Severe vibration issues

[JussiH]

GPS quality has a lot of influence on position and altitude hold stability. So Id need to see a log from the same flight to get an idea if that could be part of it.

Also, accellerometer calibration has influence. Is this a DIMU or a AIMU? Did you fully calibrate or just tare?

Have you covered the barometer with foam and shielded it from sunlight?

[Astudillo]

Hi Jussi,

Thanks for the info, unfortunately i don't have the log file of that flight anymore. But i did a new flight and a new log.



I'm using the DIMU, and did a full calibration. I put some foam form a peli case between the AQ board and the DIMU.
But its not completely shielded from the sun as you can see in the photo. Is that i problem?


Cheers,

Ray

[hogster]

Hi all

Just to update this thread, I've finally had some time to give to my X8, but I'm still going nowhere fast

I've redone my calibration using this guide:

http://autoquad.org/wiki/wiki/autoquad- ... de/3352-2/

The HUD is nice and level when on a level surface, ACCX and ACCY sit at 0, and ACCZ around -9.8. My magnetic vector magnitude is around 1.93, although on moving the vehicle to another position it went up to 2.11 ...

However if you look at this log file of a recent hop (the flight is the last thing in the log file - the rest is moving it around on the ground):

http://www.horizonimaging.co.uk/rcgroup ... es/7-6-14/

and in the attachments, you'll see that:

• I've still got quite bad vibration (ACCZ for example goes to +- 4 ... ACCX and ACCY seem to go from -3 to +9 though! ). This is better than before as my props are all balanced now (almost all of them were OUT of balance before!)

• Why does the ACCX readout suddenly click to an average of around 0.92 at the 30 second mark (assuming those are seconds on the X axis?)

• The motor outputs are still horribly noisy (which I think is what caused one of the motors to 'cough' mid hover causing that side of the X8 to drop)

• The MAG values seem to be strongly affected by the running motors

I've run the motors up to almost take-off throttle with the X8 sitting on a bench, and the ACC values are much smaller, and the motor demands don't sound nearly so noisy / scratchy – it's only once I'm airborne that it sounds so nasty

Presumably I'm getting EMI on the magnetometer from the ESCs, all 8 of which are about 5cm from the AQ at their nearest corner. Should I try raising the AQ up further? It's currently on 15mm standoffs from the top chassis plate, directly underneath which are the ESCs. This is an old photo showing the AQ in the box I made for it:



It's not in its box at the moment, which has actually put it about 10mm closer to the chassis plate and the ESCs ... the ESCs live underneath the oval holes you can see in that photo.

I would say it's flying the smoothest it's flown so far, but it's still not at all useable, particularly with the nasty scratchy sounds coming from the motors – I'm sure that's not good for them, or the ESCs for that matter.

Any advice on how to proceed is much appreciated, I'm starting to get tired of this thing sitting on the bench all the time

Many thanks

David

[LPR]

David

Don't feel like you're all alone having vibration problems. My overlapping prop octo is having vibration problems also. To get it flying, but not well, I set the Tilt Angle P down to 150, default is 290. With default PID setting it wobbles so bad I will not fly it.
I removed four of the motors and arms and fly it as a quad now but I still have much the same result.

https://www.youtube.com/watch?v=OeF3jmrqiAk

The same AQ board flies very well in a small quad with default settings. Here's an out of focus video of the small quad flying.

https://www.youtube.com/watch?v=gyiYNa8IUpE

I'm thinking the octo's 4014 T-motors with 17" T-motors are creating just the wrong frequency of vibration.
I'm changing the frame now and hoping to get better results.

Larry

[bluuu]

Some of vibrating problem:

I alwayce make it in 3 steps:

1. Balancing motors - running motors separately and reading acc values, after that I take small piece of tape and put it on side of motor. Reading values, if higher I replace tape and put it a little further and next reading values.
I do all 360 deg like this.

After that I put this small piece of tape in point of lowest vibration.

After that, repeting operation while vibrating lowering.

You must remember, point lowering viprations can be in different places but in most cases this is one point.

2. Balancing proppelers
After balancing motors it's time to proppelers.
Normal static balancing on balancing tool.

After that we could check result one more time.

3. Balancing motor and prop
Run motor with prop and read acc values. Mark current position of prop relative to motor.
Then rotate prop relative to motor by ... 10 deg, read acc. Find the best position.
If proppelers are screwedin 2 points, use tape

All proccess takes much time but result is worth the work.

(can anybody correct my english and put it to wiki ?)

[plexus]

how you see the ACC of one motor ? Do you disconnect the power supply of the others motors ?

or have you an other solution ?

[bluuu]

you can set it in mixing table.
I'm seting 1 motor, throttle 100 other 0.
On CAN I only changing CAN number on PWM changing port number.
Then you can start telemetry and see acc values.

[plexus]

Ah ah ok
Very good .

Many thanks bluuu.

Alain

[hogster]

Many thanks for the replies everyone

bluuu - I will certainly try going through those steps again. The motors all seem extremely well balanced, and I believe I once ran them all at the same time and felt no discernible vibrations. The props have all been balanced on a DuBro Tru Spin prop balancer. Are you saying it might be worth adding *extra* tape until it feels like the vibration is at its minimum (when spinning up each motor with the prop attached), even if it means putting the props slightly out of balance again?

Do you think the noisy motor outputs are a direct result of the noisy accelerometer / magnetometer readings? I'm assuming so, as I wouldn't have thought PID values could have such a strong effect on the hasrhness of the motor outputs. I will certainly try turning down the Tilt Angle P value on your suggestion Larry.

By the way, is there any way of providing additional electromagnetic shielding from the ESCs, apart from moving the AQ further away? How about sticking grounded copper foil to the underside of the upper chassis plate around the ESCs? I appreciate that you don't want to affect the Earth's magnetic field though ...

Many thanks for your help guys

David

[bluuu]

about vibrations - you must see curves, sometimes this could be resonance freq
If you have all done, PIDs settings for hexa will be quite different. Default are for quad.

For first look at Tilt angle D, default value is high, lower it to .... 30000 and check logs