Good morning - another day, another attempt at getting rid of the wobble on the X5.

This morning I took apart the main gear and autorotation assembly. I replaced the main gear and checked the assembly was solid, tightening the screws carefully to ensure pressure was equal around the entire gear.

I flashed the firmware to the latest beta version which includes better low freq. vibe updates (does the update contain better vibe analysis or better vibe compensation?).

Here's the results:

1. Vibe analysis of the X Y and Z axis (using the Android app + BT module) is clean within no RPM range showing more than approx. 20-24 range. I 20-24 at 9450 rpm - that's with a head speed of 2300. That's a great improvement on my previous vibe analysis where (from memory) I had mid 60's on the y-axis. Both times I removed main and tail blades.

2. At a head speed of 1860 I can run 55-60 head gain with no wobble during aileron inputs.

3. If I move to a head speed of 2200 - 2300, the wobbles are very bad.

4. Head speed is measured using a small hand held meter - not an iPhone etc.

I haven't tried different blades yet, that'll be my next move (need to order some).

Any other ideas I can try out here?

Thank you.

_________________
--
John Clayton

Hello,

that is very strange.
So the special firmware does not helped at all?

You can also enable Extended signal processing which may help.

In the firmware nothing changed except new special bandpass filter. So vibration analysis should give you same results.

_________________
Spirit System developer

Hi

It is possible that the special firmware helps as I can at least fly at 50 head gain now - although unless I put the old main gear back in I won't know if that's the s/ware or the gear.

What I'm finding is that there's a head speed zone at which the heli shakes. It's OK at 70% throttle, it's bad at 77-80% and it starts to calm down a LOT at 85%.

I should also clarify, the heli is stable(ish) in hover. If i go slowly vertically down slight shakes occur. If I hover and the give a reasonable aileron input, e.g. so the heli tilts quickly to 45 degrees to the right, the shakes start. I can't recall if the shakes start BEFORE I can bring the heli back to level, or stop it by using left aileron to counter the initial right aileron. If you think that's important I can try to make a close up video.

At 70% and 85% I can get up to about 50 on the head gain, and I can see the tiniest of wobbles in the heli.

I tried some Revolution blades that I had sitting around - the wobble was actually worse. I re-checked the balance on the older (Gaui X5) blades, and put them back in - visibly less wobble than the Revo's, but it's still there.

Extended signal processing doesn't appear to make much of a difference.

I'd also be happy to record video of the vibe analysis across the throttle band if that would help... let me know if you want that.

I've got some rail blades on order, and I'll try to get the bearings and thrust bearings etc replaced, in case they are binding.

I'm really keen on trying anything else. Thanks for your help so far.

--
John

_________________
--
John Clayton

Thank you for the report.

It is hard to tell you what can be a reason for this problem. But in case that updated firmware helped then it is most likely caused by low frequency vibrations that are visible only with blades. Frequency of this is most likely same with main rotor or even lower.
Spirit is able to react extremely fast (at 1000Hz rate) so with fast servos + such noise it could be unfortunately visible, because such noise is passing to low band used for flying. To verify that it could be caused by this, you can simply try to set cyclic servo frequency to 50Hz. Just for hovering test.
This will cause much slower servo reactions. But definitively it is not solution.

I also know cases that similar shaking was caused by bad one way bearing and also too tight main blades.

Please make sure, that nothing touching the unit (free connector or cable).

Vibration analysis with blades on (especially analysis for aileron axis) may tell us a lot, but it is not safe, so I can't recommend it.

_________________
Spirit System developer

I can try that - are there any other settings to slow the servo's down? Or is there a cut-off at 50 hz where the s/ware reacts with a different algorithm?

_________________
--
John Clayton

Algorithm is working always in the same way.
I am afraid, if it is caused by vibrations, then it must be few times higher than normal. Unfortunately problem occuring only in specified RPM range and with load (rotor blades) so it is hard to check where is exactly the problem without blades.

I recommend to loose bolts little bit in the blade holders if they are so tight.

_________________
Spirit System developer

I attached the X5 to a wooden board and weighted the board heavily. This enabled me to run the Spirit vibe analysis software with and without blades in the following three ways:

1. No blades
2. Only tail blades
3. Tail and Main blades

I ran each of the 3 test scenarios above at 70%, 75% and 80% throttle. I have zipped the image files taken from my Android device during vibe analysis, I've also pasted them into an Evernote doc - so you can choose your poison so to speak

The directory structure is simple, e.g. for the x/y/z axis images for 70% throttle:
no-blades/70/<datetime>.png
or for 85% throttle:
no-blades/85/<datetime>.png

there's three pictures per directory, always X, Y and Z axis snapshots - the axes markings can be seen on the bottom left of each of the picture.



Seems like with blades, the vibes are huge - but I don't know if this is normal or not. Any help is appreciated trying to decode the story that this tells.

Evernote link is:
https://www.evernote.com/l/AAdi1wcb0OdO ... KEjnecBLfQ

Thanks.

_________________
--
John Clayton

Thank you very much for posting your graphs.

Unfortunately graphs with blades attached are showing more than extreme vibrations. Y axis is the most problematic, my estimate is that spikes are approx. 400% high. Normal vibrations are 50% (half of the graph), 70% is acceptable, 100% and more are extreme.
It is most probably caused by disbalance of the rotor blades, which may be caused by something wrong in the blade holders. It could be enough if there are different washers (or missing) or axial/radial bearings are not fitting exactly then weight of one blade is moving whole shaft out of symmetry.

Without blades it is perfect, so the problem can be observed only in a load (with blades).

_________________
Spirit System developer

In my continued search to isolate the problem, I rechecked the blades - they are perfectly balanced.

I then took a closer look at the head. Hmmm... I see that I have about 0.5 - 1.0mm lateral movement of the feathering shaft - this seems like a bad thing me, but is it? Should I expect 0mm?

Here's a link to a vid showing the movement:
https://gofile.me/2fjCa/d3OXy0wq

Thanks.

PS: I also posted this on Heli Freak asking others if they have a similar play.

_________________
--
John Clayton

I now have reasonably clean X/Y/Z vibe analysis with main blades on.

I have not yet tested the X5 in flight, but will do so in the coming week.

To get this, I simply tightened the feathering shaft as much as I could, basically to the point where I thought the strength being used would soon strip the bolts, or tools. This reduced the play significantly.

I attached appropriate commented images showing the difference.

I'll keep you posted as to progress.

_________________
--
John Clayton