in my basic x AP ive been using a whatchdog timer in place of a failsafe and it has worked fine . i have even had bad gps , waypoint data ect and have had no problems recovering manual control . would use of a watchdog timer on the arduino project help keep down cost?
You need to be a member of diydrones to add comments!
"too safe" what the heck is that ? never heard of it . the point is this can avoid cascading effects . a glitched out CPU can take out your RX .Ive seen that on the bench !!! that's why i don't like a mux alone .that's the thing i think has been missed. that's why i went through the trouble. that is what I've found in my testing and sim time . which fyi has been extensive . i would like someday to start selling my APs and i plan on getting CE and UL approval some day so "mars mission" is exactly what i need . has anyone else out there played with any other type of error handling other then the three we have mentioned ?it is definitely an extensive subject and important part of any AP project .
No, I think you still need a failsafe. Failures can range from software bugs that either freeze the CPU or send wild commands to the servos, to hardware failure with connection or power loss. I can't imagine a timer that could accommodate all of those. That's why we always use a hardware failsafe with redundant power and no dependency on the CPU.
Replies