4-line LCD Help - using PortA instead of B

[Tom Gonser]

Interestingly, I see that the Crystalfontz displays require a pretty substantial intitialization routine. I can't tell if that is the same as what PBP send as standard or not.

I downloaded a demo app, and it sets the bits E,RS,etc. high and low for about 10 seconds in a pattern that looks like it must be matching the tech notes for their displays. It takes the PC parallell port driver about 10 seconds to send all the bit settings, so I'd guess it is much more than 'standard'..

The Crystlfontz guys don't seem to be able to provide details about exactly WHAT bits need to be sent and in what order to initialize the LCD.

I think my time is better spent switching to an LCD that we KNOW works with PBP, which is too bad because CrystalFontz LCDs really look nice.

Thanks as ususal for all the help folks!

TG

[mister_e]

can you try something for me and post your results?

' place your LCD define here and all the others

pause 2000
FLAGS=0
pause 500
LCDOUT $fe,1
pause 500

start:
     LCDOUT $FE,1,"Hello"
     Here: goto Here

[Tom Gonser]

Thanks - I ran the following:

' VERY SIMPLE LCD TEST

OSCCON=%01111000
DEFINE OSC 8

Include "modedefs.bas" ' Include serial modes
@ __CONFIG _CONFIG1H, _OSC_INTIO67_1H
@ __CONFIG _CONFIG2H, _WDT_ON_2H & _WDTPS_128_2H
@ __CONFIG _CONFIG3H, _PBADEN_OFF_3H & _MCLRE_OFF_3H
@ __CONFIG _CONFIG4L, _LVP_OFF_4L & 0bfh ;_XINST_OFF_4L

ADCON1 = 7

TrisA = %00000000
PortA = %00000000

DEFINE LCD_DREG PORTA ' Set data pin of LCD to
DEFINE LCD_DBIT 0 ' PORTA.0-PORTA.3

DEFINE LCD_RSREG PORTA ' Set RS bit of LCD to
DEFINE LCD_RSBIT 5 ' PORTA.5

DEFINE LCD_EREG PORTA ' Set E bit of LCD to
DEFINE LCD_EBIT 4 ' PORTA.4

DEFINE LCD_LINES 4 ' 4 Lines LCD


pause 2000
FLAGS=0
pause 500
LCDOUT $fe,1
pause 500

start:
LCDOUT $FE,1,"Hello"
Here: goto Here

END

-- No response from the LCD. I did use an LED to see if the lines were getting data - testing E, RS, and the 4 data lines, there is activity going on, so the PIC is sending data. I have to assume that the Crystalfontz LCD needs some sort of more advanced setup routine be sent to it.

TG

[NavMicroSystems]

Tom,

I guess you have double and triple checked you wiring.

if so
there must be something special with CrystalFontz.
I use a lot of LCDs and have tried a lot of brands (except CrystalFontz), without any problems at all.

Has anyone on the forum ever used a CrystalFontz - CFAH2004L-YYB-JP with a PIC ?

[Archilochus]

I just took a peek at the Hitachi 44780 LCD controller data sheet.
The initialization routine shown in the Crystalfontz part is the same - taken word-for-word from the Hitachi 44780 specs. So PBP should be initializing it properly.

Arch

[NavMicroSystems]

Arch,

that's what I thought, as it has an embedded 44780.

Assuming Tom has got his hardware wired up properly I don't know how to slove his problem.

[Tom Gonser]

The Mfg sent along this note:


--------------------------------------------------------------------------------
quote: "All I see are row 1 and 3 with shading, rows 2 and 4 blank, and no text."
--------------------------------------------------------------------------------

This is a sign that the lcd controller is not initialized. A typical init sequence is:

0x38 [wait 4mS] 0x38 [wait 100uS] 0x38 0x38 0x08 0x01 0x06 0x0C

just like the data sheet shows. Apparently LCDOUT is not performing the init. I think it has an internal flag to keep track of the init, and it probably thinks it doesn't need to do it again. I'm not really familiar with that code.
__________________

Does anyone know how to 'manually send' those commands? What is 0x38, etc?

Tom

[NavMicroSystems]

Tom,

You definitely don't have to worry about "initialisation" with PBP and a 44780 (or compatible),
PBP does it for you.

you have either got a bug in your wiring or your LCD is "dead" (or even both)

[bbarney]

I have to agree with Ralph,you have a short somewhere or bad connection.

[Tom Gonser]

OK, folks. Thanks for the help. I am going to rewire this to another LCD and see what happens. I agree something must have gotten messed up or fried, it just is not that complicated.

Tom