PIC18F4550 to 18F4553 Comm Issue?

[ShaneMichael]

I'm having trouble figuring this one out. I have code that works on my circuit board with a 4550, I changed to a 4553 so I could use a 12bit A/D. I moved the code that works with the 4550 to the 4553. Everything is working but I can't get the EUSART communication to work. It seems like it's not getting the interrupt, to trouble shoot I put an LCD message in the receive interrupt routine with a pause, the message is never displayed. I use DT interrupts for timers/clock and that part works, the RX_INT is in the same section, so I believe that should be OK. So my thought is there may be some difference in the DEFINES for the serial port baud rates etc. Trouble is the data sheet for the 4553 doesn't have a section for the EUSART so I can't put the data sheets side x side to compare. The data sheet for the 4553 is 46 pages, 4550 data sheet is 438 pages.

Tomorrow I plan to solder some wires to the board so I can connect the serial to my PC to see if I can send/receive anything to my computer.

For now I guess I have two questions, where is the complete data sheet for the 18F4553, and do you know if the serial defines I included would work on the 4553? These work just fine on the 18F4550.

DEFINE HSER_RCSTA 90h ' Enable serial port & continuous receive   
DEFINE HSER_TXSTA 164 ' Enable brg, Enable transmit, BRGH = 1
DEFINE HSER_SPBRG 10 ' 115200 Baud @ 20MHz, -1.36%  pg 245 spbrg vals
DEFINE HSER_CLROERR 1 ' Clear overflow automatically

Thanks

[Demon]

http://www.microchip.com/wwwproducts/en/PIC18F4553

At quick glance, they both use the same pins for EUSART, so that's a good thing.

Check out this funky register on the 4553: DEVID1. EDIT: never mind, The Device ID registers are “read-only” registers

Don't look for more pages in a 4553 datasheet, there are none:

This document is supplemented by the “PIC18F2455/2550/4455/4550 Data Sheet”

It is compatible with the 4550, so you're missing something very small if you already have it working between two 4550.

Hmm, very dumb question: have you changed the model when you compile?

Robert

[richard]

page 7 of the "46 page abridged version"

Note:
This data sheet documents only the
devices’ features and specifications that are
in addition to the features and specifications
of the PIC18F2455/2550/4455/4550
devices. For information on the features
and specifications shared by
the PIC18F2458/2553/4458/4553 and
PIC18F2455/2550/4455/4550 devices,
see
the “PIC18F2455/2550/4455/4550

[Demon]

to trouble shoot I put an LCD message in the receive interrupt routine with a pause, the message is never displayed.

PAUSE within an interrupt aren't a good idea.

Set a flag to 0,
Set it within your routine,
Then check it within your main program.

What if you copy your programs for both pics down to bare EUSART code and try to get them talking?

Are you sure it worked right before with two 4550?

Uh, did you put the 4553 in backwards?

Robert

[ShaneMichael]

Thanks for the responses, I can stop looking for the rest of the data sheet. I was too busy looking for fine print I overlooked the highlighted section that says it's the same as the 4550.

I did get wires connected between the 4553 board and my PC. I put an HSEROUT in my main routine to send a message out once per second. All I get is qbert "«Š¥h”ejg«üa©Š¥Gh”", I can move the wires to the 4550 board and the message displays fine.

The 4553 isn't in backward (I've done that before, so it's the first thing I check now!). I also have an LCD and LED's that are driven from the 4553, those all work fine.

I'll start a new program for the 4553 and just put in the bare EUSART as suggested. Hopefully I can find my error.

Let me know if you have any other idea's.
Thanks

[Demon]

If your LED and LCD work fine on the 4553, then that PIC is working fine.

If your 4550 can talk to another 4550, then that PIC is working fine.

Are you using the right TX/RX pins on the 4553?

There's really not a lot of possibilities here. The 4553 is EUSART compatible with a 4550. It is basically a 4550 with a different name.

Do you use different wires?

Do you do a simple drop-in substitution for the 4553, or do you have a separate breadboard circuit set up?

What about a heartbeat LED?

Main:
LEDout = 0
PAUSE 500
LEDout = 1
PAUSE 500
GOTO Main
END

If you're using a separate circuit for the 4553, does this do a 1 second pulse?
(I was thinking of a timing issue, LED and LCD might work fine, but EUSART would be more finicky)

Just trying to eliminate all possibilities I can come up with.

Robert

[Scampy]

No expert, but these are things I look at when having issues with serial coms.

Timing: The crystal and OSC value need to be right. On one project the code ran and the program functioned, but when communications was tested I got garbage streamed out to the PC. I then changed the xtal from 12mhz to a 10mhz and it worked fine.

Try something like this with a 10mhz crystal fitted

DEFINE  OSC 20                          ' 18F4520, 10mhz crystal
ADCON1 = $0F
clear 

 ;----[UART Settings]------------------------------------------------------------

DEFINE HSER_RCSTA 90h ' Enable serial port & continuous receive
DEFINE HSER_TXSTA 24h ' Enable transmit, BRGH = 1
DEFINE HSER_CLROERR 1 ' Clear overflow automatically
DEFINE HSER_SPBRG 8   ' 9600 Baud @ SPBRGH = 2
BAUDCON.3 = 1         ' Enable 16 bit baudrate generator


RCSTA = $90   ' Enable serial port & continuous receive
TXSTA = $24   ' Enable transmit, BRGH = 1
SPBRG = 8     ' 9600 Baud @ -0.03%
SPBRGH = 2
BAUDCON.3 = 1 ' Enable 16 bit baudrate generator


RCIF                VAR PIR1.5                    ' USART receive flag
Char                VAR BYTE                      ' USART byte received
GIE                 VAR INTCON.7

Then try your HSEROUT routine

There was a useful calculator that worked out the settings based on the timings which might help http://www.picbasic.co.uk/forum/showthread.php?t=4994

[Demon]

Another option is to use the built-in oscillator. That's what I used here on the slave 18F44K22 and ran at 115,200:

http://www.picbasic.co.uk/forum/show...524#post130524

Master 18F4550 at 48 MHz using external 20MHz oscillator
Slave 18F44K22 at 64 MHz using internal oscillator

If I read the datasheet right, the internal oscillator is limitted at 8 MHz on the 4550. Maybe that's enough for your needs.

[Demon]

I see no reason why you couldn't use an external 20 MHz crystal on the 4550 and 4553, and run them at 48 MHz.

Try with these settings, I know these worked.

' 184550 external oscillator 20MHz
...
DEFINE OSC 48
...
DEFINE HSER_RCSTA 90h               ' Enable serial port & continuous receive
DEFINE HSER_TXSTA 20h               ' Enable transmit, BRGH = 0
DEFINE HSER_CLROERR 1               ' Clear overflow automatically
DEFINE HSER_SPBRG 25                ' 115200 Baud @ 48MHz, 0,16%
SPBRGH = 0
BAUDCON.3 = 1                       ' Enable 16 bit baudrate generator

I used 22pF ceramic caps on the crystal.

Like these round brown ones:
http://www.capacitorguide.com/ceramic-capacitor/

Also, notice this has a low error percentage despite the higher rate, only 0.16% compared to your 1.36% (I'm assuming negative is irrelevant on this value, I could be wrong).

One last thing, 48 MHz on the 4550 would let you use USB later if that's a requirement.

Robert

[ShaneMichael]

Still no luck, the data I receive on my PC is garbled from the 4553, I've tried dozens of baud rates and settings. I plugged the com back into a board with a 4550 and the data from it is fine. Both are running external 20 MHz crystals, boards and design is similar. I copied the code and board design from the 4550 and just changed what I needed to for the new design. I tried the suggested settings from here and had the same results. I got an error when I changed the OSC to 48.

Also, my title may be confusing. What I'm doing is upgrading from a 4550 to 4553 on a board design/upgrade, the EUSART comm isn't working correct on the 4553 yet.

Tomorrow I'll take one of the boards, de-solder the 4553 and put the 4550 on it just so I can completely double check for an issue with the board.

As a little backstory the boards take external values and send the information via a cell phone to a remote server. This is working fine on boards with the 4550, but we wanted the 12bit AD for the next phase.

The cell phone needs a baud of 115200.

I'm not sure if something that I have configured might load defaults a little different from the 4550 to 4553?

Below is the current DEFINES and CONFIGS, after this I set some variables and have a 1 second loop the sends HSEROUT ["Hello",13,10], toggles an LED and updates a seconds counter on the LCD. Data received by PC acts like the baud rate its bad, LED blink and LCD updates fine.

Let me know if you think of anything that may help.
Thanks

Define OSC 20

   
DEFINE LCD_DREG          PORTD  'Port used for LCD Data
DEFINE LCD_DBIT          0      'First Pin connected to LCD DB4
DEFINE LCD_RSREG         PORTB 'Port used for LCD RS line
DEFINE LCD_RSBIT         2     'Pin used for LCD RS line

DEFINE LCD_EREG          PORTA 'Port used for LCD E Reg
DEFINE LCD_EBIT          4     'Pin used for LCD E line
DEFINE LCD_BITS          4     'Four bit comm mode
DEFINE LCD_LINES         2     '2 Line LCD
DEFINE LCD_COMMANDUS     900    'Delay between sending LCD commands
DEFINE LCD_DATAUS        75      'Delay time between data sent
  
DEFINE HSER_RCSTA 90h ' Enable serial port & continuous receive
DEFINE HSER_TXSTA 24h  'External Clock, Enable transmit, BRGH = 1
DEFINE HSER_SPBRG 10 ' 115200 Baud @ 20MHz, -1.36%   
DEFINE HSER_CLROERR 1 ' Clear overflow automatically  
 BAUDCON.3 = 0        ' BRG16 = 0      
 
  TRISC.7=1
  TRISC.6=1      
        
        
  'Re-Configure default values from the PBP 18F4553.inc and MCSP/MPASM .inc
        
   ' Ext XTAL setting
     '_PLLDIV_5_1L  20Mhz HS / 5 for PLL 4 Mhz
    '_CPUDIV_OSC3_PLL4_1L  24Mhz CPU for USB 1 speeds, PLL 96Mhz/4 =24Mhz
    'HS oscillator, PLL enabled (HSPLL)  
       '                   
    'asm
    #CONFIG
          __CONFIG  _CONFIG1L, _PLLDIV_5_1L & _CPUDIV_OSC3_PLL4_1L & _USBDIV_2_1L
          __CONFIG  _CONFIG1H, _FOSC_HSPLL_HS_1H
    #ENDCONFIG   
    'endasm
    
    INCLUDE "DT_INTS-18.bas"     ; Base Interrupt System
    INCLUDE "ReEnterPBP-18.bas"     ; Include if using PBP interrupts
    INCLUDE "Elapsed_INT-18.bas"  ; Elapsed Timer Routines
    ASM
INT_LIST  macro    ; IntSource,        Label,  Type, ResetFlag?
            INT_Handler    TMR1_INT,  _ClockCount,   PBP,  yes
            INT_Handler    RX_INT,     _Check_Command, PBP, yes 
        endm
        INT_CREATE               ; Creates the interrupt processor
    ENDASM

   
@ INT_ENABLE    TMR1_INT     
@ INT_ENABLE    RX_INT

[richard]

#CONFIG
__CONFIG _CONFIG1L, _PLLDIV_5_1L & _CPUDIV_OSC3_PLL4_1L & _USBDIV_2_1L
__CONFIG _CONFIG1H, _FOSC_HSPLL_HS_1H
#ENDCONFIG

with a 20 MHz xtal would result in a fosc of 24 MHz


Define OSC 24

DEFINE HSER_SPBRG 12

[ShaneMichael]

Ok, so now it works, but I'm more confused...Define OSC 24 causes and error (Frequency not defined), but I changed HSER_SPBRG to 12 any way, and everything is working fine.
So, why? Apparently my understanding of figuring out the settings is incorrect, although I must have understood it at one point

Can anyone explain this?
Thanks

[richard]

for pbp3 Define OSC 24 is a valid define
from the book

2.3.1 DEFINE OSC
There are a limited number of valid numbers that can be used: 3 (3.58MHz), 4, 8,
10, 12, 16, 20, 24, 25, 32, 33, 40, 48, 64. These are the only frequencies for which
PBP is able to accurately calibrate its timing. If you use a system clock that runs at
a frequency that isn't listed here, your timing will be scaled when the program
executes.

24 is in the list

osc setting

                                                   USBDIV_2_1L
                                    / =========== 96/2 = 48 ----   usb
20 mhz  /5  = 4 :-  pll =  96mhz  :
      PLLDIV_5_1L                   \============  96/4 = 24 ---   fosc
                                                   CPUDIV_OSC3_PLL4_1L

[Demon]

' 184550 external oscillator 20MHz
...
DEFINE OSC 48
...
DEFINE HSER_RCSTA 90h               ' Enable serial port & continuous receive
DEFINE HSER_TXSTA 20h               ' Enable transmit, BRGH = 0
DEFINE HSER_CLROERR 1               ' Clear overflow automatically
DEFINE HSER_SPBRG 25                ' 115200 Baud @ 48MHz, 0,16%
SPBRGH = 0
BAUDCON.3 = 1                       ' Enable 16 bit baudrate generator

I should have included the CONFIGs though, they are still in PBP v2.60c format but the settings are easily convertible:

@   __CONFIG    _CONFIG1L, _PLLDIV_5_1L & _CPUDIV_OSC1_PLL2_1L & _USBDIV_2_1L
@   __CONFIG    _CONFIG1H, _FOSC_HSPLL_HS_1H & _FCMEN_OFF_1H & _IESO_OFF_1H
@   __CONFIG    _CONFIG2L, _PWRT_OFF_2L & _BOR_ON_2L & _BORV_1_2L & _VREGEN_ON_2L
@   __CONFIG    _CONFIG2H, _WDT_OFF_2H
@   __CONFIG    _CONFIG3H, _CCP2MX_OFF_3H & _PBADEN_OFF_3H & _LPT1OSC_OFF_3H & _MCLRE_ON_3H
@   __CONFIG    _CONFIG4L, _STVREN_ON_4L & _LVP_OFF_4L & _ICPRT_OFF_4L & _XINST_OFF_4L & _DEBUG_OFF_4L

But Richard is a senior member for a reason.

Robert

[ShaneMichael]

Thought I give an update. I did get the communication working at 115200 using the 24 Mhz clock setting. But then my DT interrupts timer was running 20 % fast (24/20). If enable the HSPLL it seems PLL becomes my primary clock, my understanding is the FOSC3:FOSC0 and OSCCON would select the primary oscillator if set? So I guess I'm not reading the data sheet correct or just not getting it set correct.

To keep DT timer correct I changed to _FOSC_HS_1H and _CPUDIV_OSC1_PLL2_1L to disable the PLL. DT interrupts are timing correct again.

Then I changed communication settings to

DEFINE HSER_RCSTA 90h ' Enable serial port & continuous receive
DEFINE HSER_TXSTA 24h  'External Clock, Enable transmit, BRGH = 1
DEFINE HSER_SPBRG 10 ' 115200 Baud @ 20MHz, -1.36%  
DEFINE HSER_CLROERR 1 ' Clear overflow automatically  
BAUDCON.3 = 0        ' BRG16 = 0

The OSC define is set to 20. Communication and timing is all working correct again.

As a side note, since I couldn't get Define OSC 24 to work I tried frequencies from the list

2.3.1 DEFINE OSC
There are a limited number of valid numbers that can be used: 3 (3.58MHz), 4, 8,
10, 12, 16, 20, 24, 25, 32, 33, 40, 48, 64

The frequencies that didn't cause an error are; 4, 8, 12, 16 and 20. The others give me an error of frequency not defined or previously used. Not sure what's causing that?

I'm still not sure why the Clock/frequency/PLL configuration settings on the 4550 work, but on the 4553 I needed to change _FOSC_HS_1H and _CPUDIV_OSC1_PLL2_1L to keep DT clock running at 20Mhz and Communications clocking at 20 Mhz? The circuit board setup is the same, crystal is the same, same DT interrupts loading etc. The thing that isn't the same is the 4550 is programmed with a boot loader, and the 4553 I'm programming with PicKit 3. I wonder if the boot loader was changing some of the configurations on the 4550 in the background that I wasn't setting?

Anyway, I'm up and running, thanks for the help! Let me know if you have any more insight that would help me get unconfused.
Thanks
Shane

[Demon]

I wouldn't be surprised if either the bootloader or the Pickit 3 was overriding your config fuses.

But I've never used either.

Robert

[tumbleweed]

The thing that isn't the same is the 4550 is programmed with a boot loader

You probably should have mentioned that in post #1.

Most bootloaders won't allow you to reprogram the CONFIG section (since that will likely break the bootloader), while the PICkit3 will load whatever you've set in the hex file.

[ShaneMichael]

I defiantly would have mentioned Pickit vs bootloader if I thought that would have anything to do it. I thought they would program what ever I had configured. Sounds like that may not be how the boot loader works though? Good to know.

Thanks

[tumbleweed]

It depends on the bootloader you're using... some will, some won't, so it's worth checking.

It's usually a good idea NOT to allow the config to be reprogrammed. For example, if you goof up/change the osc setting you can easily turn your board into a brick.