Everyone In My lab

The whole team met in my lab today to prepare to for the CTD workshop.  Before everyone arrived I rushed to calibrate the ArduIMU glove gyroscope accelerometer and magnetometer.  Rather than do things from scratch, I used snippets of firmware and PC software from a  forthcoming x-io product.  I calibrated the gyroscope and accelerometer by fixing the ArduIMU to a ‘calibration cube’.  This allowed me to sample ±1g for all accelerometer axes and ±200 degrees per second (i.e. 33 1/3 RPM) for all gyroscope axes, it is straight forward to then calculate individual axes bias and sensitivities.

ArduIMU of Calibration Cubehttp://dev-blog.mimugloves.com/wp-content/uploads/2013/01/ArduIMU-of-Calibration-Cube.jpg 816w" sizes="(max-width: 300px) 100vw, 300px" />

Unfortunately the magnetometer proved more of  challenge.  I successfully calibrated it first time but subsequent attempts would fail as the Bluetooth module would disconnect and so prevent me from collecting a calibration dataset.  I suspect this could be due to settings on the RN-42 module.  When developing the x-IMU I found that the RN-41 connection could be lost if it stream arbitrary bytes and the RN-41 was in remote config enabled period.

Becky and I assembled the right ArduIMU gloves and connected only four of the bend sensors. Different coloured wires were used for each bend sensor so we could alter this if necessary.

Right ArduIMU Glove With Coloured Wireshttp://dev-blog.mimugloves.com/wp-content/uploads/2013/01/Right-ArduIMU-Glove-With-Coloured-Wires.jpg 816w" sizes="(max-width: 300px) 100vw, 300px" /> Right ArduIMU glove wiring noteshttp://dev-blog.mimugloves.com/wp-content/uploads/2013/01/Right-ArduIMU-glove-wiring-notes.jpg 816w" sizes="(max-width: 300px) 100vw, 300px" />

Both gloves were working but we needed to get things working with Kelly’s laptop so Imogen could test existing set-ups.  Unfortunately, Kelly’s laptop just wouldn’t behave.  When eventually got Glover working but immediately found that bend sensor calibration would result in a tiny operating range.  Tom and I suspected this is due to packet errors causing extreme measurement values and correcting range.  This could be fixed if we processed the checksum when decoding but it was getting late.  We’ll fix it next time!

Leave a Reply