Horizon off alittle...
- Thread starter meissner58
- Start date
I dunno, are we? It's off significantly. Table is level but it's maybe 1/4 to 1/2 bubble off at the props. Maybe those are the people who are having trouble with level horizons: the ones with "uneven legs" end up with bad calibrations.
Mike
Oso
Premium Pilot
There are some here who have posted above leveling with the motor hubs using a piece of glass or something similar across all four motors.
BTW I also looked at the leveling based on the arms and that matched the motors. Looks like I have uneven legs.
Mike
Mike
- Joined
- Mar 10, 2016
- Messages
- 224
- Reaction score
- 29
- Age
- 54
Well it's still in the mid 40,s in Wisconsin today . Probably a good day to do it outside
Yes, that chill likely will, but see here:Well it's still in the mid 40,s in Wisconsin today . Probably a good day to do it outside
Tilted Horizon 100% perfect, fix after vary cold IMU calibration
Colder the better...
RedHotPoker
- Joined
- Mar 10, 2016
- Messages
- 224
- Reaction score
- 29
- Age
- 54
- Joined
- Apr 18, 2016
- Messages
- 287
- Reaction score
- 159
- Age
- 57
The gimbal is using the compass to keep the horizontal level. Slight changes in the magnetic field may cause slight tilts - these will invert when changing the heading 180 degrees.
A sudden and significant horizon tilt while flying might be a warning sign of a compass alignment issue and I would recommend bringing back the bird immediately and land for a new compass calibration, especially during a distance flight beyond visual range. You don't want to lose both the compass/GPS and the video at the same time.
//Tom
A sudden and significant horizon tilt while flying might be a warning sign of a compass alignment issue and I would recommend bringing back the bird immediately and land for a new compass calibration, especially during a distance flight beyond visual range. You don't want to lose both the compass/GPS and the video at the same time.
//Tom
- Joined
- Apr 18, 2016
- Messages
- 287
- Reaction score
- 159
- Age
- 57
Most Phantoms do. If nothing else, that little foam pad under the landing gears may be of different thickness. It is better to calibrate using the prop shafts - remember, the drone is effectively "hanging" from the props in flight, so that's a better horizontal reference.
Before calibrating the IMU, I just remove the propellers, place a glass pane (from an old IKEA picture frame) on top of the prop shafts, place a bubble level on top of the glass, and then use sheets of paper under the landing gear legs to make sure that the bubble level is absolutely horizontal in every direction.
//Tom
I have calibrated my gimbal on uneven ground, and it still kept a parallel horizon. But a crooked IMU calibration, not so much. Hahaha
Ah, but getting it 'perfectly' level is nearly impossible. As close as humanly possible will suffice. ;-)
Having a trouble free drone, is the best starting point, to a good calibration.
RedHotPoker
Ah, but getting it 'perfectly' level is nearly impossible. As close as humanly possible will suffice. ;-)
Having a trouble free drone, is the best starting point, to a good calibration.
RedHotPoker
Last edited:
Just ran a cold IMU calibration: P3P sitting next to A/C vent, facing due north. Then a gimbal calibration immediately after. When I started (after about 30 minutes with it sitting near the vent), the case on the P3P read 61F on the far edge (away from the vent). I'll go out in an hour or so, recalibrate the compass, and run a test and let you know. One interesting thing I noticed: I had the app open already so I'd be ready to click the IMU calibration as soon as I turned the P3P on. Turned the P3P on, waited for the gimbal calibration button to be active (and the numbers in the table to show up), and clicked the IMU cal button. It calibrated it and the text message eventually said IMU calibration success, and the numbers in the table were zeroed out, proving it had done its job. BUT... the progress meter stayed blank (at zero percent) the whole time. That might just be a bug where the progress meter doesn't update if you start the calibration that early, but thought I'd mention it. On Android and the latest Go app.
Mike
Mike
Now go fly it. See what the differences may be. Hopefully you are pleased and the horizon is square to the ground.
Good luck, it should be fine now.
RedHotPoker
Good luck, it should be fine now.
RedHotPoker
Itching to... RC is charging.
I neglected to mention the most important part. I've done this same process multiple times, trying to fix the horizon (to no avail), but these are the things I changed this time, hoping it'll make some difference:- Cold calibration: did the others at room temp (76F)
- Leveling was based on the motors (props) and not the legs sitting on a level table
- Faced it north this time (why not, figured I'd hit all the points I've read in one shot)
- Moved the RC about 15 feet from the AC to reduce any RF interference with the IMU board
I took pictures of the setup while calibrating. I'll post those along with shots of the horizon later. Best I've gotten before this is abot .25 to .5 degree accuracy: and that was after doing a gimbal calibration mid flight/hovering. I'm hoping this will fix it for good.
Mike
Funny the RC always takes the longest to recharge. I also need to go fly down a couple or three batteries myself today, before the preliminary bouts begin.
So, I hope this cold IMU cal., was the fix you needed. Sweet.
If not, you may utilize the fridge next. Hahaha
RF? Electrical interference only bothers your compass.
Ok, now it's time for me to chill... ;-)
RedHotPoker
So, I hope this cold IMU cal., was the fix you needed. Sweet.
If not, you may utilize the fridge next. Hahaha
RF? Electrical interference only bothers your compass.
Ok, now it's time for me to chill... ;-)
RedHotPoker
OK, the results are in. Bottom line: closest so far: if it stays this good, I consider it FIXED!
Here's the setup showing how I cooled it and leveled it. I left the props on because I just don't like taking them off and retightening them a lot as I'm concerned what that does to the threads over time:
And the result, pointing N, W, S, and then E:
About as close as you can get. That last one (pointing east) has a minute clockwise rotation, but that's really splitting hairs. I also took a 7 minute video while flying around and other than the type of hair splitting above, it was level the entire time with me flying about 1000 feet from home and doing a lot of yawing.
Here's hoping it stays that way! Oh, and even though mine was more of a "cool" IMU calibration, the warmup time is now something on the order of 10 seconds. It was like 2 minutes before. So that's a huge improvement!
Mike
Here's the setup showing how I cooled it and leveled it. I left the props on because I just don't like taking them off and retightening them a lot as I'm concerned what that does to the threads over time:
And the result, pointing N, W, S, and then E:
About as close as you can get. That last one (pointing east) has a minute clockwise rotation, but that's really splitting hairs. I also took a 7 minute video while flying around and other than the type of hair splitting above, it was level the entire time with me flying about 1000 feet from home and doing a lot of yawing.
Here's hoping it stays that way! Oh, and even though mine was more of a "cool" IMU calibration, the warmup time is now something on the order of 10 seconds. It was like 2 minutes before. So that's a huge improvement!
Mike
Oso
Premium Pilot
The drifting horizon issue is mean to people. Like I noted earlier, it gives you false hope and then it pulls the rug out from under you. I think it even laughs from time to time.
Yours looks good. I have my fingers crossed for you.
Yup, looking good bud. As good as it needs to be. Ha
Glad the chilled bird did the trick for you, as it has for many.
Yes, the cold calibration seems to also defeat the initial warm-up time. ;-)
Enjoy your drone and the awesome pics and videos it provides us.
RedHotPoker
Glad the chilled bird did the trick for you, as it has for many.
Yes, the cold calibration seems to also defeat the initial warm-up time. ;-)
Enjoy your drone and the awesome pics and videos it provides us.
RedHotPoker
Hmm. Just measured all the shots (90 degrees is perfect), in order posted above:
N: 90.1 (0.1 degrees CW)
E: 90.2 (0.2 degrees CW)
S: 89.5 (0.5 degrees CCW)
W: 89.7 (0.3 degrees CCW)
Funny how the east shot looks like it might be off the most (again, splitting hairs) but the south shot is actually off the most. Probably an optical illusion based on terrain on the ground, roads, etc. Anyway, I'm happy with it since I don't think I'd be inclined to "correct" any of them in post at this level of accuracy. I just hope Oso isn't right and future flights don't find me back in the drift.
I did notice that right after yawing and coming to a stop like I did to get all 4 directions, the horizon looked off a little in the FPV display (based on the measurements above, maybe 1 degree) and then after the yaw stopped, it slowly corrected itself. I waited for it to "calm down" after the yawing to take the above shots but I wonder if it looks a little tilted until the yaw follow of the gimbal catches up?
Mike
N: 90.1 (0.1 degrees CW)
E: 90.2 (0.2 degrees CW)
S: 89.5 (0.5 degrees CCW)
W: 89.7 (0.3 degrees CCW)
Funny how the east shot looks like it might be off the most (again, splitting hairs) but the south shot is actually off the most. Probably an optical illusion based on terrain on the ground, roads, etc. Anyway, I'm happy with it since I don't think I'd be inclined to "correct" any of them in post at this level of accuracy. I just hope Oso isn't right and future flights don't find me back in the drift.
I did notice that right after yawing and coming to a stop like I did to get all 4 directions, the horizon looked off a little in the FPV display (based on the measurements above, maybe 1 degree) and then after the yaw stopped, it slowly corrected itself. I waited for it to "calm down" after the yawing to take the above shots but I wonder if it looks a little tilted until the yaw follow of the gimbal catches up?
Mike
Our P3A "calms down" after a short while at hover too, but I'm going to try a cold IMU calibration, just to see it that improves it a little more....
Similar threads
