 |
 |
| View previous topic :: View next topic |
| Author |
Message |
joergn
Joined: 15 Jul 2010
Posts: 15
Location: Hamburg Germany
|
|
 Posted: Tue Sep 13, 2016 3:44 am |
 |
|
Thanks for the hint, I wanted to create extra delay, because of the bad samples I am wondering about, meanwhile I found inside the datasheet I have to wait 200us befor each internal tempread conversion, which I think is a quite big delay for my application.
However my assembly below (5.062) looks different to yours, let's hope that in your code the assembly is not always that different!?
I was not aware that a read_adc(ADC_READ_ONLY)
includes also the test for the go bit - for my language understanding
"only" should mean only like there is only one max or min value so there should be just a read with no other commands, but you are right
there are two assembly lines extra 
| Code: | .................... read_adc(ADC_START_ONLY);
057A: MOVLB 01
057B: BSF 1D.1
.................... while (bit_test(*ADCON0, 1));
057C: BTFSC 1D.1
057D: GOTO 57C
.................... dummy=read_adc(ADC_READ_ONLY);
057E: BTFSC 1D.1
057F: GOTO 57E
0580: MOVF 1C,W
0581: MOVWF 7A
0582: MOVF 1B,W
0583: MOVLB 00
0584: MOVWF 39
0585: MOVF 7A,W
0586: MOVWF 3A
.................... |
|
|
 |
joergn
Joined: 15 Jul 2010
Posts: 15
Location: Hamburg Germany
|
|
| Posted: Wed Sep 14, 2016 4:38 am |
|
|
I have now on this 16F1765 the internal temp sensor on a first DUT in action and need to check the accuracy and tolerance on other chips.
I calibrated it at 0°C and 50°C manually.
It resolves 1°C very well - so very good option for me 
It was not possible to use the 4 diodes mode at all, which is set by TEMPERATURE_RANGE_HIGH
The unit just delivered bad ADC values no matter how long time was waited. I wonder why there is no error of this unit noted inside the datasheet?
I started with the HIGH option, because I expected a better ADC resolution and better performance out of it.
I think the internal current regulator of this module can simply not regulate it out. If this is the case I expect this as a possible problem also on other chips?
I have a MAX15007 5V reg as the supply.
Have you tested the TEMPERATURE_RANGE_LOW on the PIC10F322
also?
| Code: |
signed int16 pcbtemper;
int8 pcbc=0;
int16 cstor,dummy,pcbbuf;
#separate
void pcbtemp(void)
{
setup_vref(TEMPERATURE_INDICATOR_ENABLED|TEMPERATURE_RANGE_LOW);
set_adc_channel(TEMPERATURE_INDICATOR);
delay_us(200); // is important if set less the result changes
// depending on the load of the chip
read_adc(ADC_START_ONLY);
dummy=read_adc(ADC_READ_ONLY);
setup_vref(VREF_OFF);
pcbbuf+=dummy;
if(pcbc>16)
{
pcbc=0;
cstor=pcbbuf;
pcbbuf=0;
}
pcbc++;
}
// inside main loop:
pcbtemp();
if(pcbc==1)
{
cstor=((cstor/10)*11)/10;
pcbtemper=cstor-1385;
}
|
|
|
|
Ttelmah
Joined: 11 Mar 2010
Posts: 19552
|
|
| Posted: Wed Sep 14, 2016 8:45 am |
|
|
I still have to ask why you are fiddling around starting separately.
Just use:
dummy=read_adc();
This will start, wait, and read, without anything else being needed.
Also, 'think' about your values and maths. You are summing 16 readings, then dividing by 10, multiplying by 11, and dividing by ten again. What a lot of fiddling....
Gives effectively *1.76.
Try sampling twenty eight times instead, summing into an int32, and then dividing the result by 16. You don't even need dummy at all. Just use:
pcbbuf+=read_adc();
You will find your code size should go down, and the result will probably be better. |
|
|
|
|
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
