Originally Posted by miru
Thank you Candu1! I was already scratching my head.
What flashing sequence do you get when it is on the bench (no cardboard)?
I did change the startup in rev 0.10. Now the Arduino program looks at 4 things in the same loop: 1) is the radio ok 2) is FMODE LAND 3) have 10 seconds gone by so I can assume the drone has booted and 4) did the drone program.elf finally quit dumping messages to /dev/ttyPA0. The 'offending' condition is signaled with 1, 2, 3, or 4 blinks. If all 4 conditions are met, then it will quit the loop, get a shell started, change baudrate if neccessary and do the thing. Previous versions checked 1, 2 and 3 in one loop and 4 in the next.
Another short test just to confirm that it is the interaction between the metal desk and the ardrone/Rx/arduino, and nothing to do with the RF signal. I simply swap position between the tx and the ardrone - ie, put the tx on the metal desk at the exact spot where the ardone was and put the ardrone at the table where the tx was. As predicted, it started up OK. I also confirmed that the rx LED was on solid to indicate that it is binded and receives signals from the tx, so it's not a matter of the metal desk blocking RF signal from the tx reaching the rx.
OK, the arduino flashes from power up to the end looks exactly the same as a normal startup. ie, Double flash at the beginning, one triple flash, and then 39 or 40 1/2 second slower flashes, then, solid green for a second. At this point, for a normal start up sequence it would flash the ardrone LEDs to double green/red and then turns the LEDs to solid green. But in the failed case, the four ardrone LEDs would switched from solid green to solid red and rest in red. The single LED underneath the ardrone would then turn from solid green to solid red. At this point, any wifi connection is dropped. The arduino LED then turns off for a second then it either comes back with 2 flashes a second continuous flash or comes back with a solid green. No control is available from the Tx with a failed start up.
During the faild startup sequence, I also tested wifi connection to the ardrone wiht my iTouch. I was able to connect while the ardrone's LEDs were solid green. As soon as the ardone LEDs turned red, I would lose wifi connection on the iTouch. I was also using telnet to explore /data subdirectory to see if there were anything obvious in there left over from a previous boot that either mirumod or ardrone would check to prevent a normal start up. Did not find anything obvious...
I was speculating at one point that your onboard program would check for something after control was passed over to it and if it doesn't see some predefined conditions, it would then turn off ardrone's wifi.