| View previous topic :: View next topic |
| Author |
Message |
rafa2
Guest
|
| ADC of PIC18F8720 |
 Posted: Wed Mar 16, 2005 12:13 pm |
 |
|
Hi,
I have a doubt about the ADC of a PIC18F8720. There are a little offset between the real signal and the value measured, for example if the voltage at input is 1.414V, the measure is 1.504. The precision is 1/2 bit = 4/1024/2=2mV.
The configuration is: Vref+ = external ref = 4V (from a zener); Vref- = internal ref = VSS. Also, the micro have VDD = 5V, but the ICD2 only delivery 4.74V. If I change the configuration and the references are internal (VSS and VDD) the results are worse.
My code is:
| Code: | #define ANALOG_AN0_TO_AN5 0x09
#define VSS_VDD 0x00
#define VSS_VREF 0x10
#define ADC_CLOCK_DIV_32 0x102
…
#include <18f8720.h>
#device *=16 ADC=10
#FUSES HS,NOWDT,NOPROTECT,PUT,BROWNOUT,NOLVP
#ID CHECKSUM
#ZERO_RAM
#use DELAY(CLOCK=20000000)
#use RS232 (BAUD=9600, XMIT=PIN_C6, RCV=PIN_C7)
void main (void){
int16 val;
…
set_tris_a(0xFF);
setup_port_a( ANALOG_AN0_TO_AN5 | VSS_VREF);
// setup_port_a( ANALOG_AN0_TO_AN5 | VSS_VDD);
setup_adc( ADC_CLOCK_DIV_32 );
…
while(1){
set_adc_channel( 4 );
delay_ms(10);
val=read_adc();
delay_ms(10);
printf(displays,"%Lx",val);
printf("%2.3f",(float)val*3.99/1024);
// printf("%2.3f",(float)val*4.74/1024);
…
}
} |
Some idea??? |
|
 |
DragonPIC
Joined: 11 Nov 2003
Posts: 118
|
| printf("%2.3f",(float)val*3.99/1024); |
| Posted: Wed Mar 16, 2005 1:15 pm |
|
|
There have been some post earlier that claim problems with floating points with prontf(); look in some recent posts.
http://www.ccsinfo.com/forum/viewtopic.php?t=22210
If this is how you are reading your answer, beware.
_________________
-Matt |
|
|
bluetooth
Joined: 08 Jan 2005
Posts: 74
|
|
| Posted: Wed Mar 16, 2005 4:22 pm |
|
|
What is the int16 val when you have 1.414v input? Is it what you expect? With a 4v reference, I think you should read 362. I guess I don't exactly understand why your resolution ("precision") isn't 4/1024 or 3.9mV.
Verify this before you worry about the (potential) printf/float issue.... |
|
|
rafa2
Guest
|
|
| Posted: Wed Mar 16, 2005 4:29 pm |
|
|
| I think that the problem is the converter. In the code, I can see the measure directly from a converter (printf("%Lx",val); ), and for 1.4V it should be: 1.414*1023/4=361=169h, and I read 182h. |
|
|
PCM programmer
Joined: 06 Sep 2003
Posts: 21708
|
|
| Posted: Wed Mar 16, 2005 4:40 pm |
|
|
| Quote: | | external ref = 4V (from a zener) |
You are trying to get 4 digits of precision in your result, but your
posted zener voltage only has one digit.
Also if I go to http://www.digikey.com and type in Zener Diode
and look at the available voltages, I see 3.9v and 4.3v, but no 4.0v.
In addition to that problem, the typical tolerance is 5%.
What is your true Vref voltage, as measured with a voltmeter ? |
|
|
bluetooth
Joined: 08 Jan 2005
Posts: 74
|
|
| Posted: Wed Mar 16, 2005 4:41 pm |
|
|
Good info... now you might look at the input and your supplies with a 'scope to make sure they're steady and noise free, and make sure you're driving the input with an impedance that's < 2.5k ohms...
Adding:
PCM is probably on it... with a 3.9V zener you'll read 371, if it's exactly 3.9V... the tolerance is probably the rest of the error.....
Last edited by bluetooth on Wed Mar 16, 2005 4:47 pm; edited 1 time in total |
|
|
rafa2
Guest
|
|
| Posted: Wed Mar 16, 2005 4:46 pm |
|
|
| Perdon, it is a regulator not a zener, with a voltmeter 3.99V. |
|
|
rafa2
Guest
|
|
| Posted: Wed Mar 16, 2005 4:50 pm |
|
|
| And the input have a op-amp... |
|
|
bluetooth
Joined: 08 Jan 2005
Posts: 74
|
|
| Posted: Wed Mar 16, 2005 5:01 pm |
|
|
Rafa2:
You're killing all the theories!!!!!
OK - silly question, but are AVss and AVdd hooked up via a low impedance path? And are all signals stable and quiet? It starts to sound more like a power/ground problem....
Happy hunting.... |
|
|
PCM programmer
Joined: 06 Sep 2003
Posts: 21708
|
|
| Posted: Wed Mar 16, 2005 5:09 pm |
|
|
What's your version of the compiler ?
Post the portion of the .LST file that shows the ASM code for the lines
given below. It's possible that the ASM code for the A/D functions is
incorrect. If you post it, maybe we can see the problem.
set_tris_a(0xFF);
setup_port_a( ANALOG_AN0_TO_AN5 | VSS_VREF);
setup_adc( ADC_CLOCK_DIV_32 );
while(1){
set_adc_channel( 4 );
delay_ms(10);
val=read_adc(); |
|
|
rafa2
Guest
|
|
| Posted: Wed Mar 16, 2005 5:23 pm |
|
|
I am sorry for your theories, bluetooth... 
About the power supply, unless the ICD2 only delivery 4.7V
The version of the compiler is 3.180
LST:
| Code: | .................... set_tris_c(0x10);
0702: MOVLW 10
0704: MOVWF F94
.................... set_tris_a(0xFF);
0706: MOVLW FF
0708: MOVWF F92
.................... setup_port_a( ANALOG_AN0_TO_AN5 | VSS_VREF);
070A: MOVF FC1,W
070C: ANDLW C0
070E: IORLW 19
0710: MOVWF FC1
.................... // setup_port_a( ANALOG_AN0_TO_AN5 | VSS_VDD);
.................... setup_adc( ADC_CLOCK_DIV_32 );
0712: BCF FC0.0
0714: BSF FC0.1
0716: BCF FC0.2
0718: BSF FC0.7
071A: BSF FC2.0
.................... // ADCON0 = 0x11; // set up adc, timer/32, and adc on
.................... // ADCON1 = 0x00; // right justified result, a0,1,4 are analog, a2,3 are references, e0,e1,e2 are digital
.................... // ADCON2 = 0x82;
....................
.................... set_adc_channel( 4 );
071C: MOVLW 10
071E: MOVWF 01
0720: MOVF FC2,W
0722: ANDLW C3
0724: IORWF 01,W
0726: MOVWF FC2
.................... LCD_E = 1; // enable LCD
0728: BSF F82.2
.................... disp_lcd = TRUE;
072A: BSF 05.0
.................... disp_serial = false;
072C: BCF 05.1
.................... init_lcd();
072E: BRA 007C
.................... while(1){
.................... set_adc_channel( 4 );
0730: MOVLW 10
0732: MOVWF 01
0734: MOVF FC2,W
0736: ANDLW C3
0738: IORWF 01,W
073A: MOVWF FC2
.................... delay_ms(10);
073C: MOVLW 0A
073E: MOVWF 28
0740: RCALL 0004
.................... val=read_adc();
0742: BSF FC2.1
0744: BTFSC FC2.1
0746: BRA 0744
0748: MOVFF FC3,06
074C: MOVFF FC4,07
.................... delay_ms(10);
0750: MOVLW 0A
0752: MOVWF 28
0754: RCALL 0004 |
Rafa |
|
|
bluetooth
Joined: 08 Jan 2005
Posts: 74
|
|
| Posted: Wed Mar 16, 2005 5:36 pm |
|
|
Rafa2:
Looks like those setups are good - and since you're using an external VREF, the ICD power shouldn't matter.... so, back to my original questions: are AVss and AVdd connected via low impedance paths, and are all the signals stable and quiet when looked at on a 'scope? |
|
|
|
FAQ 
Search
Register
Profile 
Log in to check your private messages 
Log in
