Hi mr.Shaun or mr.dave. Could you please have a look at my bin file for 22 inch H-type quad-copter which I crashed after setting notch filter. I tried increasing its speed to 12 m/s and starts wobbling and crashed. Thank you
The copter had been flying along OK when all of a sudden motor outputs started oscillating - mainly Motors 1 and 2. This probably indicates an issue with Motor 1, either ESC wasnt happy with the low voltage or there was a physical problem like a breakage.
If the ESCs have a low voltage cut-out you might want to disable it.
I’m guessing you have T-Motor Flame ESCs ??
The tuning was not great, but good enough to fly normally - I dont see how attitude control or tuning caused this. There does need to be more tuning work done.
Battery voltage is going suspiciously low and even starts low as if the battery is not fully charged.
You definitely need to set these:
BATT_FS_CRT_ACT,1
BATT_FS_LOW_ACT,2
If there had not been instability and a crash then the battery would have soon hit the low voltage level and the copter would have done nothing - no attempt to save itself.
Regarding the notch filter, it’s not correct but also not the cause of the crash.
Set these for the next flight, when repairs are completed:
LOG_BITMASK,180222
INS_HNTCH_ENABLE,0
INS_HNTC2_ENABLE,0
INS_LOG_BAT_MASK,1
INS_LOG_BAT_OPT,4
and let’s see the results of an ordinary flight in AltHold and Loiter. Just some hovering, gentle movements, circles and ascent/descent.
1 Like
Thank you for your quick response mr.shawn. I will change notch setup. Currently we are checking the drones max. Endurance along side, so that I turned off battery failsafe and I will tun it ON.
While checking the log, I found the vibe z value goes to 27 (which is not good). I will add good damper for it.
Also I tried new PDB (self-made). I connect all my esc signal cables in it and to the FC through the PDB. Could this be a problem.
Also while observing visually the drone starts wobble at a speed of 12 m/s.
I will check my motors as well. Let me know that vibration of this could lead to crash or any other suggestions. Thank you once again
The vibrations were OK, only going high once there was some serious issue.
So I wouldnt change the vibration-damping yet. A CUAV X7 can potentially handle vibrations at the level you see.
Ok mr.shawn
As You suggest that could be a motor issue, I will go with branded PDB and connect esc directly to the FC. Because my motors and escs are in good condition ( I checked them after crash with test rig individually). Also Could you please suggest what are tuning parameters I need to check and tune.
Hi mr.Shawn, I tested my DRONE after repair it flies good and also hovers in alt-hold mode. so that I tried with Auto-tune. I set
AUTOTUNE_AXES - 7
AUTOTUNE_AGGR - 0.1
It shows the below data
but also shows auto-tune failed few times in the flight
So i tried again and the DRONE crashed. the log of crashed flight is as follows.
kindly let me know about the issue with this and i would grateful
Open up the access to that log
It’s open now. Sorry for the issue
Motors 1 and 2 get into an oscillation fight and and very quickly are oscllating between maximum and minimum outputs. There’s no evidence of a desync or a fault like that, but maybe this combination is just not up to the job.
These are the same motor/ESC positions that gave issues in the first crash log.
EDIT: Interestingly, current seems most effected by Motor 2 output so that may be a clue. Motors 1 and 4 hardly affect current even when driven to maximum.
There is some points I am compelled to make and I hope this doesnt sound rude.
- you didnt answer about what ESCs you have, or mention any other related components.
- you didnt provide a log of a simple test flight after making the changes I specified.
- you didnt make all the changes I specified, although they dont contribute to actual flight at this stage
If you did those we may have picked up on some further changes that would be needed.
If you did flights where Autotune failed did you take any steps to fix whatever was preventing Autotune from completing?
I’m not trying to do an “I told you so” - I’d rather never see any crash logs and I understand the costs involved. I’d like to prevent the crashes if we can, and it can take careful review of logs and knowing all the components.
1 Like
I understand mr.shawn. I will made the mentioned changes in the next flight.
here is the list of components i used
- T-motor antigravity MN6007- II
- Flame 60amps 12s ESC
- CUAV X7+ FC
- CUAV CAN PMU
- SIYI mk-32 RC and Receiver
I am using quad-H configuration
And the position of esc’s and motor are same except one. I have replaced one motor (motor 3 on the top left) that because the bolt got broken in the propeller mount
I’m uploading the simple flight log that i did before the second crash as follows
00000478.BIN - Google Drive
may i know how you find that current most effected by motor 2 (im asking to learn)
kindly let me know what should i do before next flight. I don’t want to crash anymore. thank you
1 Like
I simply graphed the battery current and all the motor outputs. This is not definitive though as there may be slight lag in the current measurement (maybe) or those T-Motor ESCs have been known to react badly to over-current spikes that dont appear in the conventional current monitoring. There could also be other conditions like wind or attitude change that mean the motor outputs dont directly correlate with current draw.
In my experience the Antigravity motors are low on torque (or power?) and could possibly be a source of instability since they may not respond well the changes. At least some of the models I’ve seen are like that. They are typically advertised as “long time” flight meaning they draw low current, but they also lack authority and can be difficult to tune and only work in particular weather conditions. So just be aware of that.
One of the things I noticed is your Rate PIDs are quite high, which could be OK, or it could be sign of the control loops trying to get good response out of those motors - unsure at this stage.
I would drop the D terms a bit to make sure those ESCs are not getting too hot:
ATC_RAT_PIT_D,0.012
ATC_RAT_RLL_D,0.001
For the T-Motor Flame ESCs you need these in addition to the related parameters you already have:
ATC_INPUT_TC,0.20
ATC_RAT_PIT_SMAX,50
ATC_RAT_RLL_SMAX,50
ATC_RAT_YAW_SMAX,50
And set these too:
BATT_ARM_VOLT,22.10
BATT_FS_CRT_ACT,1
BATT_FS_LOW_ACT,2
LOG_BITMASK,180222
INS_LOG_BAT_MASK,1
INS_LOG_BAT_OPT,4
What battery pack do you have? It seems very low C or may not be fully charged as the voltage drops fast.
I would consider getting a good 4in1 ESC around 60amps (per ESC, which is how they are rated) or four good single ESCs. Mount them inboard and run long motor wires, short battery wires. BLHELI32 or AM32 → all that sweet data!
When the copter is ready for flight again, just do a test in AltHold then Loiter - some turns, pitch, roll and ascents/descents. No Auto missions or Autotune.
Let’s see how that performs.
1 Like
Hi Shawn, apologies if this is a bit off-topic — I’ll keep it brief.
I’m using the T-Motor U7 490KV. After 4 hours of successful flying with the Hobbywing Skywalker 80A ESC, we had two bad crashes due to ESC failures. I then upgraded to the FLAME 60A 6–12S. Before upgrading, I read most of the concerns about these ESCs in the community here, but i had to buy it as it was the only available choice on the market.
To overcome or reduce the possibility of in-air resets or shutdowns, I reduced the attitude controller gains’ aggressiveness, in addition to changing ATC_INPUT_TC to 0.3.
May you please explain why do you suggest setting *_SMAX parameters. Any other advices to try operate with these ESC safely until we upgrade them will be appreciated. Thanks in advance.
Those SMAX params will help too - they limit the rate of change. You could even go to a smaller value like 30 or even 25, just experiment until attitude control seems slightly affected then increase a fraction - you dont want to compromise the ability to handle disturbances.
3 Likes
To make sure I understand your concern correctly:
You want to ensure that the controller won’t command very fast rate changes that could demand a current rate (dt) higher than the ESC’s threshold, which might trigger a reset — right?
2 Likes
Correct - this is achieved by limiting the PID outputs via those parameters.
2 Likes
So after doing this, do you think it will help overcome the ESC reset issue from a software perspective? From the hardware side, I’ve already installed extra capacitors to reduce voltage ripples, and there are no cold solder joints. Are there other hardware issues to consider?
One last thing — sorry for all the questions. Could you please clarify this? When I reduced the controller aggressiveness, it was still able to cope with disturbances. I kept this in mind while lowering the gains. Is this related to what you were trying to say, or not?”
Hello @randyV and @ganesh538
We developers feel sorry for the crashes and near crashes of your drones, and that is why we spend a lot of time writing code and documentation to help other people make the most out of their drones and avoid crashes.
All the parameters that @xfacta said to change would have already had the correct value if you had used the ArduPilot methodic configurator
The procedures are all there, step-by-step. @randyV @ganesh538 I know you guys have a busy life and do not have time to spare, but it would solve your issues if you would take the time to read it and execute the steps.
Hello,
I have completed most of the necessary configurations by carefully following the ArduPilot Wiki, community discussions, developer videos, code reviews, and by using my own developed dynamic model. I truly appreciate how much valuable knowledge is available — it respects my experience as an engineer with many years in this field, rather than just using a configurator to achieve this.
However, my current issue relates to preventing random FALME ESC resets, which I believe is beyond the normal configuration process.
By the way, asking Shawn about the *_SMAX parameters was to make sure that I understood his point of view correctly, and I do see that there is no need to set them in my case, as my tuning is already optimum without any observed issues.
As Randy and others have noted before, having a fixed support style — rather than engaging with the user’s specific points — is, in my opinion, sometimes not very helpful.
1 Like
Hello @randyV
I did not had that piece of information before. You are an experienced engineer with many years in this field.
If you are following the ArduPilot Wiki, community discussions, developer videos, GitHub PR code reviews you are doing it correctly. Thanks for that and participating in the community discussions.
Most users are however not engineers and do not want to spend time learning lots of information.
This has been my experience from reading over 112000 posts in this forum.
They just want to get the required information to complete their task. And they do need step-by-step procedures.
The comment you cited from @rmackay9 was before he even tested the software. Have you spoken to him recently?
Sure mr.shawn, I will apply all the mentioned params in my next flight after getting new pair of propellers that I ordered. and fly as mentioned and will share the log data. about the 4in1 ESC, I will try to get one and let you know about it