I have a deteriorating problem. This started with my first copter crashing. The second frame crashed due to the bug that cuts power upon switching to an auto mode with throttle at zero. now, with new motors and a rebuilt frame, my copter has a twitching issue, and it looks like my IN channels have some spiking going on (see attached). I am using a futaba controller, which i know has caused some problems. after testing the RX/TX on an o-scope, i know it is not the futaba system that is causing the spikes nor the twitches. some bench tests w/o props show that there is a particular location in the throttle range that cause the motors (or maybe just one, i cant tell) to start pulsing pretty sporadically. I was trying to get a log file from these bench tests, but they are either corrupted before they get to the computer, or the board is fubar'd
Whats more, it looks like ive had deteriorating communications with the board itself. The log download window just throws up garbage symbols, and just now it gave me some wild error. I have attached a screen shot that shows a few of the garbage symbols and characters in the log download window, as well as an error message.
I have already tried erasing the eeprom and reloading the firmware. still trying 2.9.1b. is this the right procedure to completely wipe the board and re-install the firmware? if not, what is that procedure?
does anyone have any insight as to what is happening here?
EDIT: In that error message, it refers to a file path C:\Users\hog..... but there is no user by that name or anything even near it on this computer.
Replies
Well, my copter is flying ok in the stabilize, but it still doesnt like alt hold and i am not confident enough in its ability to control itself to try anything more than that. After looking at the data sheets for the important sensors, accelerometer and barometer, it looks like the absolute maximum ratings are 6V. (and if im reading it right - 4.8V for the baraometer?) I think my voltage spikes may have caused some hardware damage. engaging alt hold - the copter simply lands. and if i give it more throttle (i know it is relative alt when in alt hold) - it does stay up, but wildly inaccurate.
i have tried reducing the THR_ALT_P value to no avail, doesnt seem to change much. i even added some small D to the thrust rate, still nothing.
Alas, I have a new APM on the way. While my 6-7V spikes may have warranted an RMA, i think my crashes may have caused some damage too.
Thanks to Olivier and Bill for the help.
Ok, so i got a voltmeter on the input rail of my apm board, and it reads 5.00V. is it possible that my logs are simply incorrect? maybe a datalogging error?
I think the problems I have been having may have been two fold, with some caused by my futaba radio and incompatibilities with the APM. So, i updated the ppm encoder even though i am using 8x pwm connections, this seems to have made the spikes on my roll, pitch and yaw IN channels disappear.
I also erased the eeprom, and compiled arducopter 3.0rc5. after a short low-altitude test flight, it seems that my twitches have gone away, but i am still concerned about the condition of my hardware.
at least my columns not appear to be giving the correct data. and whats more, no more 6V reported in the logs. BUT - the voltage does seem to go up and down a little bit - from 4.9V to 5.15V. is this normal?
should i get a new APM? this really blows. the good news is that it looks like the APM's come with the power module by default now.
IMAG0413.jpg
2013-06-09 13-48 3.log
Hi Ted. Sorry to hear about your crashes.
Seems like your APM is powered with more than 6.1V with a spike to 6.9V. We should have 5V here.
Could you check that ? This could have destroyed your new APM. Do you have a custom voltage / current sensor for the Lipo ?
Olivier
ok, i finally got the log download window to behave. still dont know why i got that error or the garbage characters.something is till messed up, and im starting to suspect my APM.
i have attached my bench test log file. you can see in the PITCH IN column (i assume that is pitch in; im having another problem where the logs dont properly display the right columns) that the data goes crazy.
2013-06-08 01-42 3.log