Rapid altitude change on PH/CRUISE

[DzikuVx]

I'm not sure if it's connected with recent GPS glitches from #431 or #346 or windy day, or new baro or whatever since SD card was loose in logger and I do not have a log from that mission.

Build based on 27822a5 , flying PH / CRUISE mission when suddenly FC decided that it is at least 15 meters lower than expected and started rapid climb. Climb was aborted after throttle movement. When landed, altitude drift was about 11m up.

Also, I've noticed, that enabling ALT_HOLD causes constant descend for few second before it settles down and works like expected again

[wilco1967]

BTW: isn't there anything I could do to filter out those resonance frequencies by software?

I remember few years ago, on MultiWii, there was an issue with vibrations causing the ACC to exceed the maximum range, which clipped the value, so the average/smoothened value seemed lower than it actually was (the highest/clipped values never made it to the software and where ignored). This was causing all kinds of weird behaviour. They changes the ACC range from 2G to 8G (if I remember correctly), and the problem was gone.....

Every copter has vibration.... just have to minimize it, and deal with the remainder....

[bk79]

Could be electrical issue...try putting a 25v 2200uF cap before the fc input and one before the bec that should cut any noise coming if the problem is that..little expense could worth the try..one more thing on the bench di you try increasing the throttle with props?

Il 20/ago/2016 17:31, "wilco1967" notifications@github.com ha scritto:

That is what I've been doing...... On the bench it seems to have no effect.....

But I tried something else just now.... fly it in position hold, but NOT in alt-hold.... Slowly increase the throttle and let it climb faster and faster...... and indeed, I see the ACC-Z go crazy slightly above 1650 (It hovers at 1600 on a 3S)....

here is the log..... LOG00214.TXT https://github.com/iNavFlight/inav/files/428386/LOG00214.TXT

Now replacing motors is kind of an expensive option.... perhaps (just perhaps) new motors will solve it, but after one crash, we're probably back were we started, and my wife won't let me replace motors every time.... ;-)

I have some broken arms (well, not me, but the copter ;-). perhaps shorter arms will shift the resonance frequency somehow...

I guess, this is a downside of the otherwise perfect F3FC board.... impossible to isolate the FC from the frame.... the FC IS the frame....

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/iNavFlight/inav/issues/461#issuecomment-241206446, or mute the thread https://github.com/notifications/unsubscribe-auth/ALSWinSTpfGgy0b6Xmqaszwpi2cAEIJwks5qhx3GgaJpZM4Jj2Su .

[wilco1967]

yes, with props.... ;-) and also without, to see what the motor/shaft itself was causing.... negible....

The powersupply is part of the FC board.... but I'll see if I can fit a cap to the 5V supply anyway.... just so we can exclude it..... (hope I can find one...)

[bk79]

Resonance is by nature a dangerous off work condition for any rotating equipment...you can't simply run in resonance it crumbles everything but you can shift it...hardening the structure somehow to make it stiff or dampening by appying some rubber o-ring between motors and arms...one more thing is changing props pitch. I've move from 5045 3 leafs to 5030 3 leafs and had 1/10 of the previous vibration pk-pk

Il 20/ago/2016 17:36, "wilco1967" notifications@github.com ha scritto:

BTW: isn't there anything I could do to filter out those resonance frequencies by software?

I remember few years ago, on MultiWii, there was an issue with vibrations causing the ACC to exceed the maximum range, which clipped the value, so the average/smoothened value seemed lower than it actually was (the highest/clipped values never made it to the software and where ignored). This was causing all kinds of weird behaviour. They changes the ACC range from 2G to 8G (if I remember correctly), and the problem was gone.....

Every copter has vibration.... just have to minimize it, and deal with the remainder....

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/iNavFlight/inav/issues/461#issuecomment-241206715, or mute the thread https://github.com/notifications/unsubscribe-auth/ALSWihsIRoM7ZeM1LFn_kQQLhe2Ui42qks5qhx8TgaJpZM4Jj2Su .

[bk79]

And when it was on the bench with props on increasing throttle had no acc-z issue on the reading?

Il 20/ago/2016 17:42, "wilco1967" notifications@github.com ha scritto:

yes, with props.... ;-) and also without, to see what the motor/shaft itself was causing.... negible....

The powersupply is part of the FC board.... but I'll see if I can fit a cap to the 5V supply anyway.... just so we can exclude it..... (hope I can find one...)

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/iNavFlight/inav/issues/461#issuecomment-241206984, or mute the thread https://github.com/notifications/unsubscribe-auth/ALSWitk8AEfU7ClFnc8lkEAGYnG4B6vVks5qhyBWgaJpZM4Jj2Su .

[wilco1967]

Not while I was holding it...... BUT: I tied it down using a few tyraps to a large stone, and now it DOES show up.....

Even without props.....

It didn't show up when I held it down by hand, or by holding it from the bottom..... Now while tied down, it does, and very clearly......

Really surprised me it wasn't visible before....

Now let's try some more if I can get rid of it.....

EDIT: it seems to be my motor 3 (left)..... The worst vibrations always seemed to come from 2, so I always concentrated on 2.... but that seems to have hardly any effect.... Motor 3 however, even without a prop, can make ACC-Z go full scale at certain speeds, but it doesn't 'feel' like it is so bad.... Very strange behaviour..... I fly copters for like 5 years already, but they still surprise me....

[bk79]

Ok for sure it's noise..when you hold it in your hand your probably absorbing part of it try using those bypass caps

Il 20/ago/2016 18:05, "wilco1967" notifications@github.com ha scritto:

Not while I was holding it...... BUT: I tied it down using a few tyraps to a large stone, and now it DOES show up.....

Even without props.....

It didn't show up when I held it down by hand, or by holding it from the bottom..... Now while tied down, it does, and very clearly......

Really surprised me it wasn't visible before....

Now let's try some more if I can get rid of it.....

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/iNavFlight/inav/issues/461#issuecomment-241208269, or mute the thread https://github.com/notifications/unsubscribe-auth/ALSWiiNX39XP0ruqLaxI5ZKzbHUMyOEFks5qhyXRgaJpZM4Jj2Su .

[wilco1967]

yep, sure looks like it......

btw, without a prop, it's MUCH worse than with a prop.....

Adding a cap (470, the biggest I got), on the 5V line makes absolutely no difference.... But just turning the motor 90 deg on the boom does.... still has a nasty point which I can't get away by adding pieces of tape / rotating the prop on the motor / etc. can solve....

upto 1670, ACC graph smooth as silk. 1700-1730 > ACC_Z drops > 0.4, RMS 0.6 !!! > 1750 smooth again..... ACC X/Y hardly move at all.... other motors, hardly anyhing showing.

I have another identical motor, which I removed earlier because of vibrations... might actually be better.... lets try.

OK..... swapped the motor.... but STILL vibrations, just slightly higher speed (might just be the battery getting lower). Without the prop, this motor is smooth..... (the previous one absolutely not...). start to suspect the left arm itself having a critical vibration around 1700

Putting the motor closer to the frame doesn't seem to make much of a difference.

Tried flying it again, but not much better..... Next step will be replacement of the ESC (just in case it is electrical...), and replacement of the arm.... might put in a wooden one, just for test (wood is a great vibration absorber).

[digitalentity]

@wilco1967 good news! Well, sort of good. I've confirmed the vibration issue on my QAVR. Above certain level of throttle accelerometer starts to read anything but true acceleration. This is quite a normal flight w/o AltHold (manual throttle) and above 1450 throttle accelerometer is everywhere from -1.5G to +2.2G:

I'm going to find a few o-rings and test if they help. I'll also try with a 1000uF caps on 5V supply and probably on the battery as well. Let's see if it's related to mechanical or electrical noise.

[digitalentity]

Please see #486 for discussing acceleration issue

[digitalentity]

Closing as cause is known and separate issue is existing to fight the cause.