I’m having a bit of trouble with my quadcopter and could use some help. When I fly it in the loiter mode while moving forward (tilting down), the copter seems to lose height and struggles to stop smoothly in the loiter mode (I can feel it while flying).
I’ve done the basic setup and used quiktune to get a basic level of tuning. One thing I’ve noticed is that my quadcopter is experiencing a lot of vibrations. I’m trying to figure out where these vibrations are coming from. I’ve securely hard mounted the pixhawk and the rotor arms seem to be solid and well in place.
Interestingly, when I fly in Althold mode, the quadcopter behaves better compared to the loiter mode. I’ve included the flight logs for reference: 2023-08-27 09-26-25.bin - Google Drive
Could someone please assist me in identifying the issue? Just to provide some context, I’ve designed a V-shaped quadcopter. The angle between the front motor arms is around 68 degrees, and the back arms are at about 44 degrees.
I’m really looking forward to any help or advice you can provide. A big thank you in advance!
Thanks @amilcarlucas for your response. Now it makes sense, there might be some propeller wash effect on the pixhawk which directly lead to the error in barometer reading. I will activate this wind estimation and try to save the Pixhawk from the propeller wash.
I will update you the result with log tomorrow. Thanks!
You need to reduce the vibration levels before doing anything else. They are very high with clipping events. In this log Vibration Compensation was activated; you never want that to happen.
Thanks a lot for your help earlier. I’ve done some tests, and I wanted to share what I found.
In the first test, I moved the pixhawk away from the propeller wind wash area. I placed it higher up and thought this might reduce vibrations, but sadly, it didn’t work as I hoped. Here’s the flight record: Link to Log.
For the second test, I changed things up. Instead of hard mounting the pixhawk directly onto the drone frame, I put it on dampers. And guess what? The vibrations went down a lot! Here’s the flight record for that: Link to Log.
Now I’m wondering why this happened. My thinking is that when I hard-mounted the pixhawk, the vibrations from the motors went straight to the pixhawk’s IMU. But the pixhawk 6C IMU is supposed to be okay with hard mounting, it has vibration isolation system.
I’m worried about using these dampers for flying fast, will it create any issue? They also take up extra space, Actually I don’t want to use them if I don’t have to. Any ideas? Can you suggest something else to reduce vibrations?
I’ve tried to figure out where the vibrations are coming from. Here’s what I’ve noticed:
The front two motors point down, and the back two point up. There’s about 1cm difference in the front and back motors propeller axis.
My drone’s center of gravity (C.G.) is a bit towards the back by 2-3cm because I haven’t attached the payload yet. When I attach it, things should balance out.
The angle between the front motors is 68 degrees, and the back motors’ angle is 44 degrees. This is the first time I’m trying this setup. Not sure if it might cause any issues. What do you think?
The front two motor arms are little flexy, it is not much but if I put more force then I can feel it.
By the way, I haven’t fine-tuned for wind yet. I thought I’d first sort out these vibrations and then do that.
Thanks a bunch! Can’t wait to hear your thoughts on this.
Thanks for the reply @Allister. Actually it is a new frame and I used 6mm carbon fiber plate cutting for the rotor arm. It is quite strong, is there any platform or software you can suggest where I can simulate and check the strength of the frame and also check the rigidity for the vibration? I know FEM can do that, I am looking for any other easy alternative.
I noticed you have it set for Quad-V. Even with the configuration you are describing I would use Quad-X.
The vibe mount is a big improvement. For that frame I think you will have to use it. Just because the IMU is isolated does not mean hard mounting/thin foam tape will work for all craft.
You have output oscillation and I’m wondering where those PID’s came from? The D-term on Pitch is very high. You may want to reset these to default and then set these:
INS_ACCEL_FILTER,10
INS_HNTCH_ENABLE,1
INS_LOG_BAT_MASK,1
INS_LOG_BAT_OPT,4
Perform an AltHold hover flight for a minute or so and let’s see what you have. The notch filter needs to be configured after that.
Thanks @dkemxr for your response.
Actually I have tried with Quad-X also but nothing much changed. Btw, what is quad-V and for what configuration it should be selected?
I will change it to Quad-X again.
Yes! but actually I am planning to mount the pixhawk 90 deg roll configuration, so I don’t know whether this vibe mount be effective in this configuration. Due to weight it will bend in this configuration.
I performed the Quiktune for these PIDs. Yeah! output are oscillating, both roll and pitch but I did the test in windy environment. Anyway, I will reset them to default and perform one more test.
I want to know how do you find that D-term on Pitch is very high? this will really help me in future, I will make a note there.
Yes! I will do that.
Thanks! looking forward to your insightful response.
I have done some testing, I found few things as per my observation and analysis.
When I flown in Althold mode, it was flying very smooth and stable, vibrations also under 10.
Here is the log: 2023-08-30 11-44-18.bin - Google Drive
Actually the drone was drifting away in Althold so I was controlling it to maintain the position, I know it affects the data for notch filter. There was not much wind during test and I don’t know why it is drifting. Is it a issue related to level calibration?
Here is the log for Loiter test: 2023-08-30 11-47-47.bin - Google Drive
Loiter mode was not much stable, The drone was decreasing and increasing altitude (for that I guess I need to do the wind estimation tuning), vibrations were high compared to Althold.
X and Y vibrations are pretty much under control but Z-vibration is still high for a gentle hover flight.
Because it was in AltHold which doesn’t hold position.
Vibration levels are OK.
Notch filter looks good with one minor change. Set INS_HNTCH_FM_RAT, 0.75
Set this back to default for now ATC_THR_MIX_MAN,0.1
Then try Auto Tune.
Yeah! I know, but generally when I fly the drone in Althold in low wind condition, it does not drift much. This one was drifting very fast in low wind.
Yeah! vibration levels look good now, what do you say about using the dampers in roll 90 orientation fix? I don’t think so it is a feasible for long term use. I really don’t want to use the dampers so trying to put some extra support and other ways to reduce the vibrations.
Just want to confirm with you, My front motor’s propeller axis and back motor’s propeller axis are not same. There is around 1-2 cm difference. This is not causing any problem right?
Thanks!
Today Morning, I planned to do the Auto tuning but ended up crashing the drone. I have increased the support to reduce the vibration and it actually worked little good (without dampers) but still the Z-axis vibration is increasing when I am speeding up the drone.
Z-axis vibration are the up & down vibration so I am guessing it should be coming from the motors up and down fluctuation, so I need to give more strength to the motor arms I guess.
Apart from that I figured out the pixhawk mounting plate is not strong enough (I can feel the flex in the plate), I need to make the plate mount more rigid and solid. The point is, it should not move in any direction at all.
I hope you’re well. I’ve been working on my drone, fixed a few broken things from the last crash. I did my best to reduce vibrations by changing the mount plate for the pixhawk (the old one was a bit weak), adding extra supports to the frame, and covering the pixhawk to protect it from propeller gusts. Still vibrations are there, So I guess I need to work more on the frame.
But now, I’m facing some new issues. Motors 3 and 4 seem to be running faster than the others, even though I’ve checked that everything is level and the arms are all in the same angle with no twists. Also, when I put the drone in loiter mode (just hovering), it’s going up and down in altitude.
Here is the test flight log: 2023-09-04 18-24-15.bin - Google Drive
I added a small TFmini-S lidar for front detection and avoidance, but that’s causing some problems too. Sometimes it doesn’t even power up, so I have to change the connection port (TeleM2 or TeleM3) and when it detects the object distance then the drone isn’t stopping or maintaining the right avoidance distance.
I also have a Maxbotix MB1030, but it’s not showing up at all, even though I connected it following the wiki instructions.
I’ve been learning a lot from this experience, and I’m determined to make it work. I really need your guidance and support. I have a lot of respect for your expertise, and I appreciate any suggestions you might have.
Thanks a lot! Let’s see if we can find solutions to these challenges together.
I hope you’re doing well. I’ve resolved most of the issues and temporarily switched back to my old frame while I work on building a new, sturdier one. Additionally, I’ve successfully tested the TF LIDAR sensor, and it’s performing well.
However, I’m currently facing difficulties in connecting my Maxbotic MB1040 LV-MaxSonar-EZ4 sensor. Despite following the instructions in the wiki for the serial connection, the sensor isn’t functioning as expected.
I’ve come across information that the signal from this sensor is inverted and we need to use some kind of signal inverter (I really don’t know much about it).
I would greatly appreciate your guidance in getting this sensor to work. Looking forward to your response.
I hope you’re doing well. I wanted to update you on the changes I’ve made to the frame. I’ve taken steps to enhance its rigidity and reduce vibrations, particularly with a hard-mounted Pixhawk. Everything seems great in terms of stiffness.
However, I’ve encountered a new issue which I have been trying to fix: The 1st and 3rd motors (CCW) are running significantly faster than the 2nd and 4th motors (CW). I’ve tried various solutions, but the problem persists. Typically, such an issue arises when motors are mounted at a tilt, but I’ve double-checked the motors, and all seem fine. I even swapped the motor-mounted arms (1st with 3rd and 2nd with 4th), but the imbalance remains (CCW faster than CW).
I’ve also swapped out the ESCs, balanced the center of gravity, and carefully calibrated the sensors. Unfortunately, none of that has solved the problem.
I’m really hoping you might have some suggestions. Your advice has been super helpful to me in the past. Thanks! Really looking forward to your response.
Thanks for the response, Yes! (1 & 2 for CCW) I was also thinking the same but I thoroughly checked the motors mounts (all are mounted proper), checked the motor arms (No twist noticed) and used the new propellers. Still the same issue, nothing changed.
Can you recommend any setup to check the motor arm level? I am using the small leveler but I think that does not work well in checking the motors arms level. How do you check your motor arms and motor mount level?