 |
 |
| View previous topic :: View next topic |
| Author |
Message |
Chaud
Joined: 11 Jan 2012
Posts: 39
|
| Gyro XV 3500CB angular rate sensor , i really need some help |
 Posted: Tue Jan 31, 2012 7:20 am |
 |
|
Hello ccs community,
Anyone here already used gyroscope XV 3500CB ?? I'm looking for some support with this component.
 |
|
 |
Battery David
Joined: 01 Feb 2010
Posts: 25
|
|
|
Chaud
Joined: 11 Jan 2012
Posts: 39
|
|
| Posted: Tue Jan 31, 2012 8:56 am |
|
|
EDIT: needed to cut this thread
Last edited by Chaud on Wed Mar 21, 2012 10:21 am; edited 1 time in total |
|
|
RF_Developer
Joined: 07 Feb 2011
Posts: 839
|
|
| Posted: Tue Jan 31, 2012 10:29 am |
|
|
This module is a rotation *rate* gyro detector. Its output is proportional to the rate of change of angle, not the angle itself. So, you get an output as you twist it, +ve for one way, -ve for the other. As soon as you stop twisting the output returns to the 1.35V quiescent state. To get the angle you would have to integrate the output of this sensor.
To get it to work well you probably need to do some analogue signal conditioning: amplifying and possibly filtering. What is the maximum rate of change of angle you are expecting? say its 90degrees/second, if so you amplify and offset the output of the sensor to make +/-90deg/sec equivalent to your ADC range, say 0-5V. The sensor gives you 0.67mV/degree/second with a 1350mV nominal offset. So 90deg/sec is 60.3mV. +/-90deg/sec is 120.6mV and you probably want that to be 5V, so you need gain of 41.46, and you'll have to offset it from 1.35V to 2.5V. All straightforward analogue stuff. I'm not saying these are the numbers you need, you'll have to decide what you really need for yourself and design your circuit to give it to you.
RF Developer |
|
|
Chaud
Joined: 11 Jan 2012
Posts: 39
|
|
| Posted: Tue Jan 31, 2012 11:05 am |
|
|
| RF_Developer wrote: | This module is a rotation *rate* gyro detector. Its output is proportional to the rate of change of angle, not the angle itself. So, you get an output as you twist it, +ve for one way, -ve for the other. As soon as you stop twisting the output returns to the 1.35V quiescent state. To get the angle you would have to integrate the output of this sensor.
To get it to work well you probably need to do some analogue signal conditioning: amplifying and possibly filtering. What is the maximum rate of change of angle you are expecting? say its 90degrees/second, if so you amplify and offset the output of the sensor to make +/-90deg/sec equivalent to your ADC range, say 0-5V. The sensor gives you 0.67mV/degree/second with a 1350mV nominal offset. So 90deg/sec is 60.3mV. +/-90deg/sec is 120.6mV and you probably want that to be 5V, so you need gain of 41.46, and you'll have to offset it from 1.35V to 2.5V. All straightforward analogue stuff. I'm not saying these are the numbers you need, you'll have to decide what you really need for yourself and design your circuit to give it to you.
RF Developer |
Hey RF Developer, thanks for reply, the first part you are right and i get it already, how to integrate and get the angle.
I am expecting like 1~3 degrees variation detection.
About second part i heard i can use I2C interface to obtain a much bigger resolution than ANA0(analog output).
The I2C interface pins are pin1 (IICCLK) and pin 2 (IICDAT) , to be honest i never used I2C interface before, do you know how i use it with PIC 18F2550? i saw this CCS example:
| Code: | /*
This configures two I2C ports, each has a different stream name.
*/
#use i2c(sda=PIN_C4, scl=PIN_C3, stream=I2C_HW)
#use i2c(sda=PIN_B1, scl=PIN_B2, stream=I2C_SW)
/*
The following function reads a data byte from a Microchip 24LC16 I2C EEPROM, using the I2C_HW stream
*/
int Read2416(int address)
{
i2c_start(I2C_HW, 1); //perform a start
i2c_write(I2C_HW, 0xA0);
i2c_write(I2C_HW, address);
i2c_start(I2C_HW, 2); //perform a restart
i2c_write(I2C_HW, 0xA1);
data=i2c_read(IC2_HW, 0);
i2c_stop(I2C_HW);
return(data);
} |
So i need to connect sda C4 and scl to B3 ( using first I2C port) , and with int Read2416 i simple obtain the data i want? And what adress i must give on method? any1 can giveme some lights about how to use I2C interface?
Thanks |
|
|
asmboy
Joined: 20 Nov 2007
Posts: 2128
Location: albany ny
|
|
| Posted: Tue Jan 31, 2012 1:12 pm |
|
|
| Quote: | | Gyro_output stable is always 1,34V ~ 1,35V , with AD conversion the values are 68~72. If I get a value higher than 72 its because i rotate gyro to right. If I get lower than 68 is because i rotate gyro to left. |
i suspect that better AND simpler approach is really as RF_devel has indicated.
this is a pretty simple task for a solid 1.35v ref - or it there is an assured return-to-zero condition , an integrator of suitably long TC --
coupled with a gain adjustable Dif/ instrumentation -amp --and THAT will get you a very easily processed analog input to your pic . |
|
|
Chaud
Joined: 11 Jan 2012
Posts: 39
|
|
| Posted: Tue Jan 31, 2012 1:58 pm |
|
|
I discovered that using I2C is better because it gives 16-bit resolution, with I2C I don't need to amplify signals.
So I need to read a 16-bit integer, I coded this:
| Code: |
i2c_start(I2C_HW, 1);
value = i2c_read(I2C_HW, 0);
value |= i2c_read(I2C_HW,0) << 8;
lcd_gotoxy(1,1);
printf(lcd_putc,"%d",value);
|
But the print value is always "-1", is the syntax correct ? How you guys do things like this on CCS compiler? |
|
|
asmboy
Joined: 20 Nov 2007
Posts: 2128
Location: albany ny
|
|
| Posted: Tue Jan 31, 2012 2:37 pm |
|
|
Look at the CCS documentation for
| Code: |
MAKE16(varhi,varlow);
|
the rest is EZ.
Remember to declare your INT16 that you pass the result of make16 TOO - as being signed or unsigned as you may require. |
|
|
Chaud
Joined: 11 Jan 2012
Posts: 39
|
|
| Posted: Tue Jan 31, 2012 4:19 pm |
|
|
Check if i did right
| Code: |
int value1;
int value2;
unsigned int16 gyro;
loop:
value1 = i2c_read();
value2 |= i2c_read() << 8;
gyro = make16(value2,value1);
lcd_gotoxy(1,1);
printf(lcd_putc,"%ld",gyro);
end loop: |
When i send the hex file it prints "255", but if i power off circuit and re power on it prints "25855" , something is wrong, it is suposed to get a variation not always 25855 :x |
|
|
temtronic
Joined: 01 Jul 2010
Posts: 9128
Location: Greensville,Ontario
|
|
| Posted: Tue Jan 31, 2012 4:45 pm |
|
|
Have you the correct pullup resistors on the I2C bus lines?
When doing 'math' functions, you should always test with known data.
Instead of
| Code: |
value1 = i2c_read();
value2 |= i2c_read() << 8;
|
try
| Code: |
value1= 0x21;
value2= 0x43;
|
then see what happens...
You should try a few keystone values like 0000, ffff, 1234, etc. to be _sure_ the math is correct! Nothing is worse than finding out the bytes are swapped (2143 when 4321 is right...) ! |
|
|
Chaud
Joined: 11 Jan 2012
Posts: 39
|
|
| Posted: Tue Jan 31, 2012 5:02 pm |
|
|
| temtronic wrote: | Have you the correct pullup resistors on the I2C bus lines?
When doing 'math' functions, you should always test with known data.
Instead of
value1 = i2c_read();
value2 |= i2c_read() << 8;
try
value1= 0x21;
value2= 0x43;
then see what happens...
You should try a few keystone values like 0000, ffff, 1234, etc. to be _sure_ the math is correct! Nothing is worse than finding out the bytes are swapped (2143 when 4321 is right...) ! |
Hey the math is correct, but I think I found the problem, look:
If
| Code: |
value1= 0x21;
value2= 0x43;
|
result = 17697 -> CORRECT
if
| Code: |
value1= 255;
value2= 120;
|
result= 30975 -> CORRECT
if
| Code: |
value1=255;
value2=255;
|
result= -1 -> is this what I am getting when I read values from i2c_read(). The first value being 255 is normal, because gyro normal output is something higher than 255 if resolution is 16-bit, but second value2 being 255 is not normal :O
About pull-ups I am using 1k resistors. How can I know which resistors I should use?
EDIT: now it prints 255 from:
| Code: |
value1 = i2c_read();
value2 |= i2c_read() << 8;
gyro = make16(value2,value1);
lcd_gotoxy(1,1);
printf(lcd_putc,"%ld",gyro);
|
If I print value1 it is 255 and if value2 it is 0, so it is correct, the thing is why value 2 |= i2c_read() << 8 ; is zero? |
|
|
asmboy
Joined: 20 Nov 2007
Posts: 2128
Location: albany ny
|
|
| Posted: Tue Jan 31, 2012 5:15 pm |
|
|
BTW: a struct of two 8 bit vars - defined as a 16 bit combined INT can do the same thing as the make() function - but first you will have to sort out the transfer / level shift problems with the i2c connection.
a confession - i endorsed the analog approach because
of my DEEP prejudice:
1- analog is my strong suit , by quite a large margin
2- i much prefer SPI having had nothing but misery/debug when
STUCK implementing i2c parts
(IMHO: I2C a concept that only Philips could love or invent )
  |
|
|
temtronic
Joined: 01 Jul 2010
Posts: 9128
Location: Greensville,Ontario
|
|
| Posted: Tue Jan 31, 2012 5:43 pm |
|
|
hmm....
value1= 0x21;
value2= 0x43;
result = 17697 -> CORRECT
for 0x4321(value2,value1) I get 17185 using 2 different 'calculators' !
Anyone else confirm,deny ???? |
|
|
Chaud
Joined: 11 Jan 2012
Posts: 39
|
|
| Posted: Tue Jan 31, 2012 5:45 pm |
|
|
| temtronic wrote: | hmm....
value1= 0x21;
value2= 0x43;
result = 17697 -> CORRECT
for 0x4321(value2,value1) I get 17185 using 2 different 'calculators' !
Anyone else confirm,deny ???? |
it was 17185, my mistake !
I think the problem is here:
| Code: | | value2 |= i2c_read() << 8; |
What I want on this line is the most significant 8 bits, and I think that line isn't doing what I want. |
|
|
asmboy
Joined: 20 Nov 2007
Posts: 2128
Location: albany ny
|
|
| Posted: Tue Jan 31, 2012 6:33 pm |
|
|
Well lucky you -
0x4321 is decimal 17185 && is
the same whether you are signed or UNSigned INT16
Assuming 0x43 is the MSB that is.
If would be more tricky if the unsigned i2c ret value was > 0x7FFF.
BUT more to the point - while you are figuring out how to organize bytes into 16 bit quantities - I might add - that if the i2c Byte return values are actually always 255 decimal - then you have HARDWARE interface trouble at a MINIMUM!!!
So I need to ask:
Have you evaluated IF you are having level translation / pull up issues with the circuit?? How do you know you DON'T ?? \methinks you focus too much on the code right now - and not enough on the circuitry -
The code will never be right if the circuity is a mess. |
|
|
|
|
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
