Post by thread:Hitachi LCD module

Hi All, I am trying to get an Hitachi LM016L LCD module working with a PIC 84, with no success. I have had the module working with a STAMP, and I understand everything I need to do to setup and use the module, but it just ain't workin'. I am using PORTA <0:3> for the LCD data port (connected to the MSB of the LCD data pins), and PORTB <4:6> for E, RS and R/^W. Obviously I am attempting to initialise the module for 4 bit operation. My question is: am I likely to have problems with PORT A, and should I stick everything on PORT B? This is a simple problem, I just need reassurance to keep trying. Andy (the other one) Lancaster, UK

Hi Andy (too), > I am trying to get an Hitachi LM016L LCD module working with a PIC 84, > with no success. I have had the module working with a STAMP, and I > understand everything I need to do to setup and use the module, but it > just ain't workin'. > > I am using PORTA <0:3> for the LCD data port (connected to the MSB of > the LCD data pins), and PORTB <4:6> for E, RS and R/^W. Obviously I am > attempting to initialise the module for 4 bit operation. My question > is: am I likely to have problems with PORT A, and should I stick > everything on PORT B? Not having used that direct configuration with the LCD I am speaking from general experience with the LCD and C84 devices. If it worked with the Stamp then your problem is probably timing related. Put pauses of 2 ms after mode, clear and home commands and 100 us after ather commands. Also send the enable pulse to the display with 3 instructions (lo-hi-lo) with possibly a short delay after the data and address times have settled, the PIC is FAST and the LCD is SLOW. > This is a simple problem, I just need reassurance to keep trying. If nothing helps, start the wiring and code from scratch and you will probably notice your mistake :-) Cheers -- Kalle Pihlajasaari spam_OUTkalleTakeThisOuTdata.co.za Interface Products Box 15775, Doornfontein, 2028, South Africa +27 (11) 402-7750 Fax: +27 (11) 402-7751

Really like what you write after your name (the other one). Rogerio At 10:33 AM 23/07/96 +0100, you wrote: {Quote hidden}>Hi All, > >I am trying to get an Hitachi LM016L LCD module working with a PIC 84, >with no success. I have had the module working with a STAMP, and I >understand everything I need to do to setup and use the module, but it >just ain't workin'. > >I am using PORTA <0:3> for the LCD data port (connected to the MSB of >the LCD data pins), and PORTB <4:6> for E, RS and R/^W. Obviously I am >attempting to initialise the module for 4 bit operation. My question >is: am I likely to have problems with PORT A, and should I stick >everything on PORT B? > >This is a simple problem, I just need reassurance to keep trying. > >Andy (the other one) >Lancaster, UK > >

I wonder if your problem is port A. Pin RA4 is an open collector output and may float if the LCD module doesn't pull it up. This could cause problems when doing the initialization. I usually use port A as the control lines and port B as the data. If you are using the Microchip app notes, there are a few errors. I have an example to control the LCD from a 16C84 but I can't find the web address at the moment. E-mail me if you are interested in the code. Norm At 10:33 AM 7/23/96 +0100, you wrote: {Quote hidden}>Hi All, > >I am trying to get an Hitachi LM016L LCD module working with a PIC 84, >with no success. I have had the module working with a STAMP, and I >understand everything I need to do to setup and use the module, but it >just ain't workin'. > >I am using PORTA <0:3> for the LCD data port (connected to the MSB of >the LCD data pins), and PORTB <4:6> for E, RS and R/^W. Obviously I am >attempting to initialise the module for 4 bit operation. My question >is: am I likely to have problems with PORT A, and should I stick >everything on PORT B? > >This is a simple problem, I just need reassurance to keep trying. > >Andy (the other one) >Lancaster, UK > >

Andy, > I am trying to get an Hitachi LM016L LCD module working with a PIC 84, > with no success. I have had the module working with a STAMP, and I > understand everything I need to do to setup and use the module, but it > just ain't workin'. > > I am using PORTA <0:3> for the LCD data port (connected to the MSB of > the LCD data pins), and PORTB <4:6> for E, RS and R/^W. Obviously I am > attempting to initialise the module for 4 bit operation. My question > is: am I likely to have problems with PORT A, and should I stick > everything on PORT B? I suspect that several others are right about timing. The following code is from a standarized Embedded C library that I have been using for a few years. I copied the following listing fragment from a project I did for Circuit Cellar. It was compiled with MPC on a 16C74. The LCD display was connected to Port D in this case. To save an I/O pin I generally connect the R/^W to ground. Walter Banks Byte Craft #include <lcd4.c> /******************************************************** * Device driver for 4 bit LCD interface using * * Hitachi H2570 controllers. These displays are in * * configurations from one line by 16 characters to * * 4 lines by 40 characters * * * * No warranty is implied or given as to * * their usability for any purpose. * * * * (c) Copyright 1990 * * Byte Craft Limited * * Waterloo, Ontario * * Canada N2J 4E4 * * (519) 888-6911 * * * * Walter Banks * ********************************************************/ 0000 registerx index; #pragma option p ; /* --------------- | | | Pn0 |--------------------------\ Pn1 |--------------------------+\ Pn2 |--------------------------++\ Pn3 |--------------------------+++\ E Clock Pn4 |-------------------\ |||| Register Pn5 |-------------------+-\ |||| Select Pn6 |- | | |||| Pn7 |- | | |||| | | | |||| | | | |||| | | | |||| | Vcc | | |||| --------------- | | | |||| |---- | | |||| | | | | |||| - | | | |||| 20K | <---+ | | |||| pot | ||| | |||| ----------+++ | |||| | 1 |||||||||||||| 14 | ---------------------------- --- | .............. | - | Liquid Crystal Display | | | ---------------------------- */ 0005 #define LCDdata PORTD /* Data output */ 0006 #define LCD_RS PORTD.5 /* Register select bit */ 0007 #define LCD_E PORTD.4 /* Clock bit */ /******************************************************** * wLCDdat - Write data word to LCD peripheral * * Enter with data word in accumulator * * Delay w120fS after data write * ********************************************************/ void wLCDdat(unsigned int tempa) 0035 { 001B 1283 BCF STATUS,RP0 001C 00B5 MOVWF 35 /* Write HI nybble word to LCD */ 001D 00A1 MOVWF 21 LCDdata = tempa >> 4; 001E 0EA1 SWAPF 21 001F 300F MOVLW 0Fh 0020 0521 ANDWF 21,W 0021 0088 MOVWF PORTD 0022 1688 BSF PORTD,5 LCD_RS = 1 ; /* RS->1 */ 0023 1608 BSF PORTD,4 LCD_E = 1 ; /* E->1 */ 0024 1208 BCF PORTD,4 LCD_E = 0 ; /* E->0 */ 0025 1288 BCF PORTD,5 LCD_RS = 0 ; /* RS->0 */ 0026 300F MOVLW 0Fh LCDdata = tempa & 0x0f; 0027 0535 ANDWF 35,W /* Write LO nybble word to LCD */ 0028 0088 MOVWF PORTD 0029 1688 BSF PORTD,5 LCD_RS = 1 ; /* RS->1 */ 002A 1608 BSF PORTD,4 LCD_E = 1 ; /* E->1 */ 002B 1208 BCF PORTD,4 LCD_E = 0 ; /* E->0 */ 002C 1288 BCF PORTD,5 LCD_RS = 0 ; /* RS->0 */ 002D 3014 MOVLW 14h index = 20; 002E 0084 MOVWF FSR 002F 0384 DECF FSR while (--index); /* 120usec $$$$*/ 0030 0804 MOVF FSR,W 0031 1903 BTFSC STATUS,Z 0032 2834 GOTO 0034h 0033 282F GOTO 002Fh 0034 0008 RETURN } /********************************************************* * wLCDctrl4 - Write control word to LCD peripheral * * configured as a 4 bit interface * * Enter with control word in accumulator * * Delay w4.5mS if A=0x01 or 0x02 else delay w120fS * *********************************************************/ void wLCDctrl4(unsigned int tempa) 0036 { 0035 1283 BCF STATUS,RP0 0036 00B6 MOVWF 36 /* Write HI nybble to LCD */ 0037 00A1 MOVWF 21 LCDdata = tempa >> 4; 0038 0EA1 SWAPF 21 0039 300F MOVLW 0Fh 003A 0521 ANDWF 21,W 003B 0088 MOVWF PORTD 003C 1608 BSF PORTD,4 LCD_E = 1 ; /* E->1 */ 003D 1208 BCF PORTD,4 LCD_E = 0 ; /* E->0 */ 003E 300F MOVLW 0Fh LCDdata = tempa & 0x0f; /* Write LO nybble to LCD */ 003F 0536 ANDWF 36,W 0040 0088 MOVWF PORTD 0041 1608 BSF PORTD,4 LCD_E = 1 ; /* E->1 */ 0042 1208 BCF PORTD,4 LCD_E = 0 ; /* E->0 */ 0043 3014 MOVLW 14h index = 20; 0044 0084 MOVWF FSR 0045 0384 DECF FSR while (--index); /* 120usec */ 0046 0804 MOVF FSR,W 0047 1903 BTFSC STATUS,Z 0048 284A GOTO 004Ah 0049 2845 GOTO 0045h 004A 3002 MOVLW 02h if (tempa <= 2) 004B 1283 BCF STATUS,RP0 004C 0236 SUBWF 36,W 004D 1903 BTFSC STATUS,Z 004E 2851 GOTO 0051h 004F 1803 BTFSC STATUS,C 0050 2860 GOTO 0060h 0051 { 0051 0384 DECF FSR while (--index); 0052 0804 MOVF FSR,W /* 4.6 ms or so delay */ 0053 1903 BTFSC STATUS,Z 0054 2856 GOTO 0056h 0055 2851 GOTO 0051h 0056 0384 DECF FSR while (--index); 0057 0804 MOVF FSR,W 0058 1903 BTFSC STATUS,Z 0059 285B GOTO 005Bh 005A 2856 GOTO 0056h 005B 0384 DECF FSR while (--index); 005C 0804 MOVF FSR,W 005D 1903 BTFSC STATUS,Z 005E 2860 GOTO 0060h 005F 285B GOTO 005Bh } 0060 0008 RETURN } void initLCD(void) { /* LCD display peripheral initialization */ 0061 3024 MOVLW 24h wLCDctrl4(0x24); 0062 2035 CALL 0035h /* Function Set-4-bit,1-line,5X7 */ 0063 300C MOVLW 0Ch wLCDctrl4(0x0c); 0064 2035 CALL 0035h /* Display on, Cursor off */ 0065 3006 MOVLW 06h wLCDctrl4(0x06); 0066 2035 CALL 0035h /* entry mode- Inc addr, no shift */ 0067 3001 MOVLW 01h wLCDctrl4(0x01); 0068 2035 CALL 0035h /* Clear Display */ 0069 0008 RETURN } void putcharLCD(char sci_char) /* Send a character to the LCD display */ 0037 { 006A 1283 BCF STATUS,RP0 006B 00B7 MOVWF 37 006C 300A MOVLW 0Ah if ((sci_char == 0x0a) || 006D 0237 SUBWF 37,W (sci_char == 0x0d)) /* Look for a LF */ 006E 1903 BTFSC STATUS,Z 006F 2874 GOTO 0074h 0070 300D MOVLW 0Dh 0071 0237 SUBWF 37,W 0072 1D03 BTFSS STATUS,Z 0073 2877 GOTO 0077h 0074 3001 MOVLW 01h wLCDctrl4(0x01); /* Clear Display */ 0075 2035 CALL 0035h 0076 287A GOTO 007Ah else 0077 1283 BCF STATUS,RP0 wLCDdat(sci_char); 0078 0837 MOVF 37,W 0079 201B CALL 001Bh 007A 0008 RETURN }

Andy Errington <.....PICLISTKILLspam@spam@MITVMA.MIT.EDU> wrote: > I am using PORTA <0:3> for the LCD data port (connected to the MSB > of the LCD data pins), and PORTB <4:6> for E, RS and R/^W. > Obviously I am attempting to initialise the module for 4 bit > operation. My question is: am I likely to have problems with PORT > A, and should I stick everything on PORT B? > > This is a simple problem, I just need reassurance to keep trying. > > Andy (the other one) Other Andy: You should be fine with PORTA... Especially since you're not using RA4. Your problem is undoubtedly timing-related; you probably need to slow your initialization routine WAY down, and you should be using the LCD's "BUSY" flag for all other operations. -Andy Andrew Warren - fastfwdKILLspamix.netcom.com Fast Forward Engineering, Vista, California http://www.geocities.com/SiliconValley/2499

Hi, Many thanks to all who have responded. I hope the information is useful to other PICLIST readers. Firstly can I thank all of those who advised me about the Murata 4MHz ceramic resonator, it's working fine, and you _can_ connect the middle pin to +5V to simplify board layout. Secondly the LCD module... I tried my module on the Stamp again, which showed me it was still working properly, and I haven't damaged it (relief!). Back to the '84. One problem I think I have is reading the busy flag, so I replaced that section with a delay loop. Now I get some life from the display. I am using portions of Microchips LCD application note, so my code sends 'M' 'i' 'c' 'r' 'o' 'c' 'h' 'i' 'p' to the 'Send_char' routine. Deep joy! On the display is some text! Hang on, what does it say? 'Mogwognow' I have worked out that the middle 2 bits of the lower nibble seem to be always set to one, except for the first character (!). Could it be that my SWAPF function is broken inside the CPU? I don't know, I shall follow people's suggestions and slow it all down some more. More news as it's made. To anyone else who is using an LCD module with no success: Take heart, and keep trying! Andy (the other one)

At 10:47 96/07/24 +0100, you wrote: <<snip>> {Quote hidden}>Secondly the LCD module... > >I tried my module on the Stamp again, which showed me it was still >working properly, and I haven't damaged it (relief!). Back to the '84. >One problem I think I have is reading the busy flag, so I replaced that >section with a delay loop. Now I get some life from the display. I am >using portions of Microchips LCD application note, so my code sends 'M' >'i' 'c' 'r' 'o' 'c' 'h' 'i' 'p' to the 'Send_char' routine. Deep joy! >On the display is some text! Hang on, what does it say? 'Mogwognow' > >I have worked out that the middle 2 bits of the lower nibble seem to be >always set to one, except for the first character (!). Could it be that >my SWAPF function is broken inside the CPU? I don't know, I shall >follow people's suggestions and slow it all down some more. More news >as it's made. > >To anyone else who is using an LCD module with no success: Take heart, >and keep trying! > >Andy (the other one) (Parts of the text are a reprint of some postings in februari 1996 by Scott Dattalo and myself) Assuming you are using AN587 to write your code: This app.note has a bug in the code to *read from* any HD44780 based LCD module (this affects ALL read operations). A quote of Scott's comment (4-bit LCD interface): The proper way to perform a read operation is to: 1) Clear RS (Register Select signal on LCD Module). 2) Set R/W (Read/Write signal on LCD Module). This will select the "read" mode. 3) Set E (The enable bit on the LCD Module). 4) Read the 4 MSB's of the data bus on the LCD Module. 5) Clear E. 6) Set E. 7) Again, read 4 MSB's. 8) Clear E. The two nibbles are then concatenated to form a byte. If you're reading the busy bit, then bit 7 will reflect its state while bits 6 through 0 are either the moste recent character or data RAM address. The way AN587 does a read operation is: 1) Clear RS (Register Select signal on LCD Module). 2) Set R/W (Read/Write signal on LCD Module). This will select the "read" mode. 3) Set E (The enable bit on the LCD Module). 4) Clear E. 5) Read the 4 MSB's of the data bus on the LCD Module. 6) Set E. 7) Clear E. 8) Again, read 4 MSB's. In other words, AN587 reverses steps 4 and 5 and steps 7 and 8. The effect is that the LCD module gets disabled before the data is read by the PIC. (end quote) This bug is in both the 4-bit and 8-bit code of AN587! A correct version of an 8-bit interface (using the 16C84 and based on AN587), the schematics and an example (see 4.1.4. Availability) can be found at: http://www.iaehv.nl/users/pouweha/lcd2.htm The code on this page makes use the busy flag successfully. You only need delay loops during LCD initialisation. Happy coding... Peer. <><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><> Peer Ouwehand .....pouwehaKILLspam.....iaehv.nl http://www.iaehv.nl/users/pouweha/ Welcome my son, welcome to the machine. (Pink Floyd) ><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><

I had similar problems when interfacing an LCD to an 8051 controller. Make sure you are reading/writing data on the correct clock edge and pay close attention to the chips timing req'ts and the timing you are generating. Good luck. Do you know a good source for cheap 2x16 line LCD displays? Send reply to 'EraseMErmudryspam_OUTTakeThisOuTbnr.ca' (I'm not on the list) Thanks, Rob In message "Re: Hitachi LCD module", you write: {Quote hidden}>Hi, > >Many thanks to all who have responded. I hope the information is >useful to other PICLIST readers. > >Firstly can I thank all of those who advised me about the Murata 4MHz >ceramic resonator, it's working fine, and you _can_ connect the middle >pin to +5V to simplify board layout. > >Secondly the LCD module... > >I tried my module on the Stamp again, which showed me it was still >working properly, and I haven't damaged it (relief!). Back to the '84. >One problem I think I have is reading the busy flag, so I replaced that >section with a delay loop. Now I get some life from the display. I am >using portions of Microchips LCD application note, so my code sends 'M' >'i' 'c' 'r' 'o' 'c' 'h' 'i' 'p' to the 'Send_char' routine. Deep joy! >On the display is some text! Hang on, what does it say? 'Mogwognow' > >I have worked out that the middle 2 bits of the lower nibble seem to be >always set to one, except for the first character (!). Could it be that >my SWAPF function is broken inside the CPU? I don't know, I shall >follow people's suggestions and slow it all down some more. More news >as it's made. > >To anyone else who is using an LCD module with no success: Take heart, >and keep trying! > >Andy (the other one)

Hi all, well I have the LCD module working on the PIC, using delays between characters rather than reading the busy flag. Thanks to Peer Ouwehand for pointing out the error of trying to read the LCD data with pin E low (which tristates the data bus!). I have looked through Hitachi's LCD documentation, and this fact is not at all obvious. However I still cannot successfully check for the busy flag. Most of the samples I have seen simply tie R/^W low (always write), and use delays only. This means you save one I/O pin, which might be satisfactory. I shall have another go at reading from the LCD when I go home tonight (yes I _do_ have a home to go to). Meanwhile I have successfully read and decoded data from a magnetic card (a credit card in this case), and displayed it on my LCD. I currently ignore parity on each character, and the final parity check, but this doesn't matter for the purposes of experimentation. Later I shall check parity, and then I shall attempt to decode the data with the card going through backwards. I hope all this drivel is of interest to someone. It is to me. Many thanks to all correspondents so far. Andy (the other one) >shameless plug< PS I shall be seeking employment in the Yorkshire area (north of England for our overseas readers) from January '97. My main skills are client/server development with VB, SQL Server and NT. As you can see I dabble with low level assembly, and I also have experience in access control and security. -- --------------------------------------------------------------------- Andrew M Errington Tel: +44 1524 593678 Microcomputer Consultant Fax: +44 1524 844011 The Computer Centre Mobile (Orange): 0976 243931 Lancaster University a.erringtonspam_OUTlancaster.ac.uk Lancaster LA1 4YW www.lancs.ac.uk/people/cpaame/cpaame.htm --------------------------------------------------------------------- "A dog is not just for Christmas, there may be some left for sandwiches on Boxing Day" - Vladimir Illich Ulyanov 1920

Andy Errington <@spam@PICLISTKILLspamMITVMA.MIT.EDU> wrote: > well I have the LCD module working on the PIC, using delays > between characters rather than reading the busy flag. .... I still > cannot successfully check for the busy flag. Most of the samples I > have seen simply tie R/^W low (always write), and use delays only. > This means you save one I/O pin, which might be satisfactory. Other Andy: It also means that your LCD routines will be slower than necessary, and it prevents you from reading the display (which can be very useful with small microcontrollers like the PIC, since it relieves you of the burden of storing a copy of the displayed data in the microcontroller's RAM). It also prevents you from using the display's otherwise-unused RAM as slow-but-plentiful external RAM for the PIC. You haven't given enough detail for me to really diagnose your particular problem. Make sure, though, that you're reading all EIGHT bits of data from the display... With your 4-bit interface, you'll need to do TWO reads to get all the data. Here's some sample code; it's written with the assumption that the 4 LCD data lines are tied to PORTD 4-7; SETPDI is equated to 1111xxxx (binary) and SETPDO is equated to 0000xxxx (binary). "PAGE0" and "PAGE1" are simple macros that set/clear the register-page-select bit: ; ; WAIT FOR THE LCD'S "BUSY" FLAG TO CLEAR. ; LCDREADY: PAGE1 ;SWITCH TO DATA-SEGMENT 1. MOVLW SETPDI ;MAKE THE LCD DATA BUS ALL MOVWF TRISD ^ 080H ;INPUTS. PAGE0 ;SWITCH BACK TO DATA-SEGMENT 0. LCDWTLP: BCF LCDRS ;SETUP TO READ THE BUSY FLAG. BSF LCDRW ; BSF LCDE ;CLOCK THE DATA IN. NOP ; MOVF PORTD,W ;GRAB THE HI-NIBBLE OF DATA ;(BIT 7 = BUSY). BCF LCDE ;FINISH CLOCKING. NOP ;WAIT A CYCLE. BSF LCDE ;CLOCK THE LO-NIBBLE OUT (AND NOP ;IGNORE IT). BCF LCDE ; ANDLW 10000000B ;MASK OFF ALL BUT THE BUSY FLAG. BNZ LCDWTLP ;IF THE LCD'S STILL BUSY, LOOP ;BACK AND TRY AGAIN. MOVLW 00001111B ;OTHERWISE, CLEAR THE DATA BUS. ANDWF PORTD PAGE1 ;SWITCH TO DATA-SEGMENT 1. MOVLW SETPDO ;MAKE THE LCD DATA BUS ALL MOVWF TRISD ^ 080H ;OUTPUTS. PAGE0 ;SWITCH BACK TO DATA-SEGMENT 0. RETURN ;RETURN. For maximum efficiency, I generally call this routine just BEFORE performing any operation on the LCD, rather than after. Hope this helps... -Andy Andrew Warren - KILLspamfastfwdKILLspamix.netcom.com Fast Forward Engineering, Vista, California http://www.geocities.com/SiliconValley/2499

On Wed, 24 Jul 1996, Andy Errington wrote: snip > section with a delay loop. Now I get some life from the display. I am > using portions of Microchips LCD application note, so my code sends 'M' > 'i' 'c' 'r' 'o' 'c' 'h' 'i' 'p' to the 'Send_char' routine. Deep joy! > On the display is some text! Hang on, what does it say? 'Mogwognow' > I recognice this ! Are you using 4bit mode ? When i used the Microchipcode i discovered that they had forgotten one of the initial delays so that my LCD was run in 8bit mode instead. Look for all the LCD FAQ's awailable. --- Sten Dahlgren CelsiusTech Systems ! "I'd rather have 39 Hp under my arm S-175 88 Jaerfaella Sweden ! than one under my backside" RemoveMEsedaTakeThisOuTcelsiustech.se +46-8 58084430 ! join your nearest karting club now !!