Jump at pos/alt hold

[JussiH]

In some cases using vibration dampeners for the board can make it worse - try and remove them. You definately have Acc saturation going on

[bluuu]

thx for hint

[Ruffio]

bluuu, do You have THROTLE on !!! SPAM !!! 1 (in log !!! SPAM !!! 0 - I'm shure that but better ask)?
Its walues jups to 880 (!!!), than to 680, but alt is solid (GPS ALT and prassure).
Than Alt increase, mot_throtle jumps up and down like "to late response".
Look at you radio quality - I think it is not too good.

[bluuu]

Ruffio - yes, channel 0 (in log)
It jumps, but it's "automatic" jump.

Read, once again my description. Link quality is good (it was 20 meters from TX) and after "jump" I switched to manual mode.
I must put all log (about 75MB) with describes.

[Ruffio]

ups, silly me, it is THROTLE then 0-1400 signal lvl, with about 700 in midle, my mystake.
So of course to get Alt Hold it should to be around midle and it is.
I will look one more time in logs.

[bluuu]

hello ... my AQ jump again - 3 times

I isolating log with this accident.

There is ACCZ peak like earlier ... but ... as we see ... after jump values are good.

There is log in attachement.

I will try to remove dumpers and try again.

Additional info ... it became 3 times (1 at flying at Alt/PH, 2 and 3 at RTH). Rest of fly time (about 40 minutes at manual, Alt, PH, Mission) ... no accident.

I add some pictures of hexa later.

[Max]

You have a lot of vibes there. When you plot ACCX/Y, you want the variance to be well within the 0.5 mark, preferably within .3 (of course lower is better). Yours vary in the 0-1 range. My guess is at some critical point the motor/prop vibes set up resonant frequencies in the rest of the body/control tower which overwhelm the sensors. I think this has been explained before.

Have you done the test on individual motors we discussed before? It can be very revealing. I quickly found a bad motor on my quad the other day which was causing the whole craft to vibrate and fly poorly. Often just by turning the prop on the motor a quarter turn you can bring the ACC X/Y variances down significantly.

Another thing to consider with plastic/thin/flexible props is they all have an RPM/thrust point at which they vibrate a lot more. You can see when this happens by watching the prop from the edge. When it hits a certain speed/resistance,the ends of the blades will start lifting up, until they get to their maximum flexing point. During this transition time they can easily vibrate 2-4x more than normal (or worse, with really crappy props). Ideally this happens at a relatively low RPM which normally wouldn't occur in flight. But if for example your multi is overpowered, the props would be more likely to hit this RPM range in flight. Perhaps only under very specific conditions, like when the right combination of pitch/roll/yaw/thrust leaves one or more motors only lightly loaded. Of course this can occur with a "normally" loaded mutli also, it's just less likely to.

Similarly, always check the tracking on your props. They might be perfectly balanced but completely loose tracking under load, which causes horrible vibrations.

Find the source of your vibes and I'm sure your "jumping" issues will go away.

-Max

[bluuu]

Allrigth Max. I think your post will be in wiki. Clearly explaining dependings
Thanks. Hexa go back on the table

[bluuu]

update:

I tested motors and props indywidualy (at 50 in mixing table). ACC and Gyro XY values jumped to 0.5 at peak at pecific (low) RPM. At other range RPM values max 0.35. (on all motors).

In Z positions ... nothing wrong (but copter was on the table and added lot of books weight (for safe)).


I do one more observation. When I toutched FC with shield and GPS motors jumped RPM. It confirmed that gummy dampers are wrong. I waiting for nylon stands. After changing I'll do tests again.

Question for future. What PID (maybe) try to lowering for lowering gain of sensors ?

[chschmid]