[Filippo94]

Today I tried to spool up the heli for the first time, but had a bad result .

Slowly rising the throttle, at middle stick some "jitter" starts, and raising the throttle to max lowers its frequency. Before doing that I searched around for vibrations, but everything seems ok. Doing further testing revealed that the servos are moving strangely, without any input. To me it seems like a high frequency type of vibration causes, with an offset to the CC3D update frequency, a rotating correction.

I then paid attention to it with no blades on (tail and head), and the problem is still there. With the heli secured (AND NO BLADES!!), I kept a finger over the CC3D while testing various throttle inputs, all I can feel is a Gamepad type of rumble, there are no oscillations on the tail, which looks rock solid, before reaching about half stick, where the tail servo starts oscillating followed the same way the ciclic servos do.

It's the first "full size" RC heli I'm building, I have no idea of what's usual and what's not, but in the 120 SR, after balancing the blades, I had a similar rumble (quite lower intensity, but that heli is about 110gr heavy) and never caused issues to the gyro.

Rate or attitude seems making no difference. I have stock stabilization values, still have to take off and see what are the needs.

So I wonder, is the CC3D really picking those vibrations and reacting, or is my setup not good enough to allow any gyro activity? Would lowering the compensation gain solve the issue?

btw, CC3D is mounted behind the main rotor, right over the tail mount.

[daveapplemotors]

Jitters like if the throttle midpoint is too far fro the minimum? If my throttle minimum is 871 after I do Manual Calibration then I set my throttle neutral to be 873.

hope it helps.

[Filippo94]

Nope, throttle response is all right, it's the way the servos act which looks like being interferred. They oscillate a bit, or a lot, depending on the throttle amount. For example, at mid stick it starts oscillating slowly, like when your tail gain is just a tad too high, going higher it gets more intense and frenetic, reaching 75% it is quite random, going further amplifies and goes slower, at full throttle looks like you're slowly turning your cyclic stick (mode 2) in circles.

I didn't check if stability was still working (if you don't mind, I'll leave a full throttle heli well anchored on the bench), but the stick response, in all the four servos, were crisp and working.

indeed, it looks like it suffers from resonance...

[daveapplemotors]

I don't understand what it's doing.
video?
UAV>

[Filippo94]

Will post both tomorrow, hopefully will be visible enough.
Does setting "Manual" in the flight mode assign the servos raw input from the transmitter (filtered through the 120° servo math)?
Maybe if I manage to keep the stabilization out of the system, I can find out if it's, indeed, a stabilization issue, or if it's a mechanical issue, even if everything seems solid and in place...

[Filippo94]

So, here's the video, and the UAV (attached). The servo motion is not that visible on the video unfortunately, but it's surely enough to cause the entire heli to oscillate while it has blades on, causing further vibrations. You can see even the tail pushrod oscillating back and foward slightly, moved by the tail servo.

Also you can see how the tail has no motions while the engine spools up (tail servo apart), showing that there are no visible vibrations. I did look even without the heli locked, and the results doesn't change.

Edit: I did tested manual stabilization for pitch, roll and yaw, after recording the video. The response is correctly filtered, but indeed has no reactions to any movement. Spooling up the heli this way, the servos do not move at all. So yeah, it's a vibrations issue. How to fix it, since the heli looks quite balanced already?

[f5soh]

Your AccelTau value is set to 0

This means the accel filtering is completely disabled: Try to put something around 0.05,  Attitude tab > Accel filtering

Edit: You can also increase the refresh rates for your servos... maybe capable more than 50Hz ?

[Filippo94]

Turned the Accelerometers filter up to 0.2, it didn't affect the problem. Remember it happens both in attitude and rate mode.

How do I know what refresh rate my servos are able to receive? Trial and error?
I've read somewhere that digital servos should be able to handle 300hz, I did try that for the tail servo, but it stopped moving at all. Maybe I can just increase over 50hz and see what happens.

[daveapplemotors]

G'Day Fillippo,

Change the 1099 (throttle neutral) to 1067, Throttle minimum is 1065 now. Neutral and minimum must be close.
 
Do you really need your servos to react at faster than 50hz? Mine all work fine at 50hz. I have burnt up an ESC --trial and error. I have ruined a few servos, too. One digital servo, E-Flite DS76t, gets hot at 50hz and it's supposed to be a 330hz kind of servo. Go figure!

Be careful with servo refresh rates. If anything gets hot or buzzes then lower the rate.
Good luck,
Dave

[Filippo94]

Changed neutral throttle in inputs settings to 1067, now the engine spins always, when I'm not in hold mode, even at min value. The oscillations are still there with both Attitude and Rate.

I'm not sure if I need more than 50hz refresh rate, to me it seems enough at least to start flying the heli. Surely if I'll need some more snappy behaviour I could try go higher, but I'm sure there's plenty of settings to play with before I'll need more accurate responses than 20ms.

I just had an idea, to check the graphs from the CC3D.
It seems the vibrations it's picking are really really wide, even if I can't see them. How could it be? (remember this was with the USB attached to the board)

I started the heli in manual, so the CC3D was not moving the servos, and slowly raised the throttle up to 50%, then activated the attitude mode, exactly when you see in the attitude graph starting oscillations. Then slowed down to stop, and spooled up again, always with attitude, you can see the tail was oscillating quite a bit.

Same settings as before (except manual instead of rate for pos. 2 flight mode).

[daveapplemotors]

Re-do Manual Calibration of TX in Configuration. After that change the neutral to be close to the minimum. I have a feeling the glitch/jitter is caused by that setting. I set my throttle neutral really close to the min. 

Board level it important to Attitude and  Rattitude. And the swash seems to have different ideas if you switch to those modes without board level being perfect.

[Filippo94]

I did like you said, now these are my new TX values (attached). It didn't solve. I discovered my throttle curve for Hold was at -10% (maybe some nitro heli precaution the Dx6I adopted?), set it back to 0%, and now the min throttle is correct even with hold off. Yet there are those oscillations.

I've checked the CC3D position, only think I could find out is that the USB port could touch the support, so I moved the CC3D a bit back, but it looks like did no differences...

Switching between rate and attitude makes no difference on the swash with a level heli, the previous graph was switching from "manual" to Attitude, which I guess could justify eventual drifts building up while Attitude was disabled.

Edit: forgot the attachment 

[daveapplemotors]

Did you recalibrate the TX with Manual Calibration after you discovered that hold was at -10%?  If not then try that.

[Filippo94]

I discovered I had hold at -10% because after calibrating the min value without the hold was really far from the min result. After fixing the config, I've of course redone the calibration. The result is what you see in that image.

[daveapplemotors]

Rats!
Back to vibration: have you tried the motor without anything else , even mainshaft spinning?