 |
 |
| View previous topic :: View next topic |
| Author |
Message |
Brooksy
Joined: 02 Jun 2008
Posts: 2
Location: Princeton NJ
|
| read_adc() taking 2.5ms, 1000x too long |
 Posted: Mon Jun 02, 2008 8:54 pm |
 |
|
Hi All,
I run the following code and I'm finding that the call to read_adc() is taking 2.5 ms. This is almost a thousand times longer than all sources of info tell me it should take. Even when I run the debugger MPLAB SIM it tells me this should take 3 to 4 us. I've tried the ADC pin tied to ground, to 5V, floating, it seems to have no effect.
Can anyone help me out here?
Thanks!
| Code: |
#include <16F688.h>
#device adc=8
#fuses HS,NOWDT,NOPROTECT,PUT
#include <stdlib.h>
#use delay(clock=20000000)
#use rs232(baud=9600, xmit=PIN_C4, rcv=PIN_C5)
#define TRIGGER PIN_A1
void main()
{
int8 v1, thedata[50], i;
int1 time_old;
setup_oscillator(OSC_NORMAL); //Will actually drive from Master's clock
//setup ports
set_tris_a(0B00000001); //A0 input TIMECLOCK, all others output
set_tris_c(0b00110111); //C0 to C2 are ADC, C4 and C5 are RS232 comm
setup_comparator(NC_NC_NC_NC);
//set ADC
setup_adc(ADC_CLOCK_DIV_2); // originally tried _DIV_32
setup_adc_ports(sAN6); //pin C2 as primary ADC
set_adc_channel(6);
delay_ms(20);
while (TRUE)
{
output_high(TRIGGER);
v1 = read_adc(); //Square wave is high for 2.5ms, low for 2.5ms
output_low(TRIGGER);
v1 = read_adc();
}
//also tried reading 200 times between setting the pin high and low.
//this gave about 1/2 second between going high and low - could time
//with a stopwatch.
}
|
|
|
 |
PCM programmer
Joined: 06 Sep 2003
Posts: 21708
|
|
| Posted: Mon Jun 02, 2008 9:23 pm |
|
|
1. What is your compiler version ?
2. Are you testing this in actual hardware ?
3. Describe the external circuits, including component values, on
all i/o pins that you're using, and also on the oscillator pins and the
MCLR pin. |
|
|
Ttelmah
Guest
|
|
| Posted: Tue Jun 03, 2008 2:40 am |
|
|
Before blaming the ADC, try a simple test:
| Code: |
while (TRUE)
{
output_high(TRIGGER);
delay_us(50);
output_low(TRIGGER);
delay_us(50);
}
|
A PCM programmer says, 'hardware details'.
Realistically, even at it's slowest clock, the ADC cannot take 2.5mSec, if your master clock _is_ 20MHz. The only way it can take this time, is that your master clock is actually vastly slower than you think. The 'loop', should involve about 20 instruction times for each ADC conversion, another couple to read each value, and perhaps four more to actually loop. If the loop is taking 5mSec, then it implies the master oscillator is actually running at perhaps 10000Hz, not 20MHz. So, test it with a simple loop like that shown. If this gives an output wavefrm at about 10KHz, then your oscillator is OK, If though it runs a lot slower, then you have found the 'real' problem.
Best Wishes |
|
|
Brooksy
Joined: 02 Jun 2008
Posts: 2
Location: Princeton NJ
|
| Problem Solved - Was Oscillator Configuration |
| Posted: Tue Jun 03, 2008 5:47 am |
|
|
Thanks for the responses. I just tried something and it fixed the problem.
This was running on actual hardware. The square wave was observed on an oscilloscope. The thing I had been trying for the first time was running this PIC's clock from an oscillator on another PIC. The master PIC had the 20MHz oscillator and I connected the OSC2 pin on that chip to the OSC1 pin on the PIC in question. I just found out about this possibility on another post. This seemed to work fine, even doing printf() over RS232 the timing worked fine.
When I pulled out the single wire to OSC1 and replaced it with the 20Mhz oscillator itself, the usual configuration, the 2.5ms square wave generated with the code above became a 3us square wave. So I'm back in business. |
|
|
|
|
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
