 |
 |
| View previous topic :: View next topic |
| Author |
Message |
bryant@balancebitsconsult
Joined: 21 Nov 2023
Posts: 38
|
| #INT_RDA triggering at start before interrupts enabled ? |
 Posted: Fri Mar 08, 2024 6:31 pm |
 |
|
Is there a reason why a PIC24 would trigger the INT_RDA interrupt handler at startup ?
I receive an interrupt even before I have enabled INT_RDA interrupts, or before enabling global interrupts.
Is there some #FUSES I need to enable to prevent interrupts from starting before I enable them ?
available FUSES:
| Code: |
//////// Fuses: WRT,PROTECT,NOGSSK,GSSK,FRC,FRC_PLL,PR,PR_PLL,SOSC,LPRC
//////// Fuses: FRC_DIV_BY_16,FRC_PS,IESO,EC,XT,HS,NOPR,OSCIO,IOL1WAY,CKSFSM
//////// Fuses: CKSNOFSM,NOCKSNOFSM,WPOSTS1,WPOSTS2,WPOSTS3,WPOSTS4,WPOSTS5
//////// Fuses: WPOSTS6,WPOSTS7,WPOSTS8,WPOSTS9,WPOSTS10,WPOSTS11,WPOSTS12
//////// Fuses: WPOSTS13,WPOSTS14,WPOSTS15,WPOSTS16,WPRES32,WPRES128,PLLWAIT
//////// Fuses: WINDIS,WDT,NOPUT,PUT2,PUT4,PUT8,PUT16,PUT32,PUT64,PUT128
//////// Fuses: BROWNOUT,ALTI2C1,ALTI2C2,ICSP3,ICSP2,ICSP1,RESET_AUX
//////// Fuses: RESET_PRIMARY,JTAG,DEBUG,AWRT,APROTECT,NOAPLK,APLK
|
Currently #FUSES are:
| Code: |
#include <24EP512GU810.h>
#device ICSP=1
#device ADC=12
#use delay(clock=120MHz,oscillator=4MHz)
#FUSES NOWDT //No Watch Dog Timer
#FUSES GSSK //General Segment Key bits, use if using either WRT or PROTECT fuses
#FUSES IESO //Internal External Switch Over mode enabled
#FUSES OSCIO //OSC2 is general purpose output
#FUSES NOIOL1WAY //Allows only one reconfiguration of peripheral pins
#FUSES CKSFSM //Clock Switching is enabled, fail Safe clock monitor is enabled
#FUSES PLLWAIT //Clock switch to PLL will wait until the PLL lock signal is valid
#FUSES WINDIS //Watch Dog Timer in non-Window mode
#FUSES PUT128 //Power On Reset Timer value 128ms
#FUSES BROWNOUT //Reset when brownout detected
#FUSES RESET_PRIMARY //Device will reset to Primary Flash Reset location
#FUSES JTAG //JTAG enabled
#FUSES APLK //Auxiliary Segment Key bits, use if using either AWRT or APROTECT fuses
#FUSES CKSFSM //Clock Switching is enabled, fail Safe clock monitor is enabled
|
|
|
 |
temtronic
Joined: 01 Jul 2010
Posts: 9243
Location: Greensville,Ontario
|
|
| Posted: Sat Mar 09, 2024 11:54 am |
|
|
I would have to read the datasheet to see IF the UART buffer is internally cleared on powerup.
Old skool, you 'init' all the RAM and clear /set registers the way YOU need.
Some areas come up 0,1 or ugh..... x.... (unknown..50/50 draw ) |
|
|
Ttelmah
Joined: 11 Mar 2010
Posts: 19539
|
|
| Posted: Sun Mar 10, 2024 7:46 am |
|
|
The classic reason it'd trigger at startup, is that the RX line into the UART
is not rising immediately. This always happened on traditional UART's, when
using the MAX232, since the charge pump in this took a while to build
up it's internal voltage. A low line on the TX output of this was seen as
received data on the device attached.
The solution is simple. Make sure the RDA line is high, and the UART buffer
is flushed _before_ you enable the interrupt,
So you have something like:
| Code: |
#USE RS232(UART1, BAUD=57600, BITS=8, ERRORS, STREAM=SERIAL, NOINIT)
//Depending on the chip you may not be able to read the line unless
//NOINIT is used
int error_cnt=0;
while (input(UARTRX_PIN)==0)
{
delay_ms(10);
if (error_cnt++>50)
{
your_error_code(); //do something to handle the error
}
}
//Now iniit the UART
setup_UART(1,SERIAL); //This starts the UART
|
Because this does not physically start the UART, till the RX line is high
no spurious characters are received. |
|
|
bryant@balancebitsconsult
Joined: 21 Nov 2023
Posts: 38
|
|
| Posted: Mon Mar 11, 2024 10:48 am |
|
|
I would tend to agree with you, but I am getting an interrupt BEFORE I enable any interrupts. In debugging I can see that I am getting a INT_RDA which is hitting the ISR breakpoint set AT STARTUP, before I even get to enabling the interrupts later on in code!
The serial line is even pulled up -- but that code happens after the interrupt is triggered too...
It also appears that the board resets very shortly after hitting this interrupt .. I am having difficulty determining why the board is getting the serial interrupt and/or why the board is resetting so quickly after startup.
I can't catch the restart code in debug code, but the restart code AFTER the first restart reports "RESTART_POWER_UP", indicating that this might be normal ?? |
|
|
Ttelmah
Joined: 11 Mar 2010
Posts: 19539
|
|
| Posted: Mon Mar 11, 2024 11:35 am |
|
|
If UART is not enabled it can't generate an interrupt. If you look at
what I'm suggesting, you physically do not turn the UART on, until the
line is seen as high. Problem is the default behaviour is to enable the
UART right at the start of the code. Even a slight delay in the line going
high, will lead to a character being seen, and hence a spurious interrupt.
An interrupt being set, should not matter at all if it is not enabled. So
if you don't want to delay the UART setup, just pause and read any
characters in the UART input buffer before enabling the interrupt.
Remember a framing error will trigger a UART interrupt, if you don't
have the errors handled separately. This bit has to be turned off or
this will keep triggering.
The only way the interrupt handler can be called before it is enabled
is something like a stack error, or a jump to an invalid address.
What happens if you reduce your code to a basic flash an LED test.
Does this work correctly???.
There was a [erson with a similar problem on the Micochip forum, and
it turned out to be inadequate smotthing on the CPU supply. I'd look
carefully at this, and add the traps.c handler, and see if you may be
triggering a processor trap. |
|
|
bryant@balancebitsconsult
Joined: 21 Nov 2023
Posts: 38
|
|
| Posted: Mon Mar 11, 2024 12:23 pm |
|
|
I see it now, the NOINIT which I assume does not enable the interrupts -- so even though I am "enabling" serial interrupts later in code the enable interrupts call is redundant since my #use block has already turned the ones related to serial on.
I will also take a look at the power issue you mention, this situation is very possible. |
|
|
Ttelmah
Joined: 11 Mar 2010
Posts: 19539
|
|
| Posted: Wed Mar 13, 2024 5:01 am |
|
|
No.
NOINIT, does not enable the peripheral. Does not configure it, and does not
turn it on at all. The #USE does not enable the interrupt.
The point is you cannot get an interrupt at all from a non enabled peripheral.
If your code is ending up in the interrupt handler routine, with the interrupt
disabled, it implies it is getting to this address 'another way'. So either a
'return' with an invalid address on the stack, a jump to an address in the
interrupt area, or a TRAP interrupt without a handler. I'd possibly suspect
the latter. |
|
|
bryant@balancebitsconsult
Joined: 21 Nov 2023
Posts: 38
|
|
| Posted: Wed Mar 13, 2024 11:48 am |
|
|
| We are saying the same things, differently. |
|
|
gaugeguy
Joined: 05 Apr 2011
Posts: 303
|
|
| Posted: Wed Mar 13, 2024 1:22 pm |
|
|
| NOINIT means do not initialize and do turn on the peripheral. It has nothing to do with interrupts at all. |
|
|
Ttelmah
Joined: 11 Mar 2010
Posts: 19539
|
|
| Posted: Thu Mar 14, 2024 1:40 am |
|
|
Bryant, you said:
| Quote: |
the enable interrupts call is redundant since my #use block has already turned the ones related to serial on.
|
That is totally untrue.
The interrupts are not enabled till you turn them on.
A disabled interrupt can become set, but it does not call the handler.
If the handler is being called, something else is going on. I suspect you
have one of possibly a dozen faults, possibly with the hardware.
You can get spurious jumps if the decoupling is not adequate.
Or as I said you might be incorrectly executing a return without a matching
call, or any one of a number of other hardware or software faults.
One classic that can cause an issue on the PIC, is EMI. |
|
|
Ttelmah
Joined: 11 Mar 2010
Posts: 19539
|
|
| Posted: Sat Mar 16, 2024 12:32 pm |
|
|
As a suggestion, add to the start of your iNR_RDA handler, the line:
if(interrupt_active(INT _RDA))
delay_cycles(1);
Put your breakpoint on the delay.
I suspect you will find he routine is arriving here without the interrupt
actually being set. Proof that it is getting here some other way. |
|
|
|
|
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum
|
Powered by phpBB © 2001, 2005 phpBB Group
|
FAQ 
Search
Register
Profile 
Log in to check your private messages 
Log in
