Measuring the time of the zero crossings

Fan174 · May 16, 2018

I want to measure the time interval between every zero crossings for 5V AC input signal using the PIC16f877a.
I understand c programming but I am struggling to measure the time interval between every zero crossings. I did google search and found that I have to use Analog Comparator, Interrupt and CCP Module.

How to detect every zero crossing ?

if crossing detect then how to measure time ?

srizbf · May 16, 2018

use a zero crossing detector circuit(opamp comparator based).
This will give you a pulse output for every zero crossing.

start a timer at first pulse and end it after you detect the second pulse.

time interval is the timer value * period of your timer clock.

Fan174 · May 16, 2018

srizbf said:
use a zero crossing detector circuit(opamp comparator based).

I want to use inbulit comparator in pic16f877a. I think it can be compare positive and negative voltage. if its positive voltage than output will high and its negative voltage then output will low

srizbf said:
start a timer at first pulse and end it after you detect the second pulse.

time interval is the timer value * period of your timer clock.

I think I need to set interrupt with cpp model

FvM · May 16, 2018

For precise timing measurements, CCP is the way to go, it's the only way to get it crystal accurate without e.g. interrupt latency effects. Internal comparator can't directly drive CCP, if you want it though, you need an external loopback connection from comparator output to CCP input.

Fan174 · May 16, 2018

FvM said:
For precise timing measurements, CCP is the way to go, it's the only way to get it crystal accurate without e.g. interrupt latency effects. Internal comparator can't directly drive CCP, if you want it though, you need an external loopback connection from comparator output to CCP input.

Exactly what do you mean ? Do i need to use external comparator?

KlausST · May 16, 2018

Hi,

I assume the PIC and it´s comparator input can´t handle negative input voltages. Read datasheet.

Also read about CCP in the datasheet.
In short: CCP just safes the timestamp of the CCP event (zero cross)
You need to save this timestamp (t0) and wait for the next CCP event.
you may calculate delta_t = t_new - t0
then save t_new at the place of t0
do this in a loop

--> delta_t is the time difference in clock ticks from one zero_cross to the next zero_cross.

Klaus

FvM · May 16, 2018

Exactly what do you mean ? Do i need to use external comparator?

I say, you can use the internal comparator with a loop back connection. Generally I want you to read the data sheet and think yourself.

I assume the PIC and it´s comparator input can´t handle negative input voltages.

Surely not. But it would be sufficient to clamp negative input voltage. I wonder if you even need a comparator. Most zero crossing solutions work without it.

Unless you are using a schmitt trigger or a comparator with hysteresis you should expect multiple trigger events for each fundamental wave zero crossing due to noise. If the frequency range is limited (e.g. mains frequency), a time filter can do the trick.

Fan174 · May 17, 2018

KlausST said:
Hi,

I assume the PIC and itÂ´s comparator input canÂ´t handle negative input voltages. Read datasheet.

Also read about CCP in the datasheet.

Klaus

Okay I looked datasheet and some links and I tried to understand basic of Capture mode in PIC

Capture mode is used to capture the value of Timer 1 when a signal at the CCP pin goes high /low. So we can get the time period from the timer count. In Capture mode, CCPR1H:CCPR1L captures the 16-bit value of the TMR1 register when an event occurs on pin RC2/CCP1.

Programming step

Configured RC2/CCP1 pin as input
Configured port as Output
set Timer 1
Enabled Global interrupts
Enabled CCP1 interrupt
Set Module to capture on Falling edge/rising edge
- On interrupt:
Read capture timer and save the value

Code:

void main(void)
{
   //Make all PORTD pins low
   PORTA = 0;
   PORTB = 0;
   PORTC = 0;
   PORTD = 0;
 
// Configured RB0 as Output
   TRISB0 = 0; //RB0 as Output PIN
 
//Configured RC2/CCP1 pin as input
   TRISC2 = 1;  
    
   T1CON = 0x01;           // set Timer 1
   CCP1CON = 0b00000101;   // Set Module to capture on rising edge
   INTCON = 0xc0;          //Enabled Global interrupts 
   CCP1IE = 1;             //Enabled CCP1 interrupt
   while(1);
}
 
void interrupt CCP1_ISP()
{
   // CCP1 Interrupt
   if(CCP1IF == 1)   // if the CCP1 Interrupt flag is set...
   {
      // Read capture timer and save the value//
   }
}

KlausST · May 17, 2018

Hi,

basically correct.
Just the math is missing.

in case you expect bouncing signals:
* You know the expectable value. Maybe 10ms. So you know when the measured value is much smaller than the expected value .. you just may dismiss the captured value.

Klaus

Fan174 · May 17, 2018

KlausST said:
Hi,

basically correct.
Just the math is missing.

in case you expect bouncing signals:
* You know the expectable value. Maybe 10ms. So you know when the measured value is much smaller than the expected value .. you just may dismiss the captured value.

Klaus

I want to write program to measure time for every crossing.

I have used two variable value1 and value2 to store count value.

How to measure time for every crossing?

Code:

#include <xc.h>

 unsigned int value1, value2;
 
void main(void)
{
   
    
   //Make all PORTD pins low
   PORTA = 0;
   PORTB = 0;
   PORTC = 0;
   PORTD = 0;
 
// Configured RB0 as Output
   TRISB0 = 0; //RB0 as Output PIN
 
//Configured RC2/CCP1 pin as input
   TRISC2 = 1;  
    
   T1CON = 0x01;           // set Timer 1
   CCP1CON = 0b00000101;   // Set Module to capture on rising edge
   INTCON = 0xc0;          //Enabled Global interrupts 
   CCP1IE = 1;             //Enabled CCP1 interrupt
   while(1);
}
 
void interrupt CCP1_ISP()
{
   // CCP1 Interrupt
   if(CCP1IF == 1)            // if the CCP1 Interrupt flag is set...
   { 
                  value1 = CCPR1H; 
                  value2 = CCPR1L;       
   }
}

KlausST · May 17, 2018

Hi,

I recommend to use one uint16 varible.

How to measure time for every crossing?

What´s unclear with the description of post#6?

Klaus

Fan174 · May 18, 2018

KlausST said:
Hi,

What´s unclear with the description of post#6?

Yes I have somethings that I don't know how to implement in program

KlausST said:
In short: CCP just safes the timestamp of the CCP event (zero cross)
You need to save this timestamp (t0) and wait for the next CCP event.
you may calculate delta_t = t_new - t0
then save t_new at the place of t0
do this in a loop

--> delta_t is the time difference in clock ticks from one zero_cross to the next zero_cross.

Klaus

When first interrupt happen it store data in the value1 and value2. How to store new data when next interrupt happen. I will have to use while loop but I need to make logic. I am struggling where to write code to store new value inside interrupt routine or outside interrupt routine

Code:

void interrupt CCP1_ISP()
{
   // CCP1 Interrupt
   if(CCP1IF == 1)            // if the CCP1 Interrupt flag is set...
   { 
                  value1 = CCPR1H; 
                  value2 = CCPR1L;       
   }
}

KlausST · May 18, 2018

Hi,

Since CCP1 represents a 16 bit value I recommend to use uint16.
For sure you may use a nother variable: new_ccp (just local in ISR)
And you need the variable: delta_t (global)

Then (pseudo code)

Code:

* New_ccp = ccp1h, ccp1l
* Delta_t = new_ccp - t0
* t0 = new_ccp

No need for a loop

Klaus

FvM · May 18, 2018

You told remarkable little about your application, e.g. expected trigger frequency, so we can only guess about specific requirements. It should be clear that the previous capture time stamp must be stored or processed before it's overwritten. How it should be done depends to a large extend on the application details. In some cases, it might be processed directly in the IRQ function.

In any case, you should consider that timer 1 overflows at times. It may be necessary to track, possibly count it.

Fan174 · May 18, 2018

KlausST said:
Hi,

Since CCP1 represents a 16 bit value I recommend to use uint16.
For sure you may use a nother variable: new_ccp (just local in ISR)
And you need the variable: delta_t (global)

Then (pseudo code)

Code:

* New_ccp = ccp1h, ccp1l * Delta_t = new_ccp - t0 * t0 = new_ccp

No need for a loop

Klaus

I am new to PIC and xc8 IDE,. Can you please provide a link or examples on how to store 16 bit data in 16 bit register example :- UINT unsigned int

I cannot seem to find information anywhere about using UINT, but I see many code examples where it is used. I am having a issue to use UINT16 due to this lack of understanding.

Can we write like following to store 16 bit data in one 16 bit variable

UINT unsigned int = CCPR1H, CCPR1L

KlausST · May 18, 2018

Hi,

Can you please provide a link or examples on how to store 16 bit data in 16 bit register example

This is a very basic job. Every tutorial will show how to do it. I even assume that the PIC datasheet tells you how to access 16 bit register values.

--> You urgently need to go through some basic tutorials. There will be even video tutorials.

****

In any case, you should consider that timer 1 overflows at times. It may be necessary to track, possibly count it.

I recommend to set counter frequency that it never counts more than 65536 with your lowest input signal frequency, then you don´t have to take care about timer overflow. Math will be correct.

Klaus

Fan174 · May 18, 2018

KlausST said:
Hi,

This is a very basic job. Every tutorial will show how to do it. I even assume that the PIC datasheet tells you how to access 16 bit register values.

--> You urgently need to go through some basic tutorials. There will be even video tutorials.

Klaus

I gone through the tutorials but still have same issue.

Generally we declare and assign variable in following way

Data type variable name = data

Example : int number = 10 or char name = 'R'

When I wrote uint8_t in program it show error

Code:

void main(void)
{  
     uint8_t T1;
}

FvM · May 18, 2018

When I wrote uint8_t in program it show error

So it's apparently not defined in the default include files. Which compiler are you using?

Fan174 · May 21, 2018

FvM said:
Which compiler are you using?

I am using MPLABX xc8 compiler and PIC16F877A

FvM · May 21, 2018

uint8_t is defined in stdint.h XC8 doesn't include it by default. A typical XC8 source starts with

Code:

#include < xc.h>
#include < stdint.h>
#include < stdlib.h>

Be aware of < arguments> being eaten up by the HTML parser.

Welcome to EDAboard.com

Measuring the time of the zero crossings

Member level 2

Advanced Member level 5

Member level 2

Super Moderator

Member level 2

Advanced Member level 7

Super Moderator

Member level 2

Advanced Member level 7

Member level 2

Advanced Member level 7

Member level 2

Advanced Member level 7

Super Moderator

Member level 2

Advanced Member level 7

Member level 2

Super Moderator

Member level 2

Super Moderator

Similar threads

Part and Inventory Search

Welcome to EDABoard.com

Sponsor