still display properly? -- Paul Anderson VE6HOP spam_OUTwackyvorlonTakeThisOuTgmail.com http://www.oldschoolhacker.com "May the electromotive force be with you."

Paul, On Tue, 25 Sep 2007 14:43:48 -0600, Paul Anderson wrote: > I recently bought this LCD from sparkfun electronics: > www.sparkfun.com/commerce/product_info.php?products_id=709# > > The driver chip is compatible with the HD44780. If you set the > function set for this chip to 1 line, instead of 2 line, even though > it is driving a 2 line display, would it still display properly? Rather depends what you mean by "Properly"! These controllers are designed to cope with much larger displays, and you'll find that line 2 doesn't follow on from line 1 in the addressing, for example, so even as you bought it you may find it doesn't behave quite how you expect. As long as you only display the first 16 positions, it should work fine (the display will be on the top line of course, not in the middle of the display). Why do you want to do that, by the way? What are you expecting to be different if you tell it it only has 1 line? Cheers, Howard Winter St.Albans, England

--- Paul Anderson <.....wackyvorlonKILLspam@spam@gmail.com> wrote: > I recently bought this LCD from sparkfun > electronics: > www.sparkfun.com/commerce/product_info.php?products_id=709# > > The driver chip is compatible with the HD44780. If > you set the > function set for this chip to 1 line, instead of 2 > line, even though > it is driving a 2 line display, would it still > display properly? You'll only get one line, and if you'd previously adjusted the contrast when it was set to 2 lines, you'll need to readjust. And I think it also reduces the instruction time, but it's been a while since I looked that deeply into the docs. Cheerful regards, Bob

> And I think it also reduces the instruction time, but it's been > a while since I looked that deeply into the docs I've never even thought to set a 16x2 as 16x1 so had a look in the 44780 datasheet to get an idea of why you might do it Might be wrong (distinct possibility, but I have the Hitachi LCD manual/big book), AFAICT there's no reason or advantage. Everything seems to be the same - execution time, scrolling, display addresses, etc. No notes about "If you select 16x1 this happens..." Is there some aesthetic reason for a backlit negative display in 16x1 rather than 16x2 ?

--- Jinx <.....joecolquittKILLspam.....clear.net.nz> wrote: > > And I think it also reduces the instruction time, > but it's been > > a while since I looked that deeply into the docs > > I've never even thought to set a 16x2 as 16x1 so had > a look > in the 44780 datasheet to get an idea of why you > might do it > > Might be wrong (distinct possibility, but I have the > Hitachi LCD > manual/big book), AFAICT there's no reason or > advantage. > > Everything seems to be the same - execution time, > scrolling, > display addresses, etc. No notes about "If you > select 16x1 this > happens..." Ahh, a challenge:) It's in typical obscure wording. Here it is, on page 249 of my book. Execution time is listed in two columns, one when in single line mode, fOSC = 250KHz, the other when in two-line mode, fOSC = 160KHz. Many of the parameters are three times faster in single line mode, whether using 5x10(1/11 duty cycle)or 5x7(1/8 duty cycle) font. Two line mode is always 5x7 with 1/16 duty cycle. Cheerful regards, Bob

> It's in typical obscure wording. Here it is, on page > 249 of my book. Execution time is listed in two > columns, one when in single line mode, fOSC = 250KHz, > the other when in two-line mode, fOSC = 160KHz. Many > of the parameters are three times faster in single > line mode, whether using 5x10(1/11 duty cycle)or > 5x7(1/8 duty cycle) font. Two line mode is always 5x7 > with 1/16 duty cycle. Well, that's interesting to know. I've looked through the Hitachi book again, trying to find some direct or implied link between lines and speed but nothing. Not for the 44100 or 66780 either. Every note or test case is cited with fosc = 250kHz, with no indication that speed is related to 1 or 2 lines. Certainly from this data you'd not expect any significant operational difference. In fact if you look at the Frame Frequency diagrams, you might conclude that there is likely no difference between 1/8 and 1/16 Duty Cycle. 1/8 takes 400 clocks and 1/16 takes 200 clocks to give each a FF of 78.1Hz

Jinx wrote: >> It's in typical obscure wording. Here it is, on page >> 249 of my book. Execution time is listed in two >> columns, one when in single line mode, fOSC = 250KHz, >> the other when in two-line mode, fOSC = 160KHz. Many >> of the parameters are three times faster in single >> line mode, whether using 5x10(1/11 duty cycle)or >> 5x7(1/8 duty cycle) font. Two line mode is always 5x7 >> with 1/16 duty cycle. > > Well, that's interesting to know. I've looked through the Hitachi > book again, trying to find some direct or implied link between > lines and speed but nothing. Not for the 44100 or 66780 either. > Every note or test case is cited with fosc = 250kHz, with no > indication that speed is related to 1 or 2 lines. Certainly from this > data you'd not expect any significant operational difference. In > fact if you look at the Frame Frequency diagrams, you might > conclude that there is likely no difference between 1/8 and 1/16 > Duty Cycle. 1/8 takes 400 clocks and 1/16 takes 200 clocks > to give each a FF of 78.1Hz > Yes, the docs are quite obscure. I have never measured the execution times to compare, but I suspect running in 2-line mode really does slow things down even though it maintains the same frame frequency. Likely the accessing of memory both for display generation and processing is the reason it takes so much longer in 2 line mode. I also wonder whether the 160KHz and 250KHz info relates to particular applications. There is some other info about the 44780 in the old display data sheets, in particular the L2570 display which is a huge document, unfortunately I only have a hard copy or I'd make it available. And of course with all the chip revisions and copies, the data sheet is only a starting point. Cheerful regards, Bob

> suspect running in 2-line mode really does slow things down even > though it maintains the same frame frequency Well, I've measured a few cases and find no difference with this particular 16x2 and a 40MHz 18F2520. Time of execution is started at 'bsf marker' and ended at 'bcf marker' 1-line write and busy periods Single character 42.91us Clear screen 1584us Change print position 24us 2-line write and busy periods Single character 42.91us Clear screen 1584us Change print position 24us Interestingly, to the real geeks, is that single-character write time is shorter after a command write. For example busy period for "A" after clearing the screen or changing print position is only 35.7us, whereas busy period for "B" after "A" takes 42.91us There's a 1ms delay between character writes (msdelay) but just 3us between command write and character write ( ie the 4 lines after ';==== code' ) So there ya go ;==== set-up mov macro litval,file ;movlw then movwf movlw litval movwf file endm disp macro litval ;display character movlw litval call write_d endm clrscrn macro ;clear LCD movlw 0x01 call write_c endm lcd_pos macro litval ;set display position movlw litval-1 call address endm ln1 = 0x00 ;line1 start address ln2 = 0x40 ;line2 start address ;==== code clrscrn disp "*" ;ready call line2 disp "2" ;line 2 off/on indicator go btfsc button ;start button bra go then either clrscrn call msdelay clrscrn call msdelay clrscrn call msdelay or disp "A" call msdelay disp "B" call msdelay lcd_pos ln1+.9 disp "C" call msdelay line1 movlw 0x00 ;line 1, column 0 call address return line2 movlw 0x40 ;line 2, column 0 call address return address addlw 0x80 ;set high bit of address command call write_c return write_c bcf rs ;write command bra d_out write_d bsf rs ;write data usec d_out bsf marker movwf latb ;write to port usec bcf rw usec bsf en usec bcf en usec bsf rw ;set up for Busy Flag read usec bcf rs usec clrf portb usec mov b'10000000',trisb bsf en usec getbusy btfsc busy ;loop until Busy Flag clear bra getbusy bcf marker bcf en usec bcf rw mov b'00000000',trisb return

On 9/26/07, Alan B. Pearce <EraseMEA.B.Pearcespam_OUTTakeThisOuTrl.ac.uk> wrote: > >Is there some aesthetic reason for a backlit negative display > >in 16x1 rather than 16x2 ? > > He has a 16x2 while doing development - simply because he already has it, > but the end product will use only 16x1 ?? > > It's actually bit sillier than that. The end product will be using a PIC16F88 microcontroller to drive the LCD, but I wanted to get a feel for driving the LCD on an Arduino board before moving to the F88. I've wired up the LCD, and I'm using code from here: http://www.arduino.cc/en/Tutorial/LCD8Bits The backlight is on, I added a line of code to turn on a status LED so I know it's up and running, but nothing. I've doublechecked wiring and don't see anything wrong with it so far. There are two possible issues. First, is in the code communicating to the LCD. That may have a typo or something and be sending the wrong signals(I notice the datasheet SparkFun gives for the module likes to list instruction as big endian, while the arduino code is talking small endian). Second, the contrast. I haven't a potentiometer handy, so I've tried three configurations: Pulled low, pulled high, and on a resistive voltage divider that should land it about half of the 5v supply. I'm thinking either the code is messed up or I have the contrast wired totally wrong. Any thoughts? -- Paul Anderson VE6HOP wackyvorlonspam_OUTgmail.com http://www.oldschoolhacker.com "May the electromotive force be with you."