Compiler problems with PIC18F45k42

[Anobium]

@Guiseppe Straight question. So, I do not waste my time.

Are you aiming to enable the Slave for this micrcontrolller? Why am I asking? I have in my IDE a AN734 library for the K42s with call backs to support incoming and outgoing slave communications. But, if you are wanting to complete without this library I would put this on bottom of the list of things to do.

Let me know.

[Anobium]

@Guiseppe. I have opened up another ticket with Microchip. I cannot get the Slave to operate, at all, even when using MPLAB-X 4.01 I used code provided by Microchip and I get no response. If you get things working that would be great.

My ticket has a couple asks.
1. Example/working code for slave responding to MCP2221 MCP/Bus discovery tool
2. Example/working code for write operation using MCP2221 MCP/Bus tool
3. Example/working code for read operation using MCP2221 MCP/Bus tool

[Anobium]

Bit of an updated... Using Great Cow BASIC....

A response..... for the initialisation of

'Now set up the I2C slave library - three commands needed
HI2CSetAddress 0x40
HI2CMode Slave
HI2CSlave_ISR_Init

I got this from the MCP2221. Which is a result!

> Bus scan started. 8 bit address used. Start address = 0x0, End address = 0x80
Slave found. 8 bit address = 0x40; 7 bit address = 0x20 Write
Bus scan finished

So, far. I have updated the chip.dat file (this will need to be done to all K42 chips), updated HWi2c.h library to add the slave support, revised the exiting two slave libraries to support KMode.

I can get the intrerface to respond once, only once. Then, I have to resprogram. Somehting is not right.

[Giuseppe D'Elia]

@Anobium and William ,

I found my old LCD (LCD03 - I2C/Serial LCD, see at http://www.robot-electronics.co.uk/htm/Lcd03tech.htm) and tested the I2C + Serial test program William sent in a previous post.

The serial sending works fine.

Also the I2C sending I believe works fine. However my LCD is not compatible with the instructions #define LCD_IO 10 or #define LCD_IO 12.
Accordingly, the words shown on the LCD are not the ones expected as shown on the terminal and some special character also is shown.

I tested also the code 310_IC2_LCD_Adapter_using_HardwareI2C_18f25k42 available for 18F25K42 in the ChipRange Demonstrations installation directory.

The behaviour is the same as for William code. The LCD changes according to the code sequence although the characters are not correct. Again I believe this is due to the fact that my LCD is not compatible with the code.

As a conclusion, I'm quite sure that the codes I tested works on 18F45k42 although I have not a full demonstration.

I thank you for the program showing the I2C registers you sent.
I think it will be very useful.

In the meantime after a lot of attempts I got a code able to manage the PIC as a slave. Obviously it is not complete and it is not really clear to me why it works and how stable it is.

It is only a first step. As the examples you sent me show, I think I must manage the interrupts in a more effective way by using GCBASIC oninterrupt commands.

I attach the code (there are 3 versions of the subroutine sertxd I use for printing on terminal. The Master send two byte each time and I want to read six bytes) :

#chip 18F45k42, 64
#option explicit
#config MCLRE = ON
#startup InitPPS, 85 '//Call InitPPS
'_____________
'   I2C
SET ODCONC.ODCC4 on
SET ODCONC.ODCC3 on
SET RC3I2C.TH0   OFF   'Port specific controls may be required - see the datasheet
SET RC4I2C.TH0   OFF   'Port specific controls may be required - see the datasheet
'Set pin I2C directions
Dir PORTC.4 Out
Dir PORTC.3 Out
' Define I2C settings
#define HI2C_BAUD_RATE 100
#define HI2C_DATA PORTC.4
#define HI2C_CLOCK PORTC.3
'Now set up the I2C slave
HI2CMode Slave
HI2CSetAddress 0xA0 '=esadecimale 'in binario= b'10100000'
SET I2C1ADR0 = 0xA0
SET I2C1CON0.MODE0=    0   'set Slave mode, four bit address
SET I2C1CON0.MODE1=    0
SET I2C1CON0.MODE2=    0

'______________
'   USART
#define USART_BAUD_RATE 9600
#define USART_DELAY 5 ms
#define USART_BLOCKING
'Set pin USART directions
Dir portC.7 IN
Dir portC.6 OUT

' ---Port settings____________
'    LEDS
#define led1 portA.6'
Dir portA.6 OUT

'__Variables_________
#define Dim_Array 10
dim Var_Array( Dim_Array )
Dim n_elm as Byte
Dim ind as Byte
Dim Address as Byte
'============= Code ======================
main:
wait 500 ms
led1 = 1
Repeat 5
    LED1 = !LED1
    wait 300 ms
    LED1 = !LED1
    wait 300 ms
end Repeat

Address=0
  sertxd(1,"=== Test"," I2C receive",1)
  for ind= 0 to Dim_Array
    Var_Array(ind)=0
  Next

  MyHI2CReceive ( n_elm )

  sertxd(1, "data number=" , n_elm , 0)
  sertxd(1,"received: address=", Address, 0)
  for ind= 1 to n_elm
    if ind =1 Then
      sertxd(1,"   Var_Array=", Var_Array(ind), 0)
    Else
      sertxd(9," ",Var_Array(ind), 0 )
    End If
  Next
  sertxd(1,"I2C1CNT=",I2C1CNT,1)
goto main

Sub MyHI2CReceive ( n_elm )
  Dim COUNT_read as byte
  n_elm = 1
PCIF= 0
I2C1CNT=100
count_read=I2C1CNT
do until count_read < 95
  SET I2C1PIR = 0
  SET I2C1PIE.ACKTIE    ON  'Enable acknowledge clock stretching
  SET I2C1PIE.PCIE      ON    'Enable STOP condition interrupt
  sertxd(1,"waiting ","inside loop",0)
  sertxd(9,"I2C1CNT=",I2C1CNT,0)
  Wait until I2C1IF = 1 '            'Wait for receive address
  ADDRESS= I2C1ADB0
  SET I2C1CON1.ACKDT OFF  ‘Ackowledge
  SET I2C1CON0.CSTR   OFF   'Release SCL

     Wait until I2C1RXIF = 1
    Var_Array(n_elm) = I2C1RXB
    n_elm = n_elm + 1
    wait 20 us
    SET I2C1CON0.CSTR   OFF       'Release SCL

    Wait until I2C1RXIF = 1
    Var_Array(n_elm) = I2C1RXB
    n_elm = n_elm + 1
    wait 20 us
    SET I2C1CON0.CSTR   OFF       'Release SCL
    count_read=I2C1CNT
    wait 50 us
Loop

  n_elm=n_elm - 1
End Sub
‘_________________________________________
SUB Sertxd(in CRLF1 as byte, in stringa as string, in inputS as String, in CRLF2 as byte)

  if CRLF1 < 2 Then
    HSerPrintCRLF CRLF1
  Else
    HSerSend CRLF1
  End if
  HSerPrint stringa
  HSerPrint inputS

  if CRLF2 < 2 Then
    HSerPrintCRLF CRLF2
  Else
    HSerSend CRLF2
  End if

END SUB
‘________________________________
SUB Sertxd(in CRLF1 as byte, in stringa as String, in inputB as Byte , in CRLF2 as Byte )

  if CRLF1 < 2 Then
    HSerPrintCRLF CRLF1
  Else
    HSerSend CRLF1
  End if

  HSerPrint stringa
  HSerPrint inputB

  if CRLF2 < 2 Then
    HSerPrintCRLF CRLF2
  Else
    HSerSend CRLF2
  End if

END SUB
‘______________________________________
SUB Sertxd(in CRLF1 as byte, in stringa as string, in inputW as Word, in CRLF2 as byte)

  if CRLF1 < 2 Then
    HSerPrintCRLF CRLF1
  Else
    HSerSend CRLF1
  End if

  HSerPrint stringa
  HSerPrint inputW

  if CRLF2 < 2 Then
    HSerPrintCRLF CRLF2
  Else
    HSerSend CRLF2
  End if
END SUB
‘__________________________________________________
Sub InitPPS
    '//sets up pins for USART

    UNLOCKPPS
        'Module: UART1
        U1RXPPS = 0x0017    'RC7 > RX1
        RC6PPS  = 0x0013     'TX1 > RC6
        'Module: I2C1
          'input
        I2C1SCLPPS = 0x0013
        I2C1SDAPPS = 0x0014
          'Output
        RC3PPS = 0x0021
        RC4PPS = 0x0022
    LOCKPPS
 End Sub

[Giuseppe D'Elia]

@Anobium
before my last post I did'nt seen your last posts (they are on page 2 and I have not noticed it).

As a matter of fact the code I sent, apart from the heavness due to the printing routines, is short and works although is not really stable. I receive two bytes at a time.

I found many difficulties because I was not able using many interrupts like ADRIF, WRIF, PCIE, although I enabled them by properly setting the PIE register bits.

And this was in contrast with the step by step procedure depicted in the PIC datasheet.

At the end I decided to use the ACKTIE enable (the one I excluded at the beginning) and to use the general interrupt I2C1IF for address receiving step managing.

I managed the received bytes by using the I2C1RXIF interrupt, described in the step by step procedure.
However, I needed a delay before releasing the SCL. (I used 20 us). I do not understand why.

All this is a result of intuition, not knowledge, because the PIC has a strange, unpredictable behaviour.
And also for me sometimes it stops and a reprogramming is needed.

Concerning your Stright question, obviously my aim is to have a working I2C slave (for receiving data only, at the moment). Although I also would like to understand how things works. A library provides a large amount of information often not exactly understandable when having the little experience I have.
Obviously in this way I lose a lot of time.

Thenk you again for your attention.

[William Roth]

@Guiseppe

You are making things difficult for youtself if your goal is to get reliable I2cSlave functionality in a project within a reasonable time . For this project you have selected the 18F45K42 which is one of the most complicated (and least supported) PIC chips available.

Unless you enjoy the frustration that goes with working with a chip that is possibly buggy and not fully supported, and you don't mind the time an frustration involved, I would suggest you use a different chip, until the issues are worked out of the K42 series of chips. It could be weeks or even months before I2C is stable on this chip.

When you you want to have a reliable working project?

Why did you select 18F45K42? Speed? Memory? Specific Peripherals?

There are other 18F PIC Chips that can do I2C slave right now. These do not have the complexity of PPS, ...Do not have the "new and unproven" I2C modues and can still run at 64MHz (if you are a speed demon). 18F45K80 ... 18F45K22 .. etc. There are many 16F Chips that can do this as well, some with PPS and some not.

---

LCD:
Your LCD module is an odball and uses special commands that are explained in the datasheet you posted. It will not work with any GCB LCD Libraries. Suggest you lay it aside and get a more common type that is fully supported. These cost about $6 - $8 on Ebay last time i checked.
Otherwise you will need to write your own code to use it effectively. And then you will likely be the only person that ever uses it.

https://www.ebay.com/itm/Yellow-Green-Serial-IIC-I2C-TWI-2004-LCD-20X4-Character-LCD-Module-For-Arduino/201013825839?hash=item2ecd5b912f:g:BHwAAOSwvg9XcPwD

https://www.ebay.com/itm/Blue-Serial-IIC-I2C-TWI-2004-204-20X4-Character-LCD-Module-Display-For-Arduino/381375068233?epid=894371540&hash=item58cbb9b849:g:oL8AAOSwFAZTviYj

[Giuseppe D'Elia]

@William and Anobium

I agree with your considerations.

I already used 18F45k22 as a Picaxe chip, with an external 16MHz cristal and an internal 4x multipler clock frequency. I programmed it using Picaxe Editor without problems. I still have one such chip available.

My application now involves the CNC axes control trough a "real time" PIC under the supervision of a Rpi zero. My application, using the Picaxe 18F45k22, runs fine at the moment.

However, I would run the PIC application code as fast as possible and, unfortunately, Picaxe uses an interpreter that slow down the code execution.

Accordingly, I decided to program by myself the PIC and I found GCBASIC. And PicKit3 and IDE and IPE...

However, I decided to not use the available 18F45k22 exploiting GCBASIC in order to not destroy the Picaxe firmware bootstrap code in the chip.

Obviously, I might have bought other PIC18F45k22.
I chosen the PIC 18F45k42 because it allows internall 64MHz clock and has huge memory and, mainly, because is an upgrade (I supposed that) with respect to 18F45k22.

Without being an expert, I did not imagine the difficulties of this my first PIC programming experience.

This is the story.

However, following your considerations, now I will buy a 18F45k22 and I will program it with a less effort, I hope.

I had not become frightened for the effort of programming the 18F45k42. It was of great interest and I still will try to solve the problems with 42 PIC as a lost time.

I sincerely thanks you and Anobium for all the support you gave me on that PIC and for your pacience.

Concerning the LCD, yes I agree with you on the benefits of buying a more update LCD.
I surely will do it.
I used the old LCD because it was the one already available at the moment and it would have given me pleasure to quicly test the demonstration codes you sent and suggested.

Thank you all

Sincerely yours
Giuseppe

[Giuseppe D'Elia]

Hi Anobium and William

within this post I do not want to re-open the last question discussed recently.
My last post still holds.

I just want report my last results on making a Slave reception with PIC18F45k42.
There are not controls on overflow or any other kind of errors handling in the subroutine.
However is what could be enough for me.

In the following I attach the PIC settings and the subroutine I used for PIC18F45k42 slave receiving.

Now everyting works and the behaviour seem to be stable.
I texted the code by sending two bytes as well as five bytes, at the moment.

I have a first PIC sending a sequence of bytes on the I2C channel when I press a button.

PIC18F45k42 receive the bytes on the same channel and print them on the Cool terminal.

Here is the PIC settings I used

#chip 18F45k42, 64
#option explicit
#config MCLRE = ON
#startup InitPPS, 85 '//Call InitPPS
'_____________
'   I2C
SET ODCONC.ODCC4 on
SET ODCONC.ODCC3 on
SET RC3I2C.TH0   OFF   'Port specific controls may be required - see the datasheet
SET RC4I2C.TH0   OFF   'Port specific controls may be required - see the datasheet
'Set pin I2C directions
Dir PORTC.4 Out
Dir PORTC.3 Out
' Define I2C settings
#define HI2C_BAUD_RATE 100
#define HI2C_DATA PORTC.4
#define HI2C_CLOCK PORTC.3
'Now set up the I2C slave
HI2CMode Slave
HI2CSetAddress 0xA0 '=esadecimale 'in binario= b'10100000'
SET I2C1ADR0 = 0xA0
SET I2C1CON0.MODE0=    0   'set Slave mode, four bit address
SET I2C1CON0.MODE1=    0
SET I2C1CON0.MODE2=    0

'______________
'   USART
#define USART_BAUD_RATE 9600
#define USART_DELAY 5 ms
#define USART_BLOCKING
'Set pin USART directions
Dir portC.7 IN
Dir portC.6 OUT

Here is the subroutine:

Sub MyHI2CReceive ( n_elm )

  n_elm = 1

  PCIF= 0
  I2C1CNT=100

  SET I2C1PIR = 0

  SET I2C1PIE.ACKTIE    ON    'Enable acknowledge clock stretching
  SET I2C1PIE.PCIE      ON    'Enable STOP condition interrupt

'______Reading the address and acknowledging
  sertxd(1,"waiting ","inside sub",0)
  sertxd(9,"I2C1CNT=",I2C1CNT,0)
  Wait until I2C1IF = 1  'Wait for receive address
  ADDRESS= I2C1ADB0
  SET I2C1CON1.ACKDT OFF
  SET I2C1CON0.CSTR   OFF   'Release SCL

'_______Reading bytes and storing in Var_Array
  do until I2C1PIR.PCIF = 1  'do until StopBit and Wait for receive data
    Wait until I2C1RXIF = 1
    Var_Array(n_elm) = I2C1RXB
    n_elm = n_elm + 1
    wait 20 us
    SET I2C1CON0.CSTR   OFF       'Release SCL
    wait 5 us
  Loop

  n_elm=n_elm - 1

End Sub

Note:
I already attempted to run a subroutine like that many times without success.
I discovered that if I use PCIF instead than I2C1PIR.PCIF I cannot detect the stop condition, i.e., PCIF stay 0 while I2C1PIR.PCIF becomes equal to 1 at the stop condition.
I don't know way.

I looked the PIC database and I noted that, while PIC name appears in the line PCIF,I2C2PIR,2 for I2C2, it appears in the line I2C1PIR_PCIF,I2C1PIR,2 for I2C1 channel.

Different ways are used for I2C2 and I2C1. Obviously, I do not know why.
Just a note.

I hope this can help you.

[William Roth]

This is because the Microchip bit naming convention is inconsistent in the Microchip files we use to make the GCB chip datafile from.

Short names like "PCIF" may not work as expected so when worlking with new chips is is generally a good practice to use "register.bit" instead of the short name.

You can always look in the Chip Datafile to see what the bit name should be used.

For example:

Open the file 18F45K42.dat in GreatDowBasic\chipdata. Then search for PCIF. You will find 2 lines. One shows its usage with I2C2 and the other with I2C1. "PCIF" is used for I2C2 while I2C1PIR_PCIF is used for I2C1 in the Great Cow BASIC dat file - this bit should be access via I2C1PIR.PCIF where you transpose the _ with a period.

Further down a few lines you will see "PC1IF,I2C1PIR,2""

This is can look inconsistent and confusing, and can lead to problems if you are not very careful

So for I2C1 you can use any of these :
I2C1PIR.PCIF
I2C1PIR.2
PC1IF

This may apply to other registers as well. Always look in the chip datafile to get the correct regiater name if you are not using (register,bit)

You can test this by creating a simple program, compiling, then looking at the ASM file

GCB Program

   Set PCIF OFF
   Set PC1IF OFF

Now Look at the ASM file. You will see ...

~~~
;set PCIF OFF
banksel I2C2PIR
bcf I2C2PIR,PCIF,BANKED
;set PC1IF OFF
bcf I2C1PIR,PC1IF,BANKED
~~~

Edited by Anobium.