MPLAB SIM & ICD2 get different result

[cindyw]

Posted: Thu Mar 03, 2005 2:31 am



Here is a testing code. I found the PORTA is correct by running MPLAB SIM , however, the actual test result by run MPLAB ICD2, PinPA5 always on(PORTA=0x20)

compiler version 3.188

Anyboday can help on this?

-----------code start----------

#include <16F676.h>
#DEVICE adc=8 //only use 8 bit ??

#fuses INTRC_IO,NOWDT,NOPROTECT//NOBROWNOUT //MCLR
#use delay (clock=4000000) //4M hz internal osc
//#endif


void init();
/************************************************************
Main program
*************************************************************/
void main() { int temp;

init(); // init for ADC, I/O, Timer
CHANNEL_1: output_high(PIN_A4 );

temp=read_adc(); //read noise
output_low(PIN_A4 ); //RA4--Status_LED
delay_us(100);
delay_us(100);


goto CHANNEL_1;


}

void init()
{
//----I/O initialization---

#asm
bcf 0x03,0x5; //bank 0
clrf 0x05; //;Init PORTA
clrf 0x07; //;Init PORTC

bsf 0x03,0x5; //bank 1


movlw 0x08;
movwf 0x85 ;//set TRISA RA0-2,RA4,5 as output, system not recognize TRISA, use address instead
movlw 0x0f;
movwf 0x87 ;//set TRISC Rc4,5 as output,other as input
#endasm
//----ADC initialization---
SETUP_ADC_PORTS(sAN4|sAN5|sAN6|sAN7|VSS_VDD); // AN4:7 A/D. others digital
setup_adc( ADC_CLOCK_INTERNAL ); //internal oscillator 2-6us
set_adc_channel(5);
SETUP_COMPARATOR(NC_NC_NC_NC); //
//----Timer initialization---
SETUP_TIMER_1(T1_INTERNAL|T1_DISABLED|T1_DIV_BY_8); //prescale=1:8,
}


---code end

[PCM programmer]

[cindyw]

PCM programmer,
You are so kind to help on this.

let's me answer your questions first.
1, there is LED connect to RA5 (pin 2) , I can see its on/off.
2, there is also LED connect to RA4 (pin 3) ,
3, LED circirt (same): MicroP pin-1k Resistor-LED-GND
4, I use tektronix scope to check voltage
5, I run my program step-by-step. It is running because I can see RA4 is blinking.
6, +5V supply
7, your sample code is working great.

Here is my new findings
1, my code is working well in continues run mode, RA5 turn on only when I Halt runing or when program run in step by step mode. Based on this I suppose it will run properly in programmer mode.
2, I remove I/O initialization portion in init{}, then it works well

So, does this means I/O not need initialization? output_high() is enough? or how to do correct initilization?


_________________
CC

[PCM programmer]

Because the simple program that I posted is working, I suggest that
you use that program as a base, and then add new code to it.

What feature do you want to add to your program next ?
If you tell me, then I'll describe the most simple way to do it.
If everything is kept very simple, there is a better chance that
it will work.

[Guest]

actually I have some more working code. This testing code is just for verify the timing to run a read_adc() function.

My testing can be done, only accidently found above mentioned problem. I am wondering whether same problem happened somewhere while I am not aware .
I need to do 3 contiues ADC reading at interval 32us by using internal 4 MHz oscillator. Each reading compare with threshold. My test result shows it needs 44us to do ADC only in debugging mode(including on-off port RA4). I did not do testing in programmer mode yet. It sounds timing very tight. The code is here.

[PCM programmer]

If you need to do three A/D readings quickly, then get rid of the for(;;)
loop. The for(;;) loop takes several instructions and at 4 MHz, each
instruction is 1 or 2 microseconds. Just repeat the A/D code 3 times.
It takes more ROM, but it's faster without the loop.

[Douglas Kennedy]

This may be the issue
Looking at your code

CHANNEL_1: output_high(PIN_A4 );

temp=read_adc(); //read noise
output_low(PIN_A4 ); //RA4--Status_LED
delay_us(100);
delay_us(100);


goto CHANNEL_1;

Notice there is no delay between the high and the low on pin_a4 except for the read_adc() action time.
Now when you step through there is time to see the transition from high to low when its running it will go low but the high time is very very short so it will not be seen with the human eye.

[cindyw]

Hi, Thanks for all of your advise.
_________________
CC