Can't set 38400 baud @ 48Mhz

[b1arrk5]

Hi Darrel,

Here are the lines from the .LST file. I hope that you find something useful here. It's great that we can come here and get professional level assistance from guys like you, Skimask and Steve, especially on the weekends when a lot of us have more time to work on our projects, so if I can provide any additional information please let me know.
code
Warning[202]: Argument out of range. Least significant bits used.
000498 0E38 07695 movlw HSER_SPBRG
00049A 6EAF 07696 movwf SPBRG ; Set baud rate generator reg
00049C 0E20 07697 movlw HSER_TXSTA
00049E 6EAC 07698 movwf TXSTA ; Set transmit control reg
0004A0 0E90 07699 movlw HSER_RCSTA
0004A2 6EAB 07700 movwf RCSTA ; Set receive control reg
07723 LIST
0004A4 07724 main
0004A4 EF1F F006 00012 goto MAINPROG
00013 include "USB18.INC"
00001 ; MUID = Microchip USB Class ID
00002 ; Used to identify which of the USB classes owns the current
00003 ; session of control transfer over EP0
00004 #define MUID_NULL 0
00005 #define MUID_USB9 1
00006 #define MUID_HID 2
00007 #define MUID_CDC 3
00008 #define MUID_MSD 4
00009
00010 ; Buffer Descriptor Status Register Initialization Parameters
00011 #define _BSTALL 0x04 ; Buffer Stall enable
00012 #define _DTSEN 0x08 ; Data Toggle Synch enable
00013 #define _INCDIS 0x10 ; Address increment disable
MPASM 5.06.4 MARCH3~1.ASM 3-4-2007 18:59:54 PAGE 23
/code
I don't pretend to be an assembler wiz, I did take a course years ago in 6800 and 68000, but I didn't use it for more than ten years, and of course the Pic is a little different. But I don't see any mention of SPBRGH anywhere here.
Once again, thanks for all your effort, and I look forward to hearing what you figure out.

Jerry.

[skimask]

(It's great that we can come here and get professional level assistance from guys like you, Skimask)

Warning[202]: Argument out of range. Least significant bits used.
000498 0E38 07695 movlw HSER_SPBRG
00049A 6EAF 07696 movwf SPBRG ; Set baud rate generator reg
00049C 0E20 07697 movlw HSER_TXSTA
00049E 6EAC 07698 movwf TXSTA ; Set transmit control reg
0004A0 0E90 07699 movlw HSER_RCSTA
0004A2 6EAB 07700 movwf RCSTA ; Set receive control reg

Jerry.

Wow! I ain't never been called no prefeshunol befowere...Gawllleee

Anyways...yep, right there at the top. The compiler isn't handling SPBRGH correctly, basically putting a value over 256 into a byte. However, I suspect that if you tell the compiler you want a baud rate of 13,636 (8 bit async, brg16 = 0, brgh = 0, 48mhz Fosc, etc.etc according to the datasheet should give you 56 in the SPBRGH, which should get automatically set according to the HSER defines), and then manually set BRGH high later on, it'll an up at 38400 baud. I would think this is something you should've been able to do manually, but I could be wrong...

[mister_e]

i suspect a asm math overflow, but can't reproduce it here... and... HSER_BAUD is not define so... i'm lost

Even worst.. if does the same thing even by writing to the USART register ???

what happen if you use only the CONFIG fuse, DEFINE OSC AND HSER DEFINEs?

edit: OK only by writing to the REGISTER, you'll have this error once you add a HSEROUT line, but it fix the problem when you use the DEFINEs as generated by the PicMulticalc as bellow

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 56  ' 38400 Baud @ 48MHz, -0.16%
SPBRGH = 1
BAUDCON.3 = 1         ' Enable 16 bit baudrate generator

So the DEFINEs have to be there IF you're using HSEROUT/HSERIN... just because PBP try to load HSER_SPBRG and everything else
simple code

DEFINE OSC 48
RCSTA = $90   ' Enable serial port & continuous receive
TXSTA = $24   ' Enable transmit, BRGH = 1
SPBRG = 56    ' 38400 Baud @ 48MHz, -0.16%
SPBRGH = 1
BAUDCON.3 = 1 ' Enable 16 bit baudrate generator

HSEROUT ["HELLO"]

and it's .LST around warning 202

Warning[202]: Argument out of range.  Least significant bits used.
000018 0E38           07668         movlw   HSER_SPBRG
00001A 6EAF           07669         movwf   SPBRG           ; Set baud rate generator reg
                      07670     ifdef HSER_SPBRGH
                      07671       ifdef SPBRGH
                      07672         movlw   HSER_SPBRGH
                      07673         movwf   SPBRGH          ; Set baud rate generator reg high
                      07674       endif
                      07675     endif
00001C 0E20           07676         movlw   HSER_TXSTA
00001E 6EAC           07677         movwf   TXSTA           ; Set transmit control reg
000020 0E90           07678         movlw   HSER_RCSTA
000022 6EAB           07679         movwf   RCSTA           ; Set receive control reg

INTERESTING !!!! HSER_SPBRGH is not even listed in the new PBP manual!!!

[Darrel Taylor]

Yup, you got it Steve.

Without the Define's, PBP defaults to 2400 baud and BRGH=0, which is too low for 48mhz.

But, in the first post, the defines were there. And those defines compile just fine.

So confused.

[mister_e]

yeah, weird... maybe a beta V2.46? who knows
12:14 here so

[Darrel Taylor]

Beta 2.46

[skimask]

Beta 2.46

You laugh...but for some reason I think I might have a one-off version of 2.46. I got it a couple of months after it came out. Then noticed that patch awhile after that, still haven't been able to get that applied successfully. Some of the PBP INC files are labelled 2.45, some much earlier than that (but then again, no need to waste time by renaming stuff if you didn't do anything to it right?). I emailed Jeff about it awhile back, he sent me an email back, I never did get around to sending a copy of my registration, etc. to get a new CD and he mentioned 2.47 was on it's way out, so I held off...
I'm wondering if this 2.46 isn't my problem with the SPI on the '4620, worked fine on the '452, never could get it to work on the '4620. Maybe 2.47 will fix that too, maybe Microchip will someday publish another errata...

At any rate... I still wonder if the original poster could easily get around this whole problem by setting the register values manually instead of relying on the compiler to handle it automatically...