 |
 |
| View previous topic :: View next topic |
| Author |
Message |
rvalor
Joined: 16 Apr 2009
Posts: 18
Location: Spain
|
| PIC 18F67J50, CCS 4.104, Strange problem with A/D converter |
 Posted: Fri Jan 28, 2011 3:24 am |
 |
|
Hello everybody,
I'm using a PIC18F67J50 to measure temperature using thermistors.
First of all I calibrate the AD (a functionality of this PIC AD converter...)
My AD converter configuration is this:
Avdd -> 3.3V (general system power...)
Avss -> GND
Vref+ -> 3.0V (from a voltage reference...)
I read 4 thermistor (using some AmpOp...) using as references 3.0V and GND.
I read the BAT level using as references VDD and GND.
When working, the device allways do the same, each minute, it reads all channels, store the readings in EEPROM and transmits them, no more. But suddenly, after some hours of work, something goes wrong and three channels (battery included) don't work.
Battery readings become 0V. But when this error occurs, if I test the PCB I find it ok. Power is OK and voltage reference too. If I measure the battery input pin I see a correct Voltage...
¿Have everybody experienced some problem like this whith the AD converter??
This is a sample of my AD configuration:
| Code: |
//In the first time, all values to 0, in order to acces ADCON0, ADCON1 I must switch the bit ADSHR in WDTCON register....
WDTCON_ADSHR = 0;
ADCON0 = 0;
ADCON1 = 0;
WDTCON_ADSHR = 1;
ANCON0 = 0;
ANCON1 = 0;
//------------------
// channel config
//------------------
WDTCON_ADSHR = 0; //normal access...
ADCON0_ADON = 0; //ADC off
ADCON1_ADFM = 1; //left justified...
ADCON1_ADCAL = 0; //normal conversion (not calibration...)
WDTCON_ADSHR = 1; //Acceso registros compartidos...
ANCON0 = 159; //Channels AN0, AN1, AN2, AN3(Vref+), AN4, AN7, AN10 y AN11 configured as analogs
ANCON1 = 12;
WDTCON_ADSHR = 0; //normal access...
ADCON0_VCFG1 = 0; //VSS
ADCON0_VCFG0 = cREF; //VDD o VRef+ (it depends on the channel...)
//channel selection:
ADCON0 &= CH_MASK; //I erase only the channel bits
ADCON0 |= cCHAN_SELECTED; //Now I set the appropiate channel...
delay_us(50);
ADCON1_ACQT2=0;
ADCON1_ACQT1=0; //Adquisition time --> 0Tad
ADCON1_ACQT0=0;
ADCON1_ADCS2=0;
ADCON1_ADCS1=0;
ADCON1_ADCS0=1;
ADCON0_ADON = 1; //ADC ON
delay_us(50);
//and now I start the conversion:
fMeasure = (float)read_adc(ADC_START_AND_READ);
|
I repeat this configuration for each channel changing the references and selected channel.
Thanks you in advance.... |
|
 |
newguy
Joined: 24 Jun 2004
Posts: 1909
|
| Re: PIC 18F67J50, CCS 4.104, Strange problem with A/D conver |
| Posted: Fri Jan 28, 2011 7:56 am |
|
|
| rvalor wrote: | | When working, the device allways do the same, each minute, it reads all channels, store the readings in EEPROM and transmits them, no more. But suddenly, after some hours of work, something goes wrong and three channels (battery included) don't work. |
Are you storing these values in the PIC's internal EEPROM? If so, that could be part of the problem, as you've likely worn it out. |
|
|
rvalor
Joined: 16 Apr 2009
Posts: 18
Location: Spain
|
|
| Posted: Fri Jan 28, 2011 8:03 am |
|
|
Hello newguy.
I'm not storing data in the internal PIC memory. I use a 24LC512 external memory.
When data must be stored I disable all the interrupts in order to avoid problems in the timing of the I2C communication.
Do you think it could be a problem related with the impedance of the circuits connected to the analog to digital channels?
In the other hand I'm monitoring some variables in order to prevent they become corrupted...
I'm a bit confused, the device works fine for a lot of hours and the problem appears suddenly.
Thanks you. |
|
|
SherpaDoug
Joined: 07 Sep 2003
Posts: 1640
Location: Cape Cod Mass USA
|
|
| Posted: Fri Jan 28, 2011 1:05 pm |
|
|
The 24LC512 datasheet says it is good for 1 million cycles. At 1 cycle per minute that is about 2 years. That is enough for a home project, but kind of limiting if this is a product to be sold.
Can you store the last few readings in RAM and check them when you detect a failure? Or maybe when you find a failure read a few more times and store them in a new section of EEPROM to be dumped for diagnosis.
Another thought... Are you using a large ring buffer that maybe does not wrap properly when it fills up? What is the buffer memory address where things start to go bad?
_________________
The search for better is endless. Instead simply find very good and get the job done. |
|
|
temtronic
Joined: 01 Jul 2010
Posts: 9241
Location: Greensville,Ontario
|
|
| Posted: Fri Jan 28, 2011 1:26 pm |
|
|
Any chance the device that receives the data is 'going to sleep' or otherwise failing?
If I assume you're transmitting to a PC, is it a minimal system. No antivirus s/w, no Internet connection, etc. Just a simple terminal program running.
Hmm..using real RS232 or a USB<>RS232 adapter ?
USB has a lot of problems compared to 'legacy' comports.
Perhaps you can add a 'reading counter' to go with the data. See if it always fails at nnnnn counts, or is random. If you could attach an LCD to the PIC to display the count and data...
..even an LED that toggles every transmission would help. |
|
|
|
|
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
