IMU Sensor Z Axis hangs ?

Hello,

I've been getting intermittent "Init failures, Retrying" Warning message during power-on cycle of my NEO and VBCt. By disconnecting battery and reconnecting I have been able to get past this warning and finish calibration.

However, this past weekend my system started displaying a new warning message "IMU Sensor Z Axis hangs" in Addition to "Init Failed, Retrying"

Please advise;
what are cause of these error / Warnings?
Is my NEO failing and no longer safe to use? or is there an procedure to correct these errors.

I have attached a few log files for your analysis. Please ignore the battery warning on one of the logs, I was testing and choose the wrong battery profile.

Many thanks,
Kourosh

Hi,

cause of the log file entries are issues with the accelerometers and/or gyros, inside IMU.

Reason for 'Init failed, retrying ...' is either the model not sitting still right after power up, before pitch-twitch.
In this case, the sensors can not settle, software can not learn center positions, or (less likely) sensors no longer OK, unquiet around center or drifting.
Also, if you move the sticks during initialization, you can interfere with the algorithm.

Reason for 'IMU Z-axis hangs' (sticks mechanically) is vibration, now or during the previous flight and powered off in this state. Not necessarily the visible or palpable vibration from the mechanics running, but high frequency/hard/sided vibration caused by teeth of gears meshing, under load. This _can_ cause the vertical sensor to stick mechanically, and release, and stick again ...
Seems you have selected Standard Initialization already (over Precision), which allows the NEO to initialize even if this particular sensor is sticking at power on. Then, the artificial horizon is calculated using the remaining sensors. If Precision, you'd get a generic IMU sensor error message, the servos would twitch 30+ times, and the NEO would flash read, to indicate there's a problem. Plus, a VBar Control would also show and sound an alarm.

Most likely, the NEO is good, and the platform (airframe) is causing this.

Remedy: check for proper gear mesh, for worn bearings etc., and if possible find a different location for the NEO (behind the mast is still the best place), and try a different suspension (thicker tape, softer tape like Kyosho Zeal, or sandwich tape-metal plate-tape), same as you'd have tried with legacy gyros if you experienced tail drift.

What can possibly go wrong: if the vibration levels are high enough to obviously interfere with the sensors, the horizon can tilt/shif/flip in a way so the NEO does no longer know up from down. All well ... until you trigger Rescue.
How can you check? Open the Rescue Status screen on VBar Control, make sure the ball is in the center of the hair cross (or almost) when the model is sitting on the bench, and that the orientation shows up/down properly. Then fly (without using Rescue), come back into a hover every now and then (like, every minute), or land, and check again: if the ball still returns to center, all well. If it would point elsewhere, Rescue would go elsewhere.
Of course you can not check before every Rescue usage, but this way, you can get a feeling for 'everything OK' vs. 'something looming'.

With a third party radio, you'd land, spool down, trigger Rescue. If the swash plate goes up straight, all well. If it pulls anywhere but up, clear and present danger.

Best,

Eddi

Born to fly ...
forced to work.



Edited 1 time(s). Last edit at 07/26/2022 09:24AM by Eddi E. aus G..

Hi Eddi,

Many thanks for your quick reply. Interesting information, but I'm a bit puzzled. Please see my notes below.

The machine is a new build Diabolo 600. I don't think the bearings are damaged, yet The machine has maybe 15 flights. The pinion and gear mesh are per instruction (very slight click at hight point of 1st stage gear) additionally, I apply dry fluid grease on gears and pinion. What else can I do to have a smooth machine / no vibration.

Power-on process:
At the field I power on the system on a solid table, no shaking of platform or moving of the TX sticks while powering on. NEO powers on with blue light and then green, Normal pitch twich and tones.

Update and Notes:
Your note about a IMU sensor being mechanically stuck made me think to try something simple.
Just now, I powered on the system on the bench observing the log. I see the IMU sensor Z Axis hang message. I waited, After observing "calibration finished" and "YGE ESC detected" message, I just lightly tapped the case of NEO with my fingers. Noticed "IMU sensor Z Axis released". I power down the system, waited a few seconds and powered back on. No more IMU Hang warning! *Have we fixed the stuck sensor by Tapping the NEO case while under power? See attached pictures of log.

I don't have the rescue function enabled for this NEO unit, so I can't check that. I don't know how to enable precise ver. standard calibration?

The NEO is installed behind the main shaft on gyro platform of the model. I am using one layer of gyro mounting tape that came with the NEO. Should I add more Tape?

Per attachments in my original message, the Logs show "Raised/High viberation Levels" when switching banks and "Extreme Vibration Levels" when I have landed and turning engine and ESC off. Is this level of vibration warnings normal, or should there be no vibration level warnings in the logs?

This NEO is practically new. only flown on this machine. Is this NEO unit to be trusted? Should I replace it, or send for service?
Should I replace standard plastic casing with metal casing?

Once more, many thanks for all your help.

Kourosh

Hi again,

some models are more likely to produce adverse vibration than others. If there is 'teeth chatter' not only to be felt radial (around the gears) but also vertical (force and opposing force, the first gear not only pushes the next gear one tooth further, but it also causes an axial force, pushing itself and the next gear against it's respective opposing bearing/bearing seat). If that doesn't run super-smooth, every infeed pinion tooth -> main gear tooth also causes a 'tic' on the vertical axis. That's normal, that's not particularly bad, it's just that the accelerometers can get tricked by it, same as if hard gears mesh in a way so each infeed is also a palpable 'tic', particularly under load at high rpm.

Also, if the mounting plate for the flybarless is somewhat loose, it can interfere.
We had that on LOGO XXtreme, where it was recommended to glue that mount to the frames.

I can not tell for sure, though.

Power on seems to be OK. Still, from time to time, a retry can happen. Nothing to write home about.

IMU, indeed, if you tap the unit sufficiently, you can cause the sensor element to release—zing. Then, you get no error message at the next power on.
But during the next flight, it can stick and release again eventually, it's in it's nature, cause (still there) and effect.

Your NEOs only seem to be Pro (which is OK), so the simplest remedy would be, disable Lightness in Flight Parameters / Main Rotor.
This is a feature which is also using the accelerometers, and if disabled (in all 3/4 banks), the sensors will not be queried, so, no error message.

You can try a different/softer or sandwichy tape, though, it can help making a difference.

Replacing the case with the metal one might or might not make a difference.
If so, then in a way where you de-tune the vibration sufficiently so it does no longer affect the sensor(s). Same goes for a different tape.
Some pilots reported an improvement when switching to plastic, some when switching to metal, some not at all.

I doubt the unit needs servicing ... again, some models are more agreeable working places for accelerometers, and some less.

—Eddi

Born to fly ...
forced to work.

Many thanks Eddi. I will work to see if I can figure a way to tweak the gears for better mesh / improved smoothness.

Is it normal to expect Raised / High / Extreme vibrations in the logs? Assuming all is setup as good as possible.

My Diabolo 700 (10 year old) has the same two stage gear arrangement, I have not seen IMU warnings from the NEO installed on it. I do see Raised/High/Extreme vibrations reported in logs during every flight. Maybe that machine has different harmonics, or maybe the NEO unit on that machine is a bit more forgiving, than the one installed on this Diabolo 600!!

Kourosh

Hi Kourosh,

vibration levels in the log occur when you do 3D, almost starting at the time when you hear the blades 'fart'.
Nothing to worry about.

In a hover, there should be no such entries.
Check vibration analysis (the graph) for details on where they occur, and how bad they are (on the bench, in a hover).

I can only guess that the 700, because of it's mass, maybe really has different harmonics—nicely put —than the 600.
You could try and swap NEOs between the two, to compare.

Of course there are production tolerances, for once with each model, for once with each piece of electronics.
If, for example, the main board has play inside the casing of the NEO, it is more susceptible to vibration or resonances.
Differently put, because if this, they can indeed be more or less forgiving, especially if operated at the edge of their capabilities.
This is probably also a reason why exchanging a case sometimes does something, sometimes not.

It's all not an exact science, I'm afraid.

Kind regards

Eddi

Born to fly ...
forced to work.