| View previous topic :: View next topic |
| Author |
Message |
cerr
Joined: 10 Feb 2011
Posts: 241
Location: Vancouver, BC
|
| reading odd 12bit adc value |
 Posted: Wed Aug 03, 2011 3:47 pm |
 |
|
Hi There,
I'm using an pic18f86k22 which features 12bit adcs.
Now in my program, i right align the values like
| Code: | setup_adc_ports (AMBIENT|TEMPFB_OVR|TEMPFB_LPR|POT1|POT2);
setup_adc (ADC_CLOCK_INTERNAL);
ADCON2|=0x80; |
which seems to work fine for me but the problem I'm having is that i read an adc value of 4095 on an analog input where i instead should be reading a lower value. I however had
in my code header. I now changed this to adc=12 - is that gonna resolve the issue? Any ideas or hints? I need to into the temperature chamber before i can test my code which is connected with somewhat lots of efforts. I would like to provide a code that works this time thus any hints or suggestions would be welcome!
As for the electronics on my analog input:
I have a NTC 10k connected to GND and a 10K pullup to +Vcc (+5V), thus the adc value should never be maxed out (unless the ntc falls off)...
Thank you very much!
Ron |
|
 |
PCM programmer
Joined: 06 Sep 2003
Posts: 21708
|
|
| Posted: Wed Aug 03, 2011 4:02 pm |
|
|
| Quote: |
setup_adc_ports (AMBIENT|TEMPFB_OVR|TEMPFB_LPR|POT1|POT2);
|
Only you know what these are.
| Quote: |
setup_adc (ADC_CLOCK_INTERNAL);
|
For years, we have said don't do this unless you are doing the A/D
conversion while in sleep mode.
No one knows what this is. |
|
|
cerr
Joined: 10 Feb 2011
Posts: 241
Location: Vancouver, BC
|
|
| Posted: Wed Aug 03, 2011 4:30 pm |
|
|
Sorry, here the defines:
| Code: | #define TEMPFB_OVR sAN7
#define TEMPFB_LPR sAN14
#define AMBIENT sAN21
#define POT1 sAN22
#define POT2 sAN23 |
... For years? Sorry, haven't been using this compiler that long myself. But I found a good explanation here:
ADC_CLOCK_INTERNAL - answer here is 'read the data sheet'. The internal clock is an RC oscillator at 'about' 500KHz. This is asynchronous to the master clock. Problem is that if the CPU clock is running much faster than the ADC clock, you can get varying numbers of cycles of the CPU clock, in each cycle of the ADC clock. Result varying amounts of noise. So the internal RC clock, is _not_ recommended for use above 1MHz CPU clock, _unless_ you stop the CPU (use sleep) while doing the ADC conversion.
That makes sense so i'll be using ADC_CLOCK_DIV_32 - here but that doesn't explain the 4095 that i'm getting at -10C already (NTC goes till - 40C - at the least...)
| Code: |
#byte ADCON2 = getenv("sfr:ADCON2")
ADCON2|=0x80;
|
From datasheet:
ADCON2: A/D CONTROL REGISTER 2
bit 7 ADFM:A/D Result Format Select bit
1 = Right justified
0 = Left justified |
|
|
PCM programmer
Joined: 06 Sep 2003
Posts: 21708
|
|
| Posted: Wed Aug 03, 2011 5:10 pm |
|
|
| Quote: | | here but that doesn't explain the 4095 that i'm getting at -10C already (NTC goes till - 40C - at the least...) |
Measure the input voltage to the PIC's ADC pin at that temperature
with a voltmeter. What does it show ?
Post the lines of code that setup, read, and display the ADC result.
Also post the declarations of any variables or #define's used in those lines.
Also post your compiler version. |
|
|
cerr
Joined: 10 Feb 2011
Posts: 241
Location: Vancouver, BC
|
|
| Posted: Wed Aug 03, 2011 5:43 pm |
|
|
Yep,
I'll go down thereto measure on my engineering unit tomorrow morning. Can't do that today anymore...
However, here's some of the code:
| Code: |
#device adc=12 // this used to be adc= 16 at the time the bug appeared
#define TEMPFB_OVR sAN7
#define TEMPFB_LPR sAN14
#define AMBIENT sAN21
#define POT1 sAN22
#define POT2 sAN23
#byte ADCON2 = getenv("sfr:ADCON2")
#define ADCCH_TEMP_LPR 14
#define ADCCH_TEMP_OVR 7
//setup adc ports
setup_adc_ports (AMBIENT|TEMPFB_OVR|TEMPFB_LPR|POT1|POT2);
setup_adc (ADC_CLOCK_DIV_32);
ADCON2|=0x80;
set_adc_channel (ADCCH_TEMP_LPR) ;
delay_us (10) ;
fprintf(PC, "LPR temperature %Ld\r\n",read_adc());
set_adc_channel (ADCCH_TEMP_OVR) ;
delay_us (10) ;
fprintf(PC, "OVR temperature %Ld\r\n",read_adc());
|
Thanks, for your assistance.
PS: Compiler version: 4.119 |
|
|
FvM
Joined: 27 Aug 2008
Posts: 2337
Location: Germany
|
|
| Posted: Thu Aug 04, 2011 3:52 am |
|
|
| If you look at the generated code, you'll realize that #device adc=12 in fact sets right justify, while adc=16 sets left justify. But you're overriding it, so there's no difference. Getting a result of 4095 seems to indicate correct right justify but some other kind of wrong setup or wrong pin connection. |
|
|
RF_Developer
Joined: 07 Feb 2011
Posts: 839
|
|
| Posted: Thu Aug 04, 2011 4:45 am |
|
|
You don't appear to be setting the reference voltage up at all. You need to include it in setup_adc_ports.
| Code: |
setup_adc_ports(AMBIENT|TEMPFB_OVR|TEMPFB_LPR|POT1|POT2|VSS_VDD);
|
or whatever is correct for your ADC reference wiring. Look at the .h for your PIC device to see what options are available.
RF Developer |
|
|
cerr
Joined: 10 Feb 2011
Posts: 241
Location: Vancouver, BC
|
|
| Posted: Thu Aug 04, 2011 9:32 am |
|
|
| RF_Developer wrote: | You don't appear to be setting the reference voltage up at all. You need to include it in setup_adc_ports.
| Code: |
setup_adc_ports(AMBIENT|TEMPFB_OVR|TEMPFB_LPR|POT1|POT2|VSS_VDD);
|
or whatever is correct for your ADC reference wiring. Look at the .h for your PIC device to see what options are available.
RF Developer |
I got the following available:
| Code: | // Optional Second argument:
#define __VDD 0x00 // Range 0-Vdd
#define VSS_VREF 0x10 // Range 0-VrefH
#define VSS_2V048 0x20 // Range 0-2.048V
#define VSS_4V096 0x30 // Range 0-4.096V
#define VREF_VDD 0x08 // Range VrefL-Vdd
#define VREF_VREF 0x18 // Range VrefL-VrefH
#define VREF_2V048 0x28 // Range VrefL-2.048V
#define VREF_4V096 0x38 // Range Vrefl-4.096V |
and it says second argument so i assume i wouldn't OR it to the analog pins but place a comma something like
| Code: | | setup_adc_ports (AMBIENT|TEMPFB_OVR|TEMPFB_LPR|POT1|POT2, __VDD); |
- which i also assume would be the default.., right? (scanned through the datasheet but didn't find it right-away and have to get ready for the temperature chamber with my dev board now...) |
|
|
cerr
Joined: 10 Feb 2011
Posts: 241
Location: Vancouver, BC
|
| [SOLVED] reading odd 12bit adc value |
| Posted: Thu Aug 04, 2011 12:23 pm |
|
|
Seems like the changes I applied fixed the issue. I think it probably was the wrong directive, you guys agree?
Thanks for the help
RF_Dev, FvM & PCM |
|
|
|
FAQ 
Search
Register
Profile 
Log in to check your private messages 
Log in
