What establishes home?

[Bluegrass]

I was testing out my new Phantom which I upgraded to 3.0. I started the Phantom in on location letting it idle and than walked to another location that had enough clearance from trees to test to see how well it would hover. After that test, I walked another 50 yards or so out into a field & proceeded to do more test flying. After a short time the Phantom started heading back towards where I did the first hover test. I lost complete control of it. I ran following to see where it was going and it flew right past the area where I did the hover test into a tree and feel straight down to the ground.

It was upside down on resting on the props when I got to it. After falling about 15' to 20', the only damage I found was the Zenmuse 3D H3 mount got bent and something shifted to where the yaw gimbal couldn't turn freely. After I thought about it, it was my fault that I started to do the last test when the battery was down below 50%. I guess it was trying to fly home but not sure where home was really established. From now on, I will (1.) make sure I have a full battery charge before I fly and (2.) I know where home is and make sure it is a place the Phantom can make it back to without going through trees or other things like power lines.

My question is and I'm sorry if the subject has been discussed, what establishes where home is when flying the Phantom?

 

[QYV]

EDIT: I wasn't completely accurate in my response so I wanted to get the right answer here. Apparently the rapid green flash is simply an indicator that the Phantom has enough GPS satellites locked to establish a home-point, but it doesn't actually establish that point until you fire up the rotors. So the answer to the question is: starting the rotors after the Phantom has enough satellites for a lock.

It IS possible to redefine the home point, both position and altitude (remember GPS is a 3-dimensional measurement) during flight... I'll let you find that procedure by searching around the forum or reading the online documentation.. it's out there and easy to find

while I'm editing, may as well mention this: GPS isn't really all that exact... according to GPS.gov the "worst case" for non-milspec GPS is only accurate to within 7.8 meters... that's a pretty big variance. Then again, that's WORST case.
usually GPS is accurate to within 3 meters or so... the point being if the Phantom is in RTH mode it's probably not going to try to land at the EXACT position you established home-point, it may be anywhere from 1 to nearly 8 meters away. For this reason most people switch back out of Failsafe/RTH when the Phantom gets close, and complete your catch/landing manually like normal.
If Failsafe is triggered, you can get normal control resumed by flipping the right toggle on your remote (S1) full down and back up. The indicator lights on the P2 which blink rapid yellow during failsafe, should return to green (or red, if you're in battery warning level)

 

[Buckaye]

I was testing out my new Phantom which I upgraded to 3.0. I started the Phantom in on location letting it idle and than walked to another location that had enough clearance from trees to test to see how well it would hover. After that test, I walked another 50 yards or so out into a field & proceeded to do more test flying. After a short time the Phantom started heading back towards where I did the first hover test. I lost complete control of it. I ran following to see where it was going and it flew right past the area where I did the hover test into a tree and feel straight down to the ground.

It was upside down on resting on the props when I got to it. After falling about 15' to 20', the only damage I found was the Zenmuse 3D H3 mount got bent and something shifted to where the yaw gimbal couldn't turn freely. After I thought about it, it was my fault that I started to do the last test when the battery was down below 50%. I guess it was trying to fly home but not sure where home was really established. From now on, I will (1.) make sure I have a full battery charge before I fly and (2.) I know where home is and make sure it is a place the Phantom can make it back to without going through trees or other things like power lines.

My question is and I'm sorry if the subject has been discussed, what establishes where home is when flying the Phantom?

It establishes a home point when it is powered on an gets a GPS Lock. That said... the longer you wait to take off the more satellites it finds and the better the home "lock" is.

In Phantom 2 mode - you can also move the Phantom and flip the S2 switch 5 or 6 times and it will set the homepoint to where you moved it to.

In my experience, however - the homepoint can be off by several feet (even if you have 7+ satellites) - so I tend to flip to ATTI mode when it get's close to home and take over the landing -unless I have a wide area (like a parking lot) where I don't mind if it lands +10 feet in any direction.

when you say you lost complete control... did you flip the switch to ATTI?

 

[QYV]

In my experience, however - the homepoint can be off by several feet (even if you have 7+ satellites) - so I tend to flip to ATTI mode when it get's close to home and take over the landing -unless I have a wide area (like a parking lot) where I don't mind if it lands +10 feet in any direction.

last time I checked and admittedly it's been a little while... non-military GPS was only accurate to a couple meters... which means theoretically it could be several meters off the exact point it took off from.

totally agree with buckeye that unless it's in it's final absolute forced auto-land, you should probably flip the switch and resume manual control for your catch/landing rather than let auto-land land the craft, which could easily result in a tipover/broken props

 

[Bluegrass]

Thanks for your quick replies. I'm really pretty new at this. I calibrated it in the Phantom mode so I don't know that flipping the switch out of GPS mode does anything. I'm afraid to fly it in anything but GPS mode at the time. I must admit that I was in a panic mode when the bird took off and I had no guidance control of it anymore but I soon realized that it must have been trying to fly home. I really don't know whether it had established GPS when I was up near my cabin and first powered it on but if it did, that answers why it was trying to fly through the trees in the yard. Fortunately this was a location out in the boondocks of the Missouri Ozarks so there were no people around. I'll post a clip of the homebound bird and crash when I get a chance.

As I studied the video I am posting a link to, I think that home was established when I did my first hover test which is the first part of the video. I think you might see from where I did the hover test was not very far from being under the tree that the bird crashed into. The second part I think you can see where the Phantom climbed up and then started heading home. I'm sorry I put more video than I wanted to in this sample of my crash. I think it is about 2:35 minutes into the testing that the Phantom started heading home. The one thing that puzzles me is it looks like it started to land in the clearing but on it's way down it started heading forward into the tree.

https://www.youtube.com/watch?v=7x3zyz6 ... e=youtu.be

 

[HarryT]

the Phantom establishes it's Home Point based on it's GPS location when you power it on... the "rapid green flash" sequence (you should be familiar with from your documentation ) tells you it has established a home-point at it's current location.

No, that's not entirely correct. The rapid green flashes tell you that the Phantom has established GPS lock, but the home point is actually recorded as it takes off.

You can test this easily for yourself. Let the Phantom power up and establish GPS lock, and then pick it up and carry it some distance away from the power-up location before taking off. If you force an RTH by switching off the Tx, you'll see that it will return to the take-off location, not the power-up location.

 

[rrmccabe]

Yea I think Harry is right. Once the Phantom has enough satellites to make home lock it says good enough and is done. More satellites does not help.

So its set until you do a manual reset which I usually do.

 

[HarryT]

Nobody has to take my word for it - it's clearly stated in the DJI wiki:

http://wiki.dji.com/en/index.php/Phanto ... Home_Point

Before take-off, the current position of the quad-rotor will be saved as home point when you push the throttle stick for the first time after 6 or more GPS satellites have been found.

 

[Bluegrass]

Thanks guys. If I had shut it down and then walked to the field and started it up again, when the battery reached it's go home point it would have landed within 10 feet of me in an open field. BTW, the Yaw gimbal got bound up but my brother was able to find an allen screw and adjust it so it wasn't binding anymore. I may be back in business if I straighten the mounting bracket that got bent a little. That $37x Zenmuse gimbal is very susceptible to getting damaged from crashes. I think more so than any other part of the Phantom.

Do any of you have any ideas why on it's final decent to land it seems to have decided to go in a forward direction into the tree or is it just my imagination it changed from a straight decent to one moving forward as it was descending.

Anyway, I think I've learned that it's pretty important to establish a good home landing area for emergencies such as I ran into on my first flight with the Phantom. I went on to shoot some videos later over the river. The only problem I saw in the videos was some jerkiness in the yaw because the yaw gimbal was locked and it would twitch once in awhile trying to adjust.

 

[Buckaye]

In my experience, however - the homepoint can be off by several feet (even if you have 7+ satellites) - so I tend to flip to ATTI mode when it get's close to home and take over the landing -unless I have a wide area (like a parking lot) where I don't mind if it lands +10 feet in any direction.

last time I checked and admittedly it's been a little while... non-military GPS was only accurate to a couple meters... which means theoretically it could be several meters off the exact point it took off from.

totally agree with buckeye that unless it's in it's final absolute forced auto-land, you should probably flip the switch and resume manual control for your catch/landing rather than let auto-land land the craft, which could easily result in a tipover/broken props

Yes there used to be a built in degradation of the accuracy by the military because there was worry that the GPS system would be used against us. Back in 2000 (I thought it was earlier... like 1997 but I just looked it up) they got rid of that degradation and the accuracy of consumer GPS accuracy increased 10 fold.

However, there is still some inaccuracy that comes from a number of variables that causes GPS systems to only be accurate to about 10 or so feet.

 

[Buckaye]

Nobody has to take my word for it - it's clearly stated in the DJI wiki:

http://wiki.dji.com/en/index.php/Phanto ... Home_Point

Before take-off, the current position of the quad-rotor will be saved as home point when you push the throttle stick for the first time after 6 or more GPS satellites have been found.

Thanks for the correction. I guess I remembered it wrong - and... I usually take off from where I get the satellite lock so I also probably assumed the sat lock was the trigger

 

[QYV]

No, that's not entirely correct. The rapid green flashes tell you that the Phantom has established GPS lock, but the home point is actually recorded as it takes off.

You can test this easily for yourself. Let the Phantom power up and establish GPS lock, and then pick it up and carry it some distance away from the power-up location before taking off. If you force an RTH by switching off the Tx, you'll see that it will return to the take-off location, not the power-up location.

good to know, thanks. I never DO that I always just set it down in the take off location but still, that's valuable to know. thanks!

 

[HarryT]

I suspect that almost everyone takes off from the power-up location. I did the experiment, though, just to confirm that what the wiki says is right. It is .

 

[Damon]

I suspect that almost everyone takes off from the power-up location. I did the experiment, though, just to confirm that what the wiki says is right. It is .

When I flew yesterday IOSD was blank for distance after powering up and the waiting for course lock and home lock lights. The act of CSC is what registered the distance of 0 on IOSD.

Probably the next test will be power up let c/l & h/l lights. Move 20 meters and do CSC to see if it shows 0 or 20 for distance in IOSD.

Not sure if this is the same for V+

 

[N017RW]

It's an FC (NAZA) based function so it is the same for all P2s.

 

[Damon]

For those that were curious the act of CSC for sure sets the actual home point. Put it in one location powered up till h/l and c/l . CSC set the 0 in my distance on IOSD.

Shut down motors leaving P2 powered and moved it and distanced increased from home point in IOSD.

When CSC was invoked the first home point was still set showing the meters moved from its first CSC command.