Just got this loaded this morning and off to test it.This was the first crash I had that broke something. I have no idea what happened, so I'm hoping someone can give me a few ideas.I have the quad equipped with a point and shoot camera (4oz or so) other than that no changes to the standard frame/electronics. No sonar, no compass. The PID's were set to the default values.Flying in simple mode.I flew the quad to about 5 feet --- it was very stable even in the 5-8 mph wind. Feeling confident in that fact, I took it up to about 10 feet. Then it made a quick 90 degree yaw CCW. It held that for no more than 1 second then it swung back 270 degrees CW --- headed straight for me. I cut the power and it crashed into the driveway breaking 2 landing gear legs.
You need to be a member of diydrones to add comments!
Is there a chance that this is related to the "Return to Land" when the battery hits a specific battery level?
Does anyone know what the voltage trigger level is?
I have been flying with an unknown C rated battery since the beginning (7 months now) It has been cycled at least 100 times. It's a GWS 2200mAh flat rectangular pack. Each time this did occur I was forcing the quad in a forward motion that would require the 2 motors in the back to draw more amps in turn lowering the voltage, if the battery could not keep up with the discharge rate. I just got some new 2200mAh 30C batteries in so I will give them a whirl.
See my post: http://diydrones.com/forum/topics/yaw-and-loiter-issues-2028 There is a video of what it does. It's because my PID's are to high (so i am assuming, it only takes a small bit of forward flight to get it going) but it sounds like what yours is doing. It does this on mine when (STABILIZE_YAW) P = 0.8 and D = 0.16 (double the default). I now am running P = 0.6 and D = 0.12 and it has went away but my yaw is still slower than i like.
I am waiting for my xbee's so i can easily tune this thing.
Just out of curisity, have you adjusted your yaw PID's? Raising mine to high causes the quad to loose rudder and yaw around about 180 degrees or so with a little forward flight.
OK,
I had it happen again today, but not yesterday. Almost the exact same thing, but this time it rotated 180* CW. The difference was where I was flying. It has happened 2 times in my driveway in relatively the same place (under HV powerlines and transformer.) Yesterday was at a park.
Could this be interference with the radio? Could a dead channel cause this to happen? I still had control over the throttle at all times.
Looking over my build I also noticed that I moved the receiver just opposite one of the ESC's --- Could this be causing the problem.
I'm a little worried to fly it again until I understand what is causing it.
Currently using
2.0.26
Spektrum DX6i
AR6100e
Please post your setup output. In the CLI you can do setup, show.. The log would also help.
Make sure your compass was not enabled.
I believe the reason the copter spun around was that your nav_yaw value was easier to reach by going the other way. There is a +- 180 error and if you pass 180 it will swing around. This isn't a bug but the intended behavior. It would at most spun 180, so the best thing to do is to not drop the throttle. The algorithm is from NG, which everyone said works better for them. I don't like it either. I fly in rate controlled yaw mode, which can appear to be buggy if your ESCs or frame are our of whack. But the performance is night and day better.
I have an alternative version from a while back that I can switch it to. Instead of incrementing a target yaw value, it limits the yaw to a max 40* error. some folks were not happy about this one either, but I think it was fine.
I pretty sure simple mode says it always points in the same direction (direction quad was in at startup) for beginners so they dont have to worry about the yaw and can concentrate on just pitch and roll. Sounds like the quick 90 turn made it want to go back but it over corrected maybe due to the 90 being quick. Im not an expert so maybe someone else can chime in.
Replies
Does anyone know,
Is there a chance that this is related to the "Return to Land" when the battery hits a specific battery level?
Does anyone know what the voltage trigger level is?
I have been flying with an unknown C rated battery since the beginning (7 months now) It has been cycled at least 100 times. It's a GWS 2200mAh flat rectangular pack. Each time this did occur I was forcing the quad in a forward motion that would require the 2 motors in the back to draw more amps in turn lowering the voltage, if the battery could not keep up with the discharge rate.
I just got some new 2200mAh 30C batteries in so I will give them a whirl.
Thanks again for all the comments...
See my post: http://diydrones.com/forum/topics/yaw-and-loiter-issues-2028 There is a video of what it does. It's because my PID's are to high (so i am assuming, it only takes a small bit of forward flight to get it going) but it sounds like what yours is doing. It does this on mine when (STABILIZE_YAW) P = 0.8 and D = 0.16 (double the default). I now am running P = 0.6 and D = 0.12 and it has went away but my yaw is still slower than i like.
I am waiting for my xbee's so i can easily tune this thing.
I had it happen again today, but not yesterday. Almost the exact same thing, but this time it rotated 180* CW. The difference was where I was flying. It has happened 2 times in my driveway in relatively the same place (under HV powerlines and transformer.) Yesterday was at a park.
Could this be interference with the radio? Could a dead channel cause this to happen? I still had control over the throttle at all times.
Looking over my build I also noticed that I moved the receiver just opposite one of the ESC's --- Could this be causing the problem.
I'm a little worried to fly it again until I understand what is causing it.
Currently using
2.0.26
Spektrum DX6i
AR6100e
Make sure your compass was not enabled.
I believe the reason the copter spun around was that your nav_yaw value was easier to reach by going the other way. There is a +- 180 error and if you pass 180 it will swing around. This isn't a bug but the intended behavior. It would at most spun 180, so the best thing to do is to not drop the throttle. The algorithm is from NG, which everyone said works better for them. I don't like it either. I fly in rate controlled yaw mode, which can appear to be buggy if your ESCs or frame are our of whack. But the performance is night and day better.
I have an alternative version from a while back that I can switch it to. Instead of incrementing a target yaw value, it limits the yaw to a max 40* error. some folks were not happy about this one either, but I think it was fine.
Sorry about the focus...
I pretty sure simple mode says it always points in the same direction (direction quad was in at startup) for beginners so they dont have to worry about the yaw and can concentrate on just pitch and roll. Sounds like the quick 90 turn made it want to go back but it over corrected maybe due to the 90 being quick. Im not an expert so maybe someone else can chime in.
Neil