PGA450-Q1: time of flight

Hello hunt.cn,

The example pga450_main.c equation to solve for time-of-flight is provided as:

//*** CODE SNIPPET BEGIN ***//
if ( MaskEcho == 0){
if ( data_FIFO[lcv] > DetectionThreshold) {

EN_CTRL = 0x8C; //log free running timer
FRT_T2.u8[0]=FRT_MSB; // ESFR 0xBA
FRT_T2.u8[1]=FRT_LSB; // ESFR addr 0xBB
if (FRT_T1.u16 > FRT_T2.u16) {
TimeOfFlight.u16 = 65535+FRT_T2.u16-FRT_T1.u16;
}
else {
TimeOfFlight.u16 = FRT_T2.u16-FRT_T1.u16;
}
MaskEcho = 1;
//break; // when enabled, break from loop as soon as an object is found
}
}
//*** CODE SNIPPET END ***//

The variable FRT_T1.u16 is captured immediately after issuing the burst-and-listen command, while the variable FRT_T2.u16 is captured at the time of the threshold interrupt. This difference does not need to factor in the SAT_TIME since T1 already includes this time.

Refer to section 7.3.11 Free-Running Timer of the PGA450-Q1 datasheet on FRT details.

Hello Akeem,

I don't want to use Free-Running Timer ,becauce the test results have about +5cm error when using the example pga450_main.c .

So I used Equation 2  in the 《LIN Demostration Using PGA450Q1EVM》， this Equation's Accuracy  is  about ±1.5cm. But I don't know why have not SAT_TIME.

code as:

//*** CODE SNIPPET BEGIN ***//
if ( MaskEcho == 0){
if ( data_FIFO[lcv] > DetectionThreshold) {

lcv_temp1=lcv

MaskEcho = 1;
//break; // when enabled, break from loop as soon as an object is found
                                        }
}

TimeofFlightTemp.u16 =(lcv_temp1<<4) + (BLANKING_TIMER<<4);

//*** CODE SNIPPET END ***//

hunt.cn,
The transducer saturation time measurement is independent from the FIFO record start time. Therefore, the saturation time is not included in the Time-of-Flight calculation.

SAT_TIME does not offset the FIFO record start time, only the BLANKING_TIMER offsets the FIFO record start time. SAT_TIME is intended to be used for diagnostic purposes, and can be used to set the BLANKING_TIMER. For instance, if the saturation time is known to be 1000us, then the blanking timer can be set to a value of 62x16us to ignore the saturated region of the record window to enable minimum distance detection.

hunt.cn,

If your code sequence is

EN_CTRL = 0x85; // start burst, saturation check, and log FRT
...
while(STATUS2 & 0x02 == 0){} // wait for Echo signal to be below saturation threshold
...
EN_CTRL = 0x0C; // start filling FIFO
...
// time-of-flight calculation

then you will need to add the SAT_TIME to the time-of-flight equation.

Hi ,Akeem,

Thanks for your reply ,  It 's great help for me . 

My code sequence is your mention ,  not changed  codes sequence in the original <example pga450_main.c>, but some issues in my test.

1、 TOF calculation equation without SAT_TIME ,  but it had good accuracy。

I added one code to read SAT_TIME after <code> EN_CTRL = 0x0C </code> ,  and throw it out by UART ,  it 's value is 0 .

So SAT_TIME does not offset the FIFO record start time too , That is why the equation have good accuracy whitout add SAT_TIME to the TOF equation.

2、Why SAT_TIME had a zero value .

it seem to be impossible,  because there had a  code like  :

SAT_DEGLITCH = 0x09;                   // 2us per count || 1/58kHz (17.2uS)

So SAT_TIME should have 1 value at least.        //  8bit at 16 us resolution

3、I got a picture with PGA450Q1EVM GUI 1.50.1 , maybe why SAT_TIME = 0.

configure GUI 1.50.1 : in Digital Test MUX set GPIO1 to output SAT_DONE 

ch1 is burst waveform , ch2 is SAT_Done from gpio1.