Clarification for PIC16F690

[Arakel]

Good day! I decided to buy 2 pieces from this pic with a Pickit 3 and a debugger. I want to use 10 bits one the ADC and this PIC should have a 10bit ADC. So when I try to use the functions:

[Ttelmah]

Of course it will.....

This is the ADC _clock_ setting. The ADC needs to be fed with a clock that is typically 500KHz or slightly slower. Look at the chip's data sheet. Look at what divisors can be specified for the clock (/4, /8 /16 /32 and /64 typically). This is nothing to do with the ADC resolution.

You can set the value returned by the adc functions, at the top of the code (line after the include for the processor)

#device ADC=8 //specifies to return an 8 bit value from the ADC
#device ADC=10 //same for 10bit
#device ADC=12 //same for chips with a 12bit ADC
#device ADC=16 //returns the value left justified to 16 bits

Choose the setting you want.

Remember you need to be using an int16 variable to handle the results over 8bits.

[Arakel]

Device = 10 is already set.

I will try with DIV_16. Since device is 10 it will only read the first 10 bits?

I am sorry for the stupid question, I just am too tired sometimes and I rush.
_________________
Yo! I love learning and technology! I just do not have experience so do not be angry if I ask a stupid question about a detail! From so much to remember sometimes I forget the details in order to remember the big problems!

[jeremiah]

[Ttelmah]

Also don't just 'try' clock settings. Read the data sheet. There is a table of what divisors are recommended for each clock rate. Use the right setting for your clock rate.

The syntax is not #device = 10. Wants #device ADC=10

[Arakel]

Yes, I know! I am not stupid like I make myself look. I am just too tired sometimes to think. Otherwise I understand the architecture of the processor and the programming I just have to learn to ask the questions in the right way and not to work so I can pay more attention to myself. After 3 shifts by 12 hours I am barely breathing, they put me on the best positions......And the bosses were near me the whole day.....

I do not want to read the datasheet because another 100 or 600 pages is not good for me! I have enough to read for now.

And here is another stupid question..... :D.
How do I disable and use the RC(C) pins on PIC16F690 for digital I/O. I think it is with setup_counters function because it affects the whole register or some other function maybe? This is the bad think of CCS you can not set individual bits on a register.

I just pressed the new topic button instead of the new post button.....
_________________
Yo! I love learning and technology! I just do not have experience so do not be angry if I ask a stupid question about a detail! From so much to remember sometimes I forget the details in order to remember the big problems!

[Ttelmah]

You really do need to read the data sheet. This is the first thing to read before trying to program any chip.....
Once you have read a few, whole sections become repeats, so you can just jump to the actual figures needed.
There is nothing involving counters on the RC pins. If you look at table1 in the data sheet..... this shows what pins are used by what device.

AN4 to 7, and the comparators.

So setup_adc_ports and setup_comparator.

setup_comparator(NC_NC_NC_NC);

Disables the comparators. Then ensure you are only selecting the s_ANx pins you actually want to use with setup_adc_ports, and RC0 to RC7 (C0 to C7) become available for normal I/O.

[Arakel]

This is very useful information:

[Ttelmah]

The conversion you need it to Binary Coded Decimal (BCD). So decimal digits, but coded as 0 to 9, rather than as ASCII.

A search here will find lots of 7 segment driver code, and most will include a binary to BCD conversion routine.

In fact if you search for bin2bcd, you should find one. It's commonly done as a macro, rather than a function.

[Arakel]

[Ttelmah]

If you start with a number, this has to be converted to BCD, before you get involved with the seven segment.

Think about it. You have the number '88' say. In binary 0b0101 1000
Where are the digits?.

Now convert this to BCD, and in binary you have 0b1000 1000
each 'nibble' here is '8'. This is what needs to feed to each digit.

Now with this, you take just one nibble (say the low one), and then use a look up table to say which segments have to be 'on'. So if you say that bit0 is segment 'a', 1 segment 'b' etc., and your electronics is wired so that a '1' turns the segment on, then you can have a constant table like:

gfedcba
0b00111111, //0
0b00000110, //1
0b01011011, //2

etc.. Then use the BCD digit to select a value and this is the seven segment pattern required.

So binary to seven segment, requires two conversions:

binary to BCD
BCD to seven segment.

[Arakel]

I know. This is my table and code:

const unsigned int8 TableForDigits[10] = {0b00000000, 0b00000001, 0b00000010, 0b00000011,
0b00000100, 0b00000101, 0b00000110, 0b00000111, 0b00001000, 0b00001001,};

FirstDigit = Display%10;
SecondDigit = Display/10;

for(i=0; i<=10; i++)
{
if(FirstDigit == i)
BinaryFormOfFirstDigit = TableForDigits[i];
if(SecondDigit == i)
BinaryFormOfSecondDigit = TableForDigits[i];
}
FinalFormOfThePORTCOutput = (BinaryFormOfSecondDigit >> 4) & BinaryFormOfFirstDigit;

output_c(FinalFormOfThePORTCOutput);
delay_us(100);
output_high(PIN_B5);

I am using two 4511 decoders with two 7 segment display.
That is why I find the first digit by dividing to 10 with remainders and getting the remainder, and I find the second digit by normally dividing to 10. I do not know if it will work. But the idea is to output the first digit on 4 pins and the second digit on 4 pins. So I need to make one 8 bit variable from 0b00000001 and 0b00000001 or from 0b0001 and 0b0001 I think.
_________________
Yo! I love learning and technology! I just do not have experience so do not be angry if I ask a stupid question about a detail! From so much to remember sometimes I forget the details in order to remember the big problems!

[Arakel]

The other option is 2 tables, but I was trying to save resources.

const unsigned int8 TableForDigits[10] = {0b00000000, 0b00000001, 0b00000010, 0b00000011,
0b00000100, 0b00000101, 0b00000110, 0b00000111, 0b00001000, 0b00001001};

const unsigned int8 TableForSecondDigits[10] = {0b00000000, 0b00010000, 0b00100000, 0b00110000, 0b01000000,
0b01010000, 0b01100000, 0b01110000, 0b10000000, 0b10010000};

FirstDigit = Display%10;
SecondDigit = Display/10;

for(i=0; i<=10; i++)
{
if(FirstDigit == i)
BinaryFormOfFirstDigit = TableForDigits[i];
if(SecondDigit == i)
BinaryFormOfSecondDigit = TableForSecondDigits[i];
}
FinalFormOfThePORTCOutput = (BinaryFormOfSecondDigit | BinaryFormOfFirstDigit;

output_c(FinalFormOfThePORTCOutput);
delay_us(100);
output_high(PIN_B5);
_________________
Yo! I love learning and technology! I just do not have experience so do not be angry if I ask a stupid question about a detail! From so much to remember sometimes I forget the details in order to remember the big problems!

[Arakel]

Here is the whole source code:

//A program to scan the values of an ADC and potentiometer and start a heater if they are different with the debouncing routine
#include <16F690.h>

#device ADC=10

#fuses NOMCLR, NOWDT, NOPROTECT, INTRC_IO
#use delay(clock=4M)

void main()
{

int i; //Local variables
unsigned int8 ADC_10bit_Sensor;
unsigned int8 ADC_10bit_Potentiometer;
unsigned int8 Display;
unsigned int8 FirstDigit, SecondDigit;
unsigned int8 BinaryFormOfFirstDigit, BinaryFormOfSecondDigit;
unsigned int8 FinalFormOfThePORTCOutput;

const unsigned int8 TableForDigits[10] = {
0b00000000, //Output for "0" for the first display from "FirstDigit = Display / 10"
0b00000001, //1
0b00000010, //2
0b00000011, //3
0b00000100, //4
0b00000101, //5
0b00000110, //6
0b00000111, //7
0b00001000, //8
0b00001001 //9
};

const unsigned int8 TableForSecondDigits[10] = {
0b00000000, //Output for "0" on the second display from "SecondDigit = Display/10"
0b00010000, //1
0b00100000, //2
0b00110000, //3
0b01000000, //4
0b01010000, //5
0b01100000, //6
0b01110000, //7
0b10000000, //8
0b10010000 //9
};

setup_comparator(NC_NC_NC_NC);

setup_adc_ports( sAN0 | sAN1 ); //Adc functions
setup_adc( ADC_CLOCK_DIV_16 );

while(TRUE)
{

set_adc_channel( 0 ); //We set the ADC to read from CHANNEL 0 (Where the potentiometer is connected)
for(i=0; i<15; i++)
{
delay_ms(1);
ADC_10bit_Potentiometer = read_adc();
if(ADC_10bit_Potentiometer != ADC_10bit_Potentiometer)
i=0;
}

set_adc_channel( 1 ); //We set the ADC to read from CHANNEL 1 (Where the sensor is connected)
for (i=0; i<15; i++)
{
delay_ms(1);
ADC_10bit_Sensor = read_adc();
if (ADC_10bit_Sensor != ADC_10bit_Sensor)
i=0;
}

if (ADC_10bit_Sensor <= ADC_10bit_Potentiometer) //If the potentiometer is lower than the sensor turn OFF the heater
{
delay_ms(1);
output_b(0b00010000);
} else {
output_b(0b00000000);
}
Display = (ADC_10bit_Sensor - 3.13)/0.01;

FirstDigit = Display%10;
SecondDigit = Display/10;

for(i=0; i<=10; i++)
{
if(FirstDigit == i)
BinaryFormOfFirstDigit = TableForDigits[i];
if(SecondDigit == i)
BinaryFormOfSecondDigit = TableForDigits[i];
}
FinalFormOfThePORTCOutput = BinaryFormOfSecondDigit | BinaryFormOfFirstDigit;

output_c(FinalFormOfThePORTCOutput);
delay_us(100);

}//Close while
}

It bothers me if I use "SecondDigit = Display / 10" will it work. It should work because we have int8, meaning we do not record floating point variables with int, if we want floating point we should use float?

I am certain that the binary summing does not work!

Here is the full schematic.

[img]https://drive.google.com/file/d/0B0kVHbb80xxTeHNLQng0cU9VdWM/view?usp=sharing[/img]
_________________
Yo! I love learning and technology! I just do not have experience so do not be angry if I ask a stupid question about a detail! From so much to remember sometimes I forget the details in order to remember the big problems!

[temtronic]

1) this...
Display = (ADC_10bit_Sensor - 3.13)/0.01;

May not give you the numbers you think it should...
Have you sent the 'display' variable to a PC to see what the numbers really are? Confirm that the range and type are correct?

2)since you're using 2 4511s, another fast/accurate way to send the data is to have a simple table that is 100 elements long.Old school, no math is needed,you just have to manually (or eXcel) create the table once.IE element 24 will have 0b00100100 in it, so 0010 goes to 10s digit 4511, 0100 goes to 1s digit 4511. If you've got the code space for it,it is that simple.....

Jay