Search the Community
Showing results for tags 'offsets'.
Found 2 results
Hi all, This is my first post here. I am helping out a friend with a project where he needs to read acc & gyro data from two independent surfaces. For this I am using two GY521s, at their available addresses of x68 and x69. I have the hardware working and after tweaking Jeff's example to get raw data to accommodate both sensors I am getting data. I also used the DMP6 example and got the three axis reading for the teapot demo. Not satisfied I started to do a little research which turned into massive confusion :-). I am not sure if the raw data that I am getting is what he needs. He is working on software that will receive the data through an Arduino and send it over to the host PC running the application. The app needs to receive the six gyro and accelerometer data items, along with an indicator of time passed. The app will create a graph of the surface's movement. My research yielded information about calibration and offsets. I am currently reading and trying to understand the raw data that I am getting and trying to determine whether I need the offsets or not. I found the following thread on this site: http://www.i2cdevlib.com/forums/topic/4-understanding-raw-values-of-accelerometer-and-gyrometer/ BTW: The link to the MPU 6050 datasheet is broken. I will continue to research and read the thread referenced above and the datasheet; however, I wanted to ask here, hoping that perhaps someone can help me get my head around this versatile sensor. My questions: What does the raw data indicate? Why does it vary so much even though the GY521 is not moving? What does the calibration do? Why is it important? The calibration process yields offsets, what do these values mean? Any helpful recommendations or clarifications are welcomed. I will follow up with this thread with whatever information I find. Thank you all for your time and assistance. Sgarv
mhickman posted a topic in MPU-6050 6-axis accelerometer/gyroscope (InvenSense)We have some Arduino MEGA 2560-based boards with an on-board MPU-6050 at the default I2C address 0x68. i2cdevlib is extremely helpful (thank you Jeff!), except we can't seem to generate offsets that zero the sensor outputs. I've been using luisrodenas' Arduino sketch, which he posted in this forum recently, to calculate MPU6050 offsets. I modified the output to print the mean (2000 samples in this case) of the sensor outputs and the revised offsets calculated based on those mean values each time through the loop. The full data is attached and I've included a sample from gyro_z below. As you can see, the offsets change to try to move the mean values to zero. The sensors outputs seem to resist this adjustment, then at some offset threshold, the sensor values jump way past zero. This cycle then repeats. Only the gyro_y value responds as expected. Does anyone have any ideas what could be causing this? I've verified this on several PCBs, so it doesn't seem to be a bad MPU. I saw similar results setting offsets manually, so I don't think the issue is in the calibration sketch. Thanks, Mike MPU6050 connection successful Reading sensors for first time... Calculating offsets... Configuration: buffersize=2000 acel_deadzone=16 giro_deadzone=2 mean_gz gz_offset 492 140 931 -170 -93 -139 -89 -110 -88 -81 -89 -52 -88 -23 -89 6 934 -305 -1112 65 934 -246 -89 -217 -89 -188 -89 -159 -89 -130 -88 -101 -89 -72 -89 -43 -89 -14 -89 15 934 -296 -1113 75 output.txt