search

sandeepmistry / node-sensortag

Node.js lib for the TI SensorTag
MIT License
218 stars 111 forks source link

Accelerator - Gyroscope sensor mistaken #59

Closed nixcorn closed 8 years ago

nixcorn commented 9 years ago

Hi

I load the sensordata to a InfluxDB and visualize the data with Grafana. There I found the problem, that the gyroscope data are the data from the accelerator and vise versa. See the attached screenshot.

accel_gyro

I turned the sensor in different directions (12:56:50 or 12:57:05) At 12:57:35 i threw the sensor in the air and at 12:57:50 i let fall it on the table.

You can check out the source here: https://github.com/makeplantsnotwar/node-bluetooth-connector/blob/master/app.js

Thanks

sandeepmistry commented 9 years ago

@unicorn42 what Sensortag model are you using?

It's probably best for you to dive into the code and submit a pull request with any fixes you make. You can use the following links for reference to the data formats and formulas:

nixcorn commented 9 years ago

@sandeepmistry We used the CC2650 Sensortag.

I just found the bug while i setup Grafana. I have no clue about javascript nor node specific stuff. So i can't really dive into the code. It should be a short thing, just exchange the identifiers and you're done ;-)

sandeepmistry commented 9 years ago

@unicorn42 alright, I double-checked the conversion code with the TI docs, and the calculations match.

The TI SensorTag iOS app also shows the same values. However they are probably wrong as well. Which firmware version are you running on the SensorTag? I just updated to 1.20, I wonder if they change something ....

nixcorn commented 9 years ago

I'dont have the SensorTag anymore available to answer that question. Just change the position of de Sensor and you will see the graphs changing at the accelerator. It the tag is not moving accelerator should be 0, but it's not. Thats why its wrong.

sandeepmistry commented 8 years ago

Some notes from Gitter:

@zywind Nov 24 11:31

I asked about this on the TI support forum here: https://e2e.ti.com/support/wireless_connectivity/f/538/t/471879 It turns out the sensor data are non-calibrated raw data. So I need to calibrate it in my application For the magnetometer, I could probably use the method described here: https://github.com/kriswiner/MPU-6050/wiki/Simple-and-Effective-Magnetometer-Calibration

sandeepmistry commented 8 years ago

I'm marking this as a duplicate of #62, I've updated the accelerometer conversion formulas, as they were off by a value of 4.

Acceleration should be +/- 1g when not moving, to account for gravity.