Hello folks:
I’m showing here two Vibration Metrics graphs taken about a year and a half apart. There is a big difference in the numerical magnitude of vibration at these two different times. Both graphs are from a Tarot X6 hexacopter running very similar missions.
Graph 1 from June 2023
Graph 2 from October 2024
The scale of graph 1 is 0 to 0.12 m/s^2.
The scale of graph 2 is 0 to 50 m/s^2.
I can’t imagine that the same drone would have an increase in vibration of more than an order of magnitude. Has there been a change in how flight review interprets the vibration data log? What is the explanation of this big change?
I don’t think its the drone because this is not the only drone where I’ve seen this difference over time.
Thanks,
Joe
Did you update the firmware between these flights? Check out the long discussion about the Cube Orange+ vibe issues here: High vibration in Cube Orange+ - #88 by mwshafer
mwshafer Thanks for the link to the other post on this issue. It covers only the Cube Orange, but there is complete agreement with my data from the Pixhawk 4. The increased vibration started with PX4 v1.13. I’ve done some testing with v1.14 and it has the same high vibration metrics. So I’m wondering if the high reported vibration actually exists, or is it being reported as higher by the firmware?? I fly both F550 and Tarot X6 hexacopters and their performance has not changed dramatically with each upgrade of PX4. Can PX4 cause higher vibration because of changes to the flight dynamics algorithms? That would be very interesting if it does.
Joe
@Joseph_Marrone I don’t know if it is simply an issue with reporting or some scaling of the vibes. I’ve notices high frequency oscillations in hover with PX4 that don’t show up when I load Arducopter. I can’t know for sure if this is a result of the vibe reporting in PX4, but is is suspect that my vibration metrics are so high and that the flight stability isn’t great. If you find out anything new, please make sure to post it here or on the other thread. Thanks.
I’ve been flying 5 drones with PX4 v1.13.3 for the past two years and have not had any flight issues. My drones fly just about the same as they did with v1.12, v1.11, and earlier. But as we have noted, the vibration metrics have gone up by one to two orders of magnitude with v1.13.3. Just doesn’t make any sense to me. The drones have not been changed an any major way. Have my control systems, ESCs and motors deteriorated to such an extent that they are ready to vibrate apart? I don’t think so. The fact that many people have seen the same increase in vibration would indicate that its not the hardware, but the software giving the wrong data. That to me is the most probable explanation. The other possibility is that the firmware is doing something strange to the power system causing it to deliver noisy PWM signals to the ESCs.
Another indication of what I’m saying is that battery usage has not changed with the “increase” in vibration as some have reported. I still get the same flight times as before.
So I’m sticking with the explanation that the data logs are wrong. Somebody changed the scaling in the graphs. If someone can prove me wrong I’ll willingly accept another explanation.
Joe
@Joseph_Marrone I tend to agree with you, but I’ve had continual high frequency low amplitude oscillations in my pitch and roll when using PX4 that I didn’t see with Arducopter. I have a separate post about that issue here: Pitch Setpoint Oscillations in Hover - #2 by mwshafer
I don’t know if that issue is related to vibration, but I wouldn’t be surprised if they were related. Those oscillations occurred after doing an auto tuning, so I don’t think it is a tuning issue unless the autotuner sometimes picks PID values that aren’t ideal. My drone is relatively large (>4kg) so I’m wondering if some custom tuning might be needed.
Hi All
Although vibrations and oscillations tend to be the cause of each other"s effect, i see a marked improvement using Px4 v1.15.3 stable over Px4 v1.14.3 stable.
In fact i hardly could use Px4 v1.14.3 for more than 2 or three times and i had to switch back to Px4 v1.13.3 especially seeing noisy props and very poor battery performance using the same hardware.
But now i am satisfied using Px4 v1.15.3 stable with similar battery performance as with Px4 v1.13.3 or Px4 v1.12.3( more better)
Also i notice that even with a cracked arm adaptor , the px4 v1.15.3 flew with acceptable results.
Guys, i recommend switching to Px4 v1.15.3 stable for CUBE ORANGE owners
