View previous topic :: View next topic |
Author |
Message |
randy.shaffer
Joined: 21 Mar 2018 Posts: 51
|
Help with rs-232 for PIC16F18323 [SOLVED] |
Posted: Fri Oct 13, 2023 2:51 pm |
|
|
Compiler is 5.116, programmer is ICD-U64. I was expecting the first message, then the second message each second thereafter:
Code: |
#include <16F18323.h>
#fuses PUT, NOWDT
#use delay(internal=32MHZ)
#pin_select U1TX=PIN_C4
#use rs232(UART1,baud=28800,parity=N,bits=8,errors,stream=VT)
void main(void)
{
fprintf(VT, "Hello World!!\r\n");
while(TRUE)
{
fprintf(VT, "Hello World??\r\n");
delay_ms(1000);
}
}
|
The above code produces this output using CoolTerm:
Code: |
Hello World!!
Hello World?ΓΏ
|
The program seems to choke on the last question mark. Same behavior whether U1RX is specified or not or if the delay_ms function is present or not. Any ideas?
Last edited by randy.shaffer on Mon Jun 10, 2024 10:32 am; edited 1 time in total |
|
|
temtronic
Joined: 01 Jul 2010 Posts: 9220 Location: Greensville,Ontario
|
|
Posted: Fri Oct 13, 2023 8:37 pm |
|
|
so have you just used a loopback wire on the PC serial link and confirmed it's not a 'coolterm' or other PC problem ??
also what is the actual hardware connecting PIC to PC ? I have to assume it's not PIC -->1488 to PC Comport pin 3...... |
|
|
randy.shaffer
Joined: 21 Mar 2018 Posts: 51
|
|
Posted: Mon Oct 16, 2023 2:24 pm |
|
|
If I re-compile with ICD = TRUE, the program works in the debugger. But it doesn't work outside of the debugger. Do I need a #device ICD = something? to make it work on its own? |
|
|
temtronic
Joined: 01 Jul 2010 Posts: 9220 Location: Greensville,Ontario
|
|
Posted: Mon Oct 16, 2023 2:50 pm |
|
|
Hopefully someone that uses that ICD-U64 will respond BUT my gut say that 'ICD= FALSE' is needed in your code
PLUS
in the IDE, set the 'build configuration' to 'release' NOT 'debug'.
Something.. IDE, programmer, man in the moon, is not configuring for 'release', instead going to 'debug' .
Again, i don't use the ICD-U64 but got 'stung' when MPLAB 8.xx defaulted to 'debug'. I called them, they changed their code, MY default is 'release'. The IT guy couldn't understand that I never use 'debug'.....3 hours later he finally understood......
Jay |
|
|
randy.shaffer
Joined: 21 Mar 2018 Posts: 51
|
|
Posted: Mon Oct 16, 2023 3:34 pm |
|
|
Jay, thank you for the insight. It now gets more weird. I tried ICD = FALSE, but the compiler didn't like it. Then I tried various ICD = something and got no change. Then I tried
And it worked! What's more bizarre is that statement seems to have flipped a switch as now I am able to compile without the device statement and everything is working as expected. So, ICD = nothing apparently made the IDE, programmer, or the man in the moon happy. If anyone out there can shed some light on how/why this worked, I would greatly appreciate it! Jay, thanks again for the inspiration. |
|
|
newguy
Joined: 24 Jun 2004 Posts: 1907
|
|
Posted: Mon Oct 16, 2023 4:15 pm |
|
|
If you're using the IDE, View -> C/ASM list. If not, there should be a .lst file in the directory.
At the bottom is the fuses; it will look something like this:
Code: | Configuration Fuses:
Word 1: FF8C NOEXTOSC RSTOSC_HFINTRC_64MHZ NOCLKOUT CKS FCMEN
Word 2: F73D MCLR PUT NOLPBOR NOBROWNOUT BORV19 ZCDDIS NOPPS1WAY STVREN NODEBUG NOXINST
Word 3: FF9F WDTSW NOWDT WDTWIN_SW WDTCLK_SW
Word 4: DFFF NOWRT NOWRTC NOWRTB NOWRTD SCANE NOLVP
Word 5: FFFE PROTECT NOCPD
Word 6: FFFF NOEBTR NOEBTRB |
Compile & look at the .lst file with & without the ICD=TRUE along with the weird one you found which works. Sounds like a compiler bug which seems to perhaps be messing with the fuses? |
|
|
randy.shaffer
Joined: 21 Mar 2018 Posts: 51
|
|
Posted: Wed Oct 18, 2023 9:43 am |
|
|
Thank you, newguy, I tried compiling with ICD=TRUE and ICD= and found no difference in the fuses. Does anyone know if there is an ICD=something that corresponds to released code? The compiler doesn't like ICD=FALSE or ICD=0. By the way, should the statement #device ICD = (whatever) come before or after the #fuses directive? |
|
|
newguy
Joined: 24 Jun 2004 Posts: 1907
|
|
Posted: Wed Oct 18, 2023 9:44 am |
|
|
With CCS, whatever comes last in the file overrides what came before when it comes to fuses. Same applies to the clock (#use delay) directive. |
|
|
randy.shaffer
Joined: 21 Mar 2018 Posts: 51
|
VDD not connected |
Posted: Mon Jun 10, 2024 10:34 am |
|
|
I didn't have VDD connected because I wasn't using the VDD from the ICD-64. However, it appears that VDD must be connected for everything to work properly, even when the MCU is powered by a source external to the programmer. |
|
|
Ttelmah
Joined: 11 Mar 2010 Posts: 19488
|
|
Posted: Mon Jun 10, 2024 11:34 am |
|
|
The programmer unit, when setup with 'target supplies power', uses
the supply connection to test that the chip has a legal supply, and will
not allow you to program, unless it is seen as OK. Hence your problwm.
A lot of the units will give you an error message to say that the supply
is not OK, but this depends on the ICD, and the firmware involved. |
|
|
randy.shaffer
Joined: 21 Mar 2018 Posts: 51
|
|
Posted: Mon Jun 10, 2024 11:37 am |
|
|
Thank you for the clarification, much obliged. |
|
|
|