h48c accelerometer

pilot00

Howdy,

Anybody got a way to calibrate the accelerometer without using a uM-FPU 3.1 as I have seen in the
attached file.

My h48c shows funny numbers when at rest.

cessnapilot

Hi pilot00,
Yes, the calibration of the H48C has nothing to do with the uM-FPU v3.1. It was just a convenience for me to use it. The algorithm of the general Least-Squares Nonlinear Optimizer is straightforward and commented enough (at least I tried to make it so) to be reproduced easily with the various float objects. For a quick and dirty solution I would start with the definition of a LONG array to mimic the FPU registers in HUB. Then I would use the register names and/or the index management of the original source code wherever appropriate. The result will be not elegant or optimized for the particular float object, but will be correct and you could get it quickly.

More important is, however, that you should translate the "AcquireSteady" object, as well. You have to collect low-pass filtered data for the described calibration.

But, please wait for about 2 weeks with the translation to start with something more worthwhile. I will post a "General Sensor Model" object at OBEX. This can be used for any 3-axes sensor, being accelero-, gyro- or magneto. Using this "General Sensor Model," each of my 5 H48C sensors achieved 1mg accuracy (RMS error about 0.01 m/s^2). That accuracy allows you to sense 0.06 deg attitude change in a tilt-sensing application. I tried this new object with a dual Memsic 2125 setup and a dual HM55B array, too. (See attached figures). I observed significant (>20) improvement of accuracy with those MEMS devices, too.

Istv

cessnapilot

Well, you can have it. Maybe that would be more interesting than those funny numbers.

JTC

I have purchased the the hc48c and would really like to give the new opject a try.
Is it available ?
thanks
Jim