I got some decent flight time this afternoon! I'm uploading the log file from my APM 2.0 controller.
I had a blast today. thanks to all who've made this platform possible. I really like it. I just ordered a Turnigy 6000ma LiPo and my landing gear/gopro mount from XproHeli. Can't wait!
Here's a link to my Quad's flight. Sorry, it' not POV, my wife manned the video cam:
I attached my flight log!
I hope some of you can help me improve my PIDs for LOITER mode. currently when I loiter I travel in a 10ft circle. You can see this in my video. Also the LiPo I have is 2200ma. So stable flight time is around
7 minutes, after that.. well landing is not going to be easy :) I am posting my flight log and current
PID settings. My loiter params started with the params for the 3D Robotics platform, here's the
stab ang rate P = .1
stab cont P = 4.5
yaw ang P = .15
yaw stab P = 7
Alt err P = .5
Thrust rate P = .2
Loiter ang rate P = 2.5
Loiter speed P = .47
My little 450mm platform appears to be very stable in stabilize mode! I can control the location
with yaw. It would be cool to nail down loiter, then I can record waypoints of where I've been and
execute a replay flight.
The ultimate goal is to get this puppy stable to fly along a seashore and video it.
Can someone confirm form me that my max altitude was @ 130ft ?
This is awesome! Here's my logfile from today's flight:
Loiter is much the same as mine was, that swing in my case was faster and more aggressive.
Been high winds and rain here for three days, managed to get out briefly and found loiter in wind was much more stable,(within a 10ft box) not sure if it was the wind or the many tweaks that was responsible for this. Could have been the change to manual mag declination...
Your video at 3.10 was that an auto land or just flying practice?
Anyway looking promising!
The park is located near Mt. Hope Bay. Located here:
The wind was calm inland, the field where we were the wind was about 5knots from the north with gusts to 10 knots. I did notice the motor at location 1 (front left with front facing me) took more power to spin up. While listening to the motors closely with the props off and all motor the motors at 40% It appeared that 3 motors sounded similar, the one on the left front had a bad bearing noise maybe once every 3 seconds. This motor is cheap at HK so I ordered two more. I'm going to try to get info on how to properly disassemble this motor and replace the bearings with some American made bearings. That's one thing we still manufacture extremely well. I think this motor is made well, but the bearings don't hold up.
The plan is to swap the motor out with another one, learn how to take the motor apart and replace the bearing with high quality bearings.
My Spektrum DX5e has only a 2 position toggle switch. I only have STABALIZE and LOITER available from channel 5.
What you was was a switch to stabalize more and me just learning how to fly.
Yes heard about poor bearings in the cheap hk motors, I keep an eye on them, but so far the motors have not given any problems!
Got out again this morning, light winds, Stabize is so good! perhaps need backing off a bit, noticed slight rapid oscillation on roll.
Alt hold still playing up, it decends about half the normal fall rate (which is an improvement) might call it fast autoland.
If I increase the throttle then alt_hold works fine on baro. very stable (within a meter I'd say)
Loiter now is staying within a ten foot box, yet it's still swinging (like yours was)
Have just added Sonar but seems quite unresponsive, putting my hand under the sonar while in alt_hold doesn't do anything until I get to within about 2 inches then it will lift slightly, need to look into that.
The main problem of Throttle remains though and has caused me to go in twice already. Having adjusted throttle to stay up in Alt_hold, switching back to stabilise I get a big throttle boost, which if I have slight forward/aft pitch at the time then it will loop, not good if your only 12 ft off the ground...
So generally I'm happy with 2.6, apart from the throttle, it just needs fine tuning.
I tried some tests while hand holding the quad, I added more than the normal throttle, but held it in place as if it was in hover, then when switching to Alt_hold there was no decrease in motor speed, so I think Mine is hovering with the throttle position below the new threshhold used for alt_hold in 2.6
Think I'll try to implement the old throttle code into 2.6, used avr studio but not familiar with arduino IDE.
I'd like to know if you manage to reduce that swing in Loiter!
Peter, sonar is not going to detect objects closer than 2". The reason why is the receiver has a inhibit signal to prevent false echos from ringing. Closer than two inches the sonar is deaf. The quads fly so the 2" limitation isn't an issue for me. I tested sonar by running with the props off, pointing at walls, furniture , etc, then graphing the log. It looked pretty accurate to me.
Jerry, thanks for the sonar info! will look into this once the throttle is sorted..
Nice work, but you're right, it does need a bit of tweaking. Though I'm not sure that adjusting parameters will help, as I think your quad is ever-so-slightly out of level. I can see that lift-off isn't exactly straight up. I know it sounds silly, and I'm sure that you have leveled it, but this is typical performance from one that is not perfect. Can you try an in-flight leveling? Just a thought.
Hi John. I ran the level routing on my workbench. I checked my bench with a level, it's level. But you are correct the lift-off wasn't exactly straight. I believe the problem is with motor 1. Motor 1 bearings don't sound right, once every 2 seconds or so you can hear a bearing vibrate inside. All the other motors sound fine. Also at startup it's the last motor to spin up at speed, so if I want to take off I need to throttle up until they are all spinning and then take off. When I am airborne I am ok until it's just about time to land. If I limit my flights to 4 minutes I can land ok, but need to be carefull at takeoff. I ordered 2 more 28-25a motors since they are cheap .. $15.00 each. When I swap it out I'm going to see if I can put good old bearings made in the USA in this puppy. hey, it's one thing the USA still makes, high quality ball bearings. I hope I can get buy using a vice to press the new bearing in. I'll update the forums to let everyone know whats involved. I think if we can improve the bearings in the NTM motors, then they would be an excellent source. If we can't then you need to order more motors than you need because it looks like you are going to get one that their QA department didn't catch. If you know what you are listening for you can hear this bearing fail.
My transmitter only has a 2 position toggle switch for channel 5. so right now I can only select stabilize and loiter. I have a 9 channel turnigy on the way that I'm going to do the 7 postion switch mod to.
How do you execute in-flight leveling ?
Also again.. I really like this project. look at my sonar ? It looks pretty clean! I did soar to 130ft quickly and
wanted to keep an eye on loiter and altitude hold to get an idea of how much power I had left so it appears
the sonar is pretty clean right ? I'm using a LV-EZ01. It's not officially blessed by the group, but I had it in my parts box so I figured I'd give it a try.
After I replace the motor and check again.. if loiter is still in a 10' - 15' circle I will attempt to reduce Loiter Speed.
to see if the circle gets smaller or larger. If that has no effect then I'll reduce Loiter Angular rate.
I have a 6000ma Turnigy 3S 25C-50C discharge rate battery on the way along with 2 replacement motors.
John. I am going to get a level and look for a better vibration are to run the level utility again before I try the new motor. There's nothing like experience and you have said you've seen this before. I figure the info would be helpful to add to the wiki. Maybe leveling is off and motor 1 is the location that the code detected as the highest point so the motor don't turn as fast. This is definitely worth checking, thanks for pointing this out.
The other place worth checking is the apm board itself to be sure its perfectly level.
This flash simulator by Jason helped me a lot to tune my loiter parameters. Take a look at it if you haven't done so: