Resolved: temperature not displaying in demo code

[HackInBlack]

i have adapted one of the demo coes for the microchip PICDEM2 plus 2010 (in the vendor boards folder)
the first call is for a temperature display followed by voltmeter and clock.the volts and time routines both work;
but somehow it seems like the temperature call doesn't work as the display goes blank.it's got me baffled.

[Anobium]

I would isolate this issue.

As you have noted that ShowI2CTemperature() does not operate as expected.

So, take that routine and change as below:

Sub ShowI2CTemperature

    InitLCDChar

  // Set temperatureVal - set to 20 as a test

        temperatureVal = 20
        Locate 0,12
        Print temperatureVal
        LCDWriteChar 0
        Print "C "

End Sub     // GCBASIC Program.

If this shows 20c on the LCD then you know the issue is in the I2C interface.

---

Have you tried I2CDiscovery to ensure the I2C address is correct?
Is the I2CTEMPSENSOR 0x9A found?

[HackInBlack]

i've made a version of I2C discovery for this chip/board and the results are these:-

it seems neither the temperature sensor nor the eeprom are found (they share a bus)

[Anobium]

Which board are you using?

[HackInBlack]

The PICDEM2 plus c2002 with a PIC18F4431

[Anobium]

Is this your board ? https://github.com/GreatCowBASIC/Demonstration_Sources/tree/main/Vendor_Boards/Microchip_PICDEM%202%20plus_board_2002

If not, what is the URL to your Board? I need the circuit diagram.

[HackInBlack]

yes thats the board i'm using; i tried to post the PDF dated 2004 revB,but this forum software doesn't like PDF as an extension...delivery failed due to a security concern.
i've checked the manual (both ones!) and the schematic is identical for both years and revisions
the pinouts are here:-

[Anobium]

I assume the board is OK.
And, I know that this is a very old chip.
Therefore, I looked as the ASM generated by your program.

It has the following - meaning a RAM variable called SSPCON2 has been created by the HWI2C library. This means the library has not been tested with that chip. The library expects the SSPCON2 I2C register, as this does not exist... it creates a byte variable.

SSPCON2 EQU 62 ; 0x3E (SA)

---

Next, open PICINFO to access the Datasheet.

Page 212. 19.3 SSP I2C Operation
The SSP module, in I2C mode, fully implements all slave functions except general call support and provides interrupts on Start and Stop bits in hardware to facilitate firmware implementations of the master functions. The SSP module implements the standard mode specifications, as well as 7-bit and 10-bit addressing.

So, this is a slave only I2C implementation.

---

So, changing the program to software I2C and it with work OK as a MASTER device. Change the I2C Discovery program ( I would use an existing Software i2C discovery program and just change the chip etc. Then, change the other program. I2C discovery is always a pre-req).

[Anobium]

Please also try this. Add the following to HWI2C.h, in the #script section of the library.

This will inspect 18F chips, ensure SSP2CON2 exists and checks for HI2C_DATA. If this failes the compiler will issue a warning.

  //Check for SLAVE only 18F chip
  If ChipSubFamily = 16000 Then
    If NOVAR(SSPCON2) Then
      If DEF(HI2C_DATA) Then
        // So, this is an 18F chip, has not got SSP2CON2 and the user has defined HI2C_DATA
        Error "Microcontroller does not support hardware I2C, please use software I2C."
      End If 
    End If 
  End If

---

I have changed from ChipFamily = 16 to ChipSubFamily = 16000. ChipSubFamily = 16000 are the older chips.

I have added this to the master build here, run all the tests and this seems to work OK. Should trap and issue a message when appropriate.

[HackInBlack]

IT LIVES! i finally got the code to do what it was meant to do.Here are the files for the 'demo' folder
I've included the assembly file to put the code 'under the microscope'

[Anobium]

Excellent.

We have improved the I2C library which is a good thing.

Will you be posting a complete suite of demos? including I2C discovery?

[HackInBlack]

i'll try 80) here's the discovery one .

[Anobium]

Thanks. Great work

Will you be publishing a suite that I can put in the demo folder?

[HackInBlack]

at the moment i'm trying to cover the gaps i left in the 16F76 examples,due to running out of pins etc. so its demo's 14 15 & 16 (the I2C eeprom routines) from the demonstration example for the early board version.i also have to sidestep the code with PPS used...watch this space!

[HackInBlack]

here's demo 14 with three errors,which i think reference other parts of the original code...not sure where they are yet!

[Anobium]

Here goes.

    // Common macro and constants
        #define LEDS_OFF            LED_D8=0:LED_D7=0:LED_D6=0:LED_D5=0:LED_D4=0:LED_D3=0:LED_D2=0:LED_D1=0

    // Variable Definitions
        Dim labNumber as Byte:          labNumber   = 0               
        Dim switchEvent as Byte:        switchEvent = 0

Sub checkButtonS2

    if btnState = NOT_PRESSED Then
        if SWITCH_S2_PORT = 0 Then  
            wait 100 ms
            btnState = PRESSED
        end If
    else if SWITCH_S2_PORT = 1 Then
            btnState = NOT_PRESSED
            switchEvent = 1                                                   
    end if

End Sub

I took these from .. Demos\Vendor_Boards\Microchip_Explorer_8_Demo_Board\PIC18F\18F67K40_Explorer8_Capability_demo.gcb

What was your source GCB? Did you purge these pieces out?

[HackInBlack]

This is my reference file:-
C:\GCstudio\gcbasic\demos\Vendor_Boards\Microchip_PICDEM_2_plus_board_2006
i unpicked the eeprom code from the multiple demo's contained in the one 'capability' program!
hence the 'orphaned' bits throwing errors...but now i'm down to three errors with,i believe, 2 root causes;getting there slowly!

[HackInBlack]

here's the errors:-

[HackInBlack]

new version with 2 GCASM errors that have me stumped

[Anobium]

Both caused by typos in the source program, specifically SW2 PORT

To resolve this sort of issue. Here is my method for the following:

Error: Variable 4PORT was not explicitly declared
 Error: GCASM: Symbol [SFR]4PORT has not been defined. Inspect ASM file to determine error  at `ANDWF 4PORT, W,BANKED` Hexfile address(see lst file): 000390
 Error: GCASM: Symbol [SFR]4PORT has not been defined. Inspect ASM file to determine error  at `ANDWF 4PORT, W,BANKED` Hexfile address(see lst file): 0003AC

1. Ensure GCBASIC source code is shown in ASM.
2. Open the LST file produced. Search for the last hex number in the error message - 00390 This will show the error source. In this case 000390 1500 ANDWF 4PORT, W,BANKED. This confirms the source.
3. Open the ASM, search for 4PORT.
4. Examine the GCBASIC source for error
5. Correct error.

[HackInBlack]

OK now i've followed your instructions,and found out how to enable basic code in the assembly listing; i eventually figured out it should be just :-
' SW2 ' and Not ' SW2 PORT ' a silly mistake; as SW2's PORT was already defined! The code now produces a hex file....BUT it doesn't seem to run!
EDIT it is writing to an external eeprom? so i can't check the contents using PK2..hmmm how would i go about this?

[Anobium]

:-) That is good re compiling. However, do you get any thing displayed ?

[HackInBlack]

yippee i get results in PUTTY

---

nope

[Anobium]

Check i2CDiscovery with a full power cycle the board.

Remove the eeprom_wr_array and eeprom_rd_array function. Do you get action on the display? Work back until you get the display working. What is serial debug showing?

[HackInBlack]

with rd and wr remmed out putty shows this:-

I2C EEPROM
Initial Array
-:1,2,3,4,5,6,7,8,9,10

Repopulate Array with #127 to ensure the EEPROM results are read correctly
    -:127,127,127,127,127,127,127,127,127,127

Read Array from EEProm results
    -:127,127,127,127,127,127,127,127,127,127

End of array operations

LCD is still blank after power cycling the board