| View previous topic :: View next topic |
| Author |
Message |
Zek_De
Joined: 13 Aug 2016
Posts: 100
|
| ADXL345 Angle Problem |
 Posted: Wed Dec 14, 2016 1:14 pm |
 |
|
Friends,
Why does device show 86 or 87 degree as max angle?
| Code: |
void Acc_Values()
{
Acc_Values_Raw();
float32 x = 0.004*(AccRaw[x_eksen]);
float32 y = 0.004*(AccRaw[y_eksen]);
float32 z = 0.004*(AccRaw[z_eksen]);
tetapitch = atan2(x,sqrt(y*y + z*z)) * 57.29577;
psiroll = atan2(y,sqrt(x*x + z*z)) * 57.29577;
} |
if I use
| Code: | | psiroll = atan2(y,z) * 57.29577; |
it's okay but angle is locked at 90. Also how can I fit in 0 - 180 degree? |
|
 |
temtronic
Joined: 01 Jul 2010
Posts: 9229
Location: Greensville,Ontario
|
|
| Posted: Wed Dec 14, 2016 1:20 pm |
|
|
We NEED to see your raw data !
AccRaw[x_eksen], y, and z......
Without KNOWING what the raw data is, no one knows if it's a raw reading problem( sensor error ) or a 'math' error somewhere in those calculations.....
Whenever you have a problem like this you need to printout(display) ALL the numbers from raw data,mid calculations as well as the final result. That way you can see where the problem is.
Jay |
|
|
Zek_De
Joined: 13 Aug 2016
Posts: 100
|
|
| Posted: Wed Dec 14, 2016 1:24 pm |
|
|
To Raw Values
| Code: | void Acc_Values_Raw()
{
unsigned int8 cikis=0;
cikis = Single_Byte_Read(ADXL345_deviceAdress_w ,ADXL345_deviceAdress_r ,INT_SOURCE);
while(bit_test(cikis,7)==0)
{
cikis = Single_Byte_Read(ADXL345_deviceAdress_w ,ADXL345_deviceAdress_r ,INT_SOURCE);
}
Multiple_Bytes_Read(ADXL345_deviceAdress_w ,ADXL345_deviceAdress_r ,DATAXL_INC , out);
AccRaw[x_eksen] = make16(out[1],out[0]);
AccRaw[y_eksen] = make16(out[3],out[2]);
AccRaw[z_eksen] = make16(out[5],out[4]);
} |
|
|
|
Zek_De
Joined: 13 Aug 2016
Posts: 100
|
|
| Posted: Wed Dec 14, 2016 1:25 pm |
|
|
| Oh sorry you mean you wanna see some values |
|
|
guy
Joined: 21 Oct 2005
Posts: 297
|
|
| Posted: Wed Dec 14, 2016 1:43 pm |
|
|
Here's a piece of code that measures between -90 to +90 deg. on a MMA8453Q accelerometer. Each of the two axis used has better accuracy on different angles, therefore the calculation is made using both.
Could be useful:
| Code: | void main()
{
byte i;
byte tiltFltr=0;
int timer=0,adcvalue=0,callini=0;
short tmp;
float f,f2;
signed int16 i16;
ini();
while(true)
{
delay_ms(5);
// handle tilt:
updateXYZ();
// x axis
i16=acc_xyz[0];
if(i16<0) i16=(-i16);
if(i16<130) {
f=acc_xyz[0];
f=asin(f/260)*180/pi;
}
else if(i16<220) {
f=acc_xyz[2]; // z axis
f=acos(f/260)*180/pi;
f2=acc_xyz[0];
f2=asin(f2/260)*180/pi;
if(acc_xyz[0]<0) f=-f; // neg. angle
f=(f+f2)/2;
}
else {
f=acc_xyz[2]; // z axis
f=acos(f/260)*180/pi;
if(acc_xyz[0]<0) f=-f; // neg. angle
}
if(f<-90) f=-90;
if(f>90) f=90;
printf("%2.1g",f);
putc(176); // deg. sign
putc('\r');
putc('\n');
}
} |
|
|
|
temtronic
Joined: 01 Jul 2010
Posts: 9229
Location: Greensville,Ontario
|
|
| Posted: Wed Dec 14, 2016 1:53 pm |
|
|
re:
Oh sorry you mean you wanna see some values
yes...
There are at least 7 reasons for the bad numbers... |
|
|
Ttelmah
Joined: 11 Mar 2010
Posts: 19516
|
|
| Posted: Wed Dec 14, 2016 3:26 pm |
|
|
Also you might want to look at this code:
<http://www.ccsinfo.com/forum/viewtopic.php?t=50920&highlight=atan>
The code I posted there is an atan2 function that is half the size and much faster than the CCS version, designed specifically to give degrees as an output directly from an input from -2047 to 2048. It's very many times faster than using atan2, more accurate than the chip, and gets rid of the need to scale the input numbers and the result.
One particular problem (which is why atan2 is available), is that using atan (which is what is actually being coded using the sin & cos here), the functions fail at the quadrant points, so may not be able to code with values as you get to 90degrees |
|
|
Zek_De
Joined: 13 Aug 2016
Posts: 100
|
|
| Posted: Thu Dec 15, 2016 2:47 am |
|
|
| Okay thanks everyone |
|
|
|
FAQ 
Search
Register
Profile 
Log in to check your private messages 
Log in
