Just loaded 2.0.55 (three times all up) and for some reason the gyros refuse to initialise. Running the IMU command in CLI just keeps displaying * characters after "Initialising gyros" but never resolves.
Is my board broken? The only change I've made recently was to solder a header on for the Attopilot current sensor. All the other test commands seem to work in CLI (except current and battery, but that's another problem for another day).
Replies
Ok, here's what I've got so far after some testing without reloading the code.
If the APM boots without the current sensor plugged in everything seems to work normally. The camera bracket rolls and tilts normally and the artificial horizon responds normally in mission planner. Additionally when I run test->IMU in CLI the gyros initialise and return proper data. I also get raw sensor data for the gyros from the Planner in flight mode.
If I boot the APM with the sensor plugged in none of the above happens. The roll tilt bracked barely responds, and the artificial horizon is sluggish. I get no data from the gyros in the flight data mode of the Planner or in the IMU test in CLI.
I'll try loading 2.0.56, and if that fails, editing the code. Afsin, does the current sensor work now you've edited the gyro offset?
Chris
If you haven't succeeded, try to edit libraries\AP_IMU\AP_IMU_Oilpan.h, increase the offset
gyro callibration offset value in the line;
static const float _gyro_max_cal_offset = 320.0;
Irrelevant of the current sensor, this happened to me a while ago, there was no answer in the forums regarding to this problem, so I increased the offset value, then the initiallization succeeded. The offset value is now 385 for my quad and it works fine. You have to load the code with arduino though. Once you load the code, perform IMU test in CLI, check the gyro offset value, if it is too low than your preset value, reset it +5 of the test output value and load the code again (actually I dont know if it is necessary to limit the offset value that much:)).
I hope it does helps.
Christopher
If you unplug the current sensor, does the APM come back to life after reloading the software. Next if you don't tell the software that you want don't want current and voltage, does it work?
Reason for the questions, I have had a simular problem with 2.24 & 2.26 plane software, and have seemed to narrow the problem down to my AttoPilot Voltage and Current Sensor. I could get a few cycles that it would work, then fail on a change of the battery. On startup again, no startup lights, and no connection with the MP execpt in CLI mode.
I have the sensor unplugged and left the software set for none, and now have 8 flights with out a problem.
Burt