PIC ADC

[jonnylazer]

Hi all,
I hope someone is able to help me. Im using the ADC on a PIC16f916 as part of my final year project. At the moment im just trying to sample one channel and send the 8 bit result to 8 LED's connected to channel B. The problem is two fold...even if I ground the input channel I still get a reading of binary 1 which equates to 19.7mV. The major problem is that all of my readings (from the LED's) seem to be around 80mV to high when compared to my voltmeter readings. BTW im using a bench power supply to vary my voltage.

ill post my code if anyone wants to help

thanks for reading and I hope you can help

[PCM programmer]

Post a short test program. Be sure to post all the #include, #fuses,
and #use statements. Post all the variable declarations. You should
be able to do this with about 10 lines of code.

[jonnylazer]

here is the part of my program that is relevant to my prob...

#INCLUDE <16f916.h>
#DEVICE ADC=8
#DEFINE DEVICEOSC 4000000
#fuses HS,NOWDT,PUT,NOPROTECT,BROWNOUT,
#use delay(clock=DEVICEOSC)



void main(void) {
int value;

// set up
setup_adc( ADC_CLOCK_DIV_8); // the adc
setup_adc_ports(ALL_ANALOG);
set_adc_channel( 0 );

do {

delay_us(30);
value = read_ADC();
output_b(value);


}
while(true);
}

[SherpaDoug]

How are you using your voltmeter? Measure from the PIC analog ground pin to the A/D channel pin. If possible put your probe on the pin where it leaves the package, before it enters a solder joint or socket pin.
_________________
The search for better is endless. Instead simply find very good and get the job done.

[PCM programmer]

I modified your code slightly and programmed it into a 16F917, which
is a similar PIC. I tested it on a on a PicDem2-Plus board, which has
a 5K trimpot connected to the AN0 pin. I turned the trimpot down to
zero ohms and the A/D reads 0. So it's working correctly. This is with
compiler vs. 4.066. Here's the output:

[PCM programmer]

[jonnylazer]

Thanks for the input, unfortunately I dont have trimpot at hand. I will have to 'borrow' one from Uni in the morning. So, judging by the fact that it worked for you I assume any problem I have is a hardware one?

I will post again tomorrow when ive tried your suggestion.

cheers mate.

[Guest]

I would take a look at your ground paths. Make sure all of your circuit grounds go to a single node. Without seeing your circuit and layout its tough to tell if it could be hardware, but it sounds like your PIC thinks that the A/D converter port is not going to ground when you connect it to ground. Also check your bypassing; these fast ADCs need a stiff supply.

[PCM programmer]

He sent me a PM that he put in the trimpot and it worked OK.
His problem is solved.

[jonnylazer]

Thanks PCM....should have posted my PM here. Anyway, thanks again.

[jonnylazer]

One final question PCM....why add the 'DELAY_ms(500)' to the routine....is it RS232 ??

[SherpaDoug]

I would guess that is so your eyes can focus on each reading before it scrolls off the edge of the screen.
_________________
The search for better is endless. Instead simply find very good and get the job done.