I'm running initial flying trials with custom quad, apm2.
So far: Level good, Stabilize good, Alt hold good
Loiter works but starts with small pendulum movements which grow to huge pendulum movements, describing an arc of about 50 ft length. Slight breeze during testing poss 3 mph. Any guidance on what needs adjusting?
You might want to check out the arducopter wiki, there's a simulator over there. I think no one has offered to help because no one wants to be the one that gives you the info, it's wrong and you crash hard. Because I've been asking. So I've been playing with the PID value simulator and tweaking the stock 3drobotics parameters.
I know you need to take into consideration your quad frame type, size, motors in kv, prop size, also from my experience so far... have you set your inclination deviation ? Dont' use manual, find out what it is where you live and set it. My quad performance in loiter improved.
I know there's a lot of reading I've got to do before I have loiter working perfectly with my platform.
Here's my quad specs.. I am using APM2 as well. If it's similar maybe we can work on t together.
Frame Size 540mm
Motors: 26-28a 1200kv
weight w/battery 680g
Try lowering loiter_rate_P. You're accelerating back to center faster than the GPS is updating the position.
Hi Jason, Thanks very much for the guidance, will test this tomorrow!
Hi Jerry, thanks for the info, think if people are afraid of crashing then they should choose a different hobby lol. safest policy is to fly in long grass, it's saved me a huge amount of work so far.
I'm glad to share results etc, although not sure what help it might be as my quest is slightly different than yours! I have built a custom Hex, this is the target airframe, it's a bit over engineered (lot of build effort) so had second thoughts of using it for initial testing, so built a quick quad.
Frame size 800mm
props 10 X 4.5
esc 30a flashed for high refresh rate
using 3S 5000mah battery (can carry two)
Sonar not yet fitted
Weight with battery 1.54 Kg
So far, impressed how accurate the barometric pressure sensor is, compared with apm1.
This info is good to know! I'm actually testing loiter mode this weekend.
Jason I have the same prop size, 1200kv motors, same size esc running high refresh rate.
I will be getting a larger battery once I master flying this. I only have a 2200ma battery right now.
So once I get a larger battery I'll probably be around the same weight. I know I can carry the load.
I've got plenty of thrust. If I take off 2/4 throttle the quad rockets up to the sky~ Very fast.
I'm using sonar from my parts box... I had a maxbotix LV-EZ01. It's not officiallu supported due to the tighter cone, but it works fine for me. LAND flight more works great.
Let us know how you get on Jerry! I've been out today:
compass heading works good
Simple mode works
Yaw very slow for max rudder stick movement.
Land not full tested, seemed to just want to hover, more testing required.
Ok I can see by the forum posts that the 3 most common problems are:
over sensitive throttle
This is similar to what I'm seeing...
For poor yaw reaction have been increasing yaw rate_P 3 times, hardly no difference as yet.
Over-sensitive throttle, Living with it atm.
Loiter still describes a very large arc, this also occurs when testing RTL, it returns to launch fine, then goes into loiter mode with the associated problem...
Decreased the Loiter rate_P 3 times without any noticeable difference, Seems to be not that sensitive! Wind then came up so testing was abandoned... so flew last two batteries in fun mode ha!
I've got an oversensitive throttle too. I was going to try to compensate by getting a better remote.
I have a Spektrum DX5e remote. It's a great remote for the money, but the only adjustments I can make on the remote are reverse and mix. Nothing else, you get what you get and have to get used to it. I think the you can set limits and dampening on a turnigy. I am going to check on that today after work.
Slow yaw is going to effect me big time since when I fly I depend on yaw to turn.
I use pitch and roll mostly if I want to bank into a turn, climb or dive on an angle. I've
been attempting to use it for compensation in a stationary hover in stable mode but
with sensitive controls it's easy to overcompensate. I use the same remote for my
Snellflight Quadpod and the remote is the "feel" of the remote is the same, sensitive.
I was flying the small quadpod in the back yard, doing laps and figure 8's just using yaw.
Pitch and Roll dead center in the middle the whole time. I was also playing with the
arducopter simulator to see what the ideal values are.
From what I read about autoland and rtl, when you are just about ready to complete the landing
you need to switch to simple mode to shut the motors off. Maybe this is a clue for you.
I think there is a rpm to thrust value you can adjust. I saw that in the simulator. I wonder what happens when you reduce that value ?
thanks for your post. I expect my prop mounts and props to be delivered on thursday. I'll be flying the quad this thursday evening.
More wind and rain coming in here so no more flying today!
I did try alt hold then landed by closing down throttle, worked ok but props rotate until switching to stabilize mode(as per wiki).
I have another similar quad which I fitted with a kk board, followed by a DJi naza, yaw reaction was slow with both controllers, so to experience the same on the apm2 never came as a surprise. Guess the torque reaction effect must be small(so the pid values will have less effect). I've also observed that rotating clockwise is much slower than anticlockwise! Think I will order some larger props just to see if it helps 11" X 4.7
Regarding tx. I use MPX EvoPro. don't know if you've ever used them but programming is by far the most extensive (and easiest) of all tx's out there. For the apm2 quad I'm using futaba ff9 (only because it's on 433mhz LR UHF band) No way of using throttle curve in acro mode, and in simple heli mode you lose the mixers, 5 neccesary for the mode switch.
Could do away with mixers and add a rotary mode switch but have grown to like the 2 & 3 toggle pair, you know instantly what mode your in.
Never tried Turnigy 9x, heard it was quite programmable and excellent value for money.
I got similar overshoot then Loiter.
Tried to change pids, without any changes.
Is that GPS slow position update?
my apm 2 updates quick enough.
Can you describe in "easy understand" language those Loiter and Nav PIDs?
I am little confused about them.
Take your props off and check your motors. It appears that my issue is tied to motor 1. I think it has a bad bearing creating more spin resistance. I will update the forum with my results.
Virgil. I am just starting to learn about this stuff.
The loiter and nav pids were posted in the forum. I don't think it made it to the wiki.. You can look through the source code or forums.