i changed my props out because one was developing a crack, they are the same props (10X4.5) that I have been using with good luck. went out to test fly stabilize worked fine, was a little more sensitive so I turned down the P for pitch and roll. When I went to alt hold it did not hold and in the attached log file just came down. before changing the props I have had no issues and have had successful full auto mode flights with this air-frame… I am still working on reading the log files so any help on what is going on would be greatly appreciated.
I think it may be a vibration issue with the new props… I changed back to my original props and everything is fine… but I would still appreciate any suggestions on what to do to get the new props working
It’s almost certainly a vibration issue. You didn’t have your IMU logging enabled, but there’s another good way to check.
Have a look at the attached plot of your BarAlt and RelAlt. BarAlt is the altitude based upon the barometer only. RelAlt is a combination of the barometer and also the accelerometers. These lines should normally be very close. But if you have a vibration issue that’s spoofing the accelerometers you’ll see the RelAlt diverge from the BarAlt. In Alt Hold and Loiter modes the APM uses RelAlt. In Stabilize it uses BarAlt. This is why a copter with vibration problems can fly OK in Stabilize but is erratic in other modes. Which is exactly what you have.
You didn’t mention if you balanced your replacement prop, so that’s a possibility. Or perhaps it wasn’t seated correctly or centered properly. And your props should probably all be of the same model/manufacturer.
The new set of props are the same manufacturer and pitch as the old ones… so there should be no issues there, I will checking the balance and make sure everything is seated properly the retest with vibration loging turned on. this batch of propellers feels stiffer than the first set I bought so they may have more material in them and be farther out of balance than I realized.
[quote=“OtherHand”]It’s almost certainly a vibration issue. You didn’t have your IMU logging enabled, but there’s another good way to check.
Have a look at the attached plot of your BarAlt and RelAlt. BarAlt is the altitude based upon the barometer only. RelAlt is a combination of the barometer and also the accelerometers. These lines should normally be very close. But if you have a vibration issue that’s spoofing the accelerometers you’ll see the RelAlt diverge from the BarAlt. In Alt Hold and Loiter modes the APM uses RelAlt. In Stabilize it uses BarAlt. This is why a copter with vibration problems can fly OK in Stabilize but is erratic in other modes. Which is exactly what you have.
You didn’t mention if you balanced your replacement prop, so that’s a possibility. Or perhaps it wasn’t seated correctly or centered properly. And your props should probably all be of the same model/manufacturer.[/quote]
OtherHand:
Wait, Stabilize mode actually uses some sort of Barometric input? I thought Stabilize was only gyro+accel. This makes no sense to me, it seems like AltHold would be Stabilize (gyro+accel) plus either BarAlt or RelAlt. Not trying to get OT, but can you help me understand this via a link or other treatment of this topic?
[quote=“britdavis”]
OtherHand:
Wait, Stabilize mode actually uses some sort of Barometric input? I thought Stabilize was only gyro+accel. This makes no sense to me, it seems like AltHold would be Stabilize (gyro+accel) plus either BarAlt or RelAlt. Not trying to get OT, but can you help me understand this via a link or other treatment of this topic?[/quote]
Ummmm…I believe you are correct. Lately I’ve been dealing with so many modes and switching between fixed wing and copter that I’ve lost track. What I’d really love to see is some sort of matrix of the various sensors and which flight modes rely on each of them. As you say, I don’t think Stabilize uses any altitude sensing, which is why I’m usually bobbing up and down like crazy. I’m pretty sure, thanks to your reminder, that Stabilize uses only gyros and accels to ensure self leveling. But because it only self levels it should be relatively immune from flyaways caused by accelerometer spoofing.
I balanced and put the new props back on… this time I logged vibrations, things seem a little better but I am still getting unexplained surges.
I have attached the log file from today attempt, I think I may still have too much vibration on the sonar but everything else is looking better to me.