Vector Waypoint Usage and Support - Page 4 - RC Groups
Thread Tools
This thread is privately moderated by JohnET, who may elect to delete unwanted replies.
Apr 19, 2016, 03:13 AM
Registered User
evo101's Avatar
hopefully in the next month or two. i'm also having the same issue. just have to find flat ground in the mean time

i have to say my plane follows the set waypoints with excellent precession so far



Quote:
Originally Posted by TxLooper
Any Idea when the height restriction will be removed? I fly in the mountains and the 400' restriction really limits my ability to fly by waypoints.
Sign up now
to remove ads between posts
Apr 19, 2016, 11:57 AM
Senior Member
Paul Penney's Avatar
Waypoints

Here is my feedback after trying my airplane using waypoints:

They worked great! My airplane flew a series of about 36 waypoints with no problems, along with height changes and all. I have a couple questions and points of feedback:

CORNER RADIUS - The main thing I think could greatly improve the experience: I am an airplane flyer. When the model approaches a waypoint on the map, it is flying at a set speed and has to maintain that inertia. The problem is that when it reaches the point on the map and the turn is very sharp, the model overshoots the line on the map and has to overcompensate in the turn in order to get back on course, creating a zig zagging effect at every sharp corner. The problem is that if I have the gains set to produce a nice smooth turn without being too jerky, it doesn’t have the gain required to get it back on course and ends up waggling quite a bit. What would solve this problem is if in the software side, we had an option to have a corner radius at each point, so the model would follow the line and make a nice smooth turn. It would be great if we could specify a corner radius at each waypoint to ensure the turns the airplane makes would be nice and smooth, without having to waggle to find the course again.

Heres my question: If I have a pitot tube airspeed indicator, will the model use that reading when maintaining a minimum airspeed?

Thanks

Paul
Apr 19, 2016, 04:52 PM
Registered User
Quote:
Originally Posted by Paul Penney
Waypoints

Heres my question: If I have a pitot tube airspeed indicator, will the model use that reading when maintaining a minimum airspeed?

Thanks

Paul
Only minimum ground speed is used at the moment which is okay for multicopters but no good for fixed wing, you need to keep the minimum ground speed high enough to avoid a stall with a strong tailwind. ET are working on it though - quote from John ET:

"We are working on having closed loop speed control that will take advantage of an airspeed sensor. No ETA but it is in the works."
Apr 19, 2016, 08:11 PM
Registered User
evo101's Avatar
one still needs to be careful with this eg if airspeed is set to 20kph and you hit a 20kph head wind you plane will not be moving across the ground


Quote:
Originally Posted by Turbiner
Only minimum ground speed is used at the moment which is okay for multicopters but no good for fixed wing, you need to keep the minimum ground speed high enough to avoid a stall with a strong tailwind. ET are working on it though - quote from John ET:

"We are working on having closed loop speed control that will take advantage of an airspeed sensor. No ETA but it is in the works."

Hi Paul
i didn't notice this on my tests over the weekend, is my twin dream doing it also in the video?
having said that i have it doing a loop of the waypoint before proceeding to the next one it appears to be on rails, might be the different planes?


Quote:
Originally Posted by Paul Penney
the model overshoots the line on the map and has to overcompensate in the turn in order to get back on course, creating a zig zagging effect at every sharp corner.

Paul
Eagletree Vector OSD Beta Waypoints Test (2 min 51 sec)
Last edited by evo101; Apr 19, 2016 at 08:23 PM.
Apr 19, 2016, 10:26 PM
Senior Member
Paul Penney's Avatar
Quote:
Originally Posted by evo101
Hi Paul
i didn't notice this on my tests over the weekend, is my twin dream doing it also in the video?
having said that i have it doing a loop of the waypoint before proceeding to the next one it appears to be on rails, might be the different planes?
Because you had your model loiter at every waypoint, you didn't get a good chance to see this effect. Basically, for every turn that is over 60 degrees or so, the model has to work significantly to get back onto the track by overcompensating and then turning slightly in the opposite direction to get back on the course of the line. This effect could be eliminated if there was an option to program in a corner radius in the software, where the model could make a nice smooth, programmed turn in the corner and end up right on track without having to compensate or backtrack.

Of course, this is possibly by hand programming several points into a turn to make it smoother, but it would be great if the software could do this instead.
Apr 20, 2016, 12:31 AM
FPV junkie
m_beeson's Avatar
Quote:
Originally Posted by Paul Penney
Waypoints

Here is my feedback after trying my airplane using waypoints:

They worked great! My airplane flew a series of about 36 waypoints with no problems, along with height changes and all. I have a couple questions and points of feedback:

CORNER RADIUS - The main thing I think could greatly improve the experience: I am an airplane flyer. When the model approaches a waypoint on the map, it is flying at a set speed and has to maintain that inertia. The problem is that when it reaches the point on the map and the turn is very sharp, the model overshoots the line on the map and has to overcompensate in the turn in order to get back on course, creating a zig zagging effect at every sharp corner. The problem is that if I have the gains set to produce a nice smooth turn without being too jerky, it doesn’t have the gain required to get it back on course and ends up waggling quite a bit. What would solve this problem is if in the software side, we had an option to have a corner radius at each point, so the model would follow the line and make a nice smooth turn. It would be great if we could specify a corner radius at each waypoint to ensure the turns the airplane makes would be nice and smooth, without having to waggle to find the course again.

Heres my question: If I have a pitot tube airspeed indicator, will the model use that reading when maintaining a minimum airspeed?

Thanks

Paul

Paul,

I don't have a fixed wing flying at this time, so I haven't flown waypoints on a fixed wing, but I have a couple of thoughts that you can look at.

first, I know that you can adjust the tolerances to the waypoints. You can extent the distance to which it acknowledges that it has reached its' waypoint.

I would have to look at it to quote the exact terms.


also, you can adjust the turn raduis so the fixed wing does not turn too sharply.

This gives you a couple of adjustments to look at.

If you have questions about these two settings, I can look at it, and give you some specifics.

-Mike
Apr 20, 2016, 10:43 AM
Senior Member
Paul Penney's Avatar
Quote:
Originally Posted by m_beeson
Paul,

I don't have a fixed wing flying at this time, so I haven't flown waypoints on a fixed wing, but I have a couple of thoughts that you can look at.

first, I know that you can adjust the tolerances to the waypoints. You can extent the distance to which it acknowledges that it has reached its' waypoint.

I would have to look at it to quote the exact terms.


also, you can adjust the turn raduis so the fixed wing does not turn too sharply.

This gives you a couple of adjustments to look at.

If you have questions about these two settings, I can look at it, and give you some specifics.

-Mike
Mike,

I didn't know you could adjust the tolerances, maybe this is what I need to try.

The problem with adjusting the turn radius is that while it does turn smoother, it also has a hard time locking onto the track because it doesn't have as much banking authority. Maybe I am missing something?

I drew a simple diagram to show the behaviour I am experiencing. In red is the model Path. The green point is the waypoint. The solid blue is the path I am experiencing with the model in relation to the waypoints. The dotted red and blue express my desire to have an automatic fillet in the corners so that the model can stay on course and not have to overshoot the next line, waver, and search to get back on course.
Apr 21, 2016, 09:26 AM
JohnET's Avatar
Hey guys,

I will bring this up to see if there is an easy way to address the course during a turn however, from my personal experience in some military grade SUAS, we had to program the setup and exit paths with additional waypoints to get the exact behavior you are looking for. Perhaps we could automate the process but setting additional waypoints is likely the fix for this behavior. Also, keep in mind that your max pitch and roll limits may be affecting this behavior. Maybe try increasing the max values so it can turn sharper to get on track quicker.

Regards, John
Apr 22, 2016, 11:23 PM
ob1
ob1
Registered User
Pixhawk/APM have had this working correctly for years. Might want to take a look into how they accomplish "approaching" a WP, and how they define the radius distance from the WP that is used to signal "WP has been reached". Also take a look at Autoflight Logic LLC Autopilot doc for how they address this.
Hope this helps.
ob1
Apr 23, 2016, 02:12 PM
Registered User
goojah's Avatar
Quote:
Originally Posted by ob1
Pixhawk/APM have had this working correctly for years. Might want to take a look into how they accomplish "approaching" a WP, and how they define the radius distance from the WP that is used to signal "WP has been reached". Also take a look at Autoflight Logic LLC Autopilot doc for how they address this.
Hope this helps.
ob1
I assume we are talking about spline here?

It is a nice function, and even nice is the choice to face the next way point while doing a spline curve/turn, or follow the route/curve/turn. Very nifty from an Aerial photography or even fpv point of view.
Apr 25, 2016, 05:16 PM
Senior Member
Paul Penney's Avatar
Quote:
Originally Posted by JohnET
Hey guys,

I will bring this up to see if there is an easy way to address the course during a turn however, from my personal experience in some military grade SUAS, we had to program the setup and exit paths with additional waypoints to get the exact behavior you are looking for. Perhaps we could automate the process but setting additional waypoints is likely the fix for this behavior. Also, keep in mind that your max pitch and roll limits may be affecting this behavior. Maybe try increasing the max values so it can turn sharper to get on track quicker.

Regards, John
I think yes, I am looking for the spline operation. Right now I have my bank and pitch angles set for about 37 and 45 degrees.

In these sharp turns, setting the pitch angle to 40 or 45 degrees does get you on track QUICKER, but presents two problems:

1) The extreme bank angle leads to significant height decrease, even with high vertical and pitch gains. I observed losses of 15-20 feet, and the required recovery by the airplane to the set height can lead to situations where the airspeed is reduced / tipstall territory if the wind conditions are correct etc. (at least this is what I observed)

2) Even with a large bank angle the airplane still follows the same waving path, just on a smaller scale. This behaviour also presents challenges if you have a gimbal on the aircraft, because the gimbal then has to compensate for the large and sudden changes in bank and often results in some sort of small anomaly from the airplane being shifted so dramatically.

Are there any plans to allow for more than 50 waypoints in a flight? Or is that a hard limit.

Thanks!

Paul
Apr 28, 2016, 08:40 AM
Senior Member
Paul Penney's Avatar
I managed to get out to the field again and programmed a course without any sharp turns, the results were much better! What a great system. I just got my airspeed pitot indicator, its going to be great when we have a closed loop throttle management system!
May 11, 2016, 10:55 PM
ChrisI
Same sort of error latest version but looks at Google Maps


I have turned in ticket. This worked for almost a week??!!


KG5BBF AMA#954079
May 12, 2016, 07:20 AM
... Always Learning
eBoon's Avatar
Quote:
Originally Posted by JCLs
Same sort of error latest version but looks at Google Maps


I have turned in ticket. This worked for almost a week??!!


KG5BBF AMA#954079


I often get the same error. Have not yet figured it out yet.
May 14, 2016, 02:45 PM
Registered User
hey quick question, i am currently using this software version Vector Software BETA Version 12.07 + Vector Firmware Version 2.13, when i hit the waypoint tab the google map never shows up, i just have a blank white square. i cant do anything. anyone know why??


Thread Tools