Can't do full forward flip during inverted forward F3C flight

Hello Sir,
this time I have an issue with the Urukay WC, that is new to me:
When flying the F3C, P6 figure : "Oval with travelling flip". So the heli is flying (fast) inverted forward, after the heli must make a forward flip (nose going up). At this point the heli seems "to get stuck" after 90degrees, so will not finalize the flip. It seems as if the swashplate cannot move enough to finalize the flip ?!

This happened today, after I changed all swash servo's from JR DS8915 to MKS HBL950.
The lenght of the servoarm of both servotypes is the same (18mm) and according to the manual of the Urukay.

When checking vbar log I see "cyclic ring active" (See attached file of bank1=aero with 1750rpm. Bank0 = hoover and bank2 is aero with 1900rpm.
I've also noticed in the software that the rates are quite high, in order to achieve the applicable "collective travel" / "cyclic throw" => do I have to go for the outer hole in the servoarms .. ??

Cyclic ring is the standard value of 80 and nothing binds and could be made higher theoretically => what's the criteria for this setting ?

I have the feeling it must be some kind of setting since in the mechanics nothing has changed.

Hope you can help me out.
thanks!
Sebastien
The Netherlands

Hi,

I guess you are exceeding either what the servos can do, or what the model can do.

Do I get it right, it did work with the JR Servos? Do you happen to have the setup(s) saved from when you had those on the model?

Things that meet the eye pretty obviously:

Trim on CH1 (elevator servo) is huge, creating an offset from the servo center position and already using a considerable amount of the overall servo pulse range.
Settings for collective and cyclic _way_ above the recommended range also cause a considerably high servo pulse output.

This together causes the VBar to run into it's own electronic limiting (cyclic ring), plus it could explain that the servos (CH1 servo, "elevator" servo, but also CH2 "left aileron" servo) and subsequently the swash plate can not do more travel, causing the model to seem to stop mid-air with full collective and full cyclic forward.

Solution: change the mechanical geometry so you get into a sensible range of values for collective and cyclic, and reset the CH1 servo arm in a way so you can trim considerably closer to zero, for zero collective. Do the same for CH2, although there it's not as critical since the cyclic ratio in the mixing is not as high as it is on CH1 elevator.

Back to bench ;-)

Kind regards

Eddi

Born to fly ...
forced to work.

Dear Eddi,
thanks for your very quick reply !
It did work flawless with the JR servo's . Unfortunately I didn't save the vbar file :-(

I will change CH1 and 2 with respect to servohorn position, maybe rotate it "one teeth" further.
Also I will shift the ball links one hole further outside to get lower values for collective and cyclic.
Still strange that with the JR's and the same length servohorns it was obviously good...

To be continued...
thanks again
sebastien

Hello Sebastien,

two servos must not have the same travel, could make a difference.

I'm curious.

Kind regards

Eddi

Born to fly ...
forced to work.

Hi Eddi,
well after several hours I got the new better set-up, for collective and cyclic. Moved the ball links on the 3 servoarms one hole further outside.
However, the trim for servo center I can't get any better for those two servo's 1 and 2. Rotated the arms as good as possible and swapped the servoarms 1 and 2, but no improvement. It's impossible to get smaller trim values. (vbar Files attached)
I get now 12degrees of pitch and 8 degrees collective (that was last time less for both, forgot to mention...).

Maybe I need another set of servoarms and hoping that the splines are a little different (although when it's manufactured the splines are on the same spot ...).
Unfortunately I've had those special servoarms with which it can be rotated "stepless" but they do not fit in the helicopter (arms are very close to the centerpiece...)

Dispite this, I think I will have to give it a try (after my ski holidays ...).

Whenever any update I'll let it know here, so others can learn from it as well :-)

regards
Sebastien

Hi,

OK, looks different now :-) but the trims. Unfortunately again proof that design and function are often not what they should be :-/
On the other hand, if you re-set the servo arm one step, I can not imagine that you get a trim value of the same amount to the other direction, I would assume less, but with conversion. Or do the splines have such coarse teeth?

If you do not get binding now (links, servo arms, swash plate), increase the cyclic limiter ring back to 85 (default), this will also help with a little more deflection of the swash plate.

You could also think about putting plates between the servo mounting ears and the servo mounts, so you get more room to the center piece, for the other servo arm design.

Happy skiing! I had at least one day on my board this year, must be enough

Eddi

Born to fly ...
forced to work.

Hello Eddi,
well unfortunately I'm back again since the issue is not solved yet...

Today I did some testflights with the latest mentioned settings : cyclic ring = 85 without binding.
Then there was very little improvement flying "the" maneuvre, while flying , the heli could make an angle of max 100 degrees (from the start).
Then I thought, ok, let me increase the cyclic limiter setting without binding and came up to 100% (bank aero 1 and 2). The angle I could reach was about 110 degrees and still not enough to make the full forward flip while flying inverted forward.
Then the same was flown with the highest headspeed, but not any difference (.

In the picture attached you can see how much the swashplate can move when giving full forward nick and some negative pitch on the bench when the vbar menu is active (about the same when I'm flying "the" maneuvre) and more seems not possible without binding.

Another check I did was flying some forward- and backward static flips but that didn't give any issue. They are quick and clean.

Then I checked to see that the servo's got enough power, by looking at the maximum current of the BEC, which appeared to be 4.4A only (see picture) and should be no issue for the built-in 8V , 15A BEC (+ 2x extra low esr elco's!) of the jeti mezon 130A controller.

It seems that somehow the servo's do not rotate enough or that the PWM signal that they get is not reaching it's full range...
Could it be something to do with the frequency that the MKS HBL950 servo's can operate on, as it is 333Hz maximum ???

So, now I'm a bit puzzled, something like this never happened to any other of my vstabi heli's before, so what to test next.
I've attached the settings for all banks again for a check.

Maybe you have any idea's or suggestions...

thanks again
regards
Sebastien

Hallo Sebastien,

the Vstabi-FBL "pushes" the swashplate-servos only with 143 Hz. So this should be no problem for your MKS-Servos.

Gruß, Wolfram

Hello,
after testing on the bench something unbelievable happens; when giving full pitch AND full forward elevator, the servo seems to come in some kind of "undefined" state and can be moved by hand !
In this way, when flying, the pitch will decrease when flying the mentioned figure and the flip cannot be finished, so this explains why the figure could not be flown correctly.

In any other way the servo moves correctly and also with a seperate servotester everything works ok. I've never seen this strange behaviour before. Could this be caused by the v-bar eg. some kind of cyclic ring "overdrive"?
I will do some more testing coming days.

A video can be found here : https://youtu.be/p8osN8EkbpY


regards
Sebastien

I know this effect. Let me guess … a lot of trims in the servo and high cyclic and collective?

Have a look in the Trim tab and read the text on the left side …

You have:

- Trim over 40 … rotate 1-2 tooth!
- cyclic and collective over 100 - use longer horn


Reason: Some Servo manufacturer let the servos be forceless if they get a PWM pulsewidth out of their internally defined range.

regards (EN) / Gruss (DE)

Rainer Vetter

_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
Posting and answering - please use a PC and a standard keyboard. Answer all questions fully, take the needed time, don't forget probably requested attachments.

Für Posts und Antworten bitte einen PC und eine richtige Taststaur benutzen sowie alle Fragen vollständig abarbeiten. Nehmt Euch Zeit, beantwortet Rückfragen in Ruhe und vollständig.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
Für Einzelkommunikation bitte ausschließlich Email verwenden! - PN, Messenger, Chat usw. werden nicht bearbeitet.
For single user communication please send a email only! PM, Messenger, Chat etc. won't be read.

RV Wrote:
-------------------------------------------------------
> Reason: Some Servo manufacturer let the servos be forceless if they get a PWM pulsewidth out of their internally defined range.


Aha, yes that's indeed a very plausible explanation Rainer!

I just did a test that confirms this, by swapping servo 1 (=nick) to one of the roll servo's and indeed the behaviour moves along with that servo. Luckily it's not a servo-issue.

However, I'm currently at the maximum outer hole of the original SAB servo-arms, a longer one will not fit due to limited space of the nick servo (1). Next to this, the distance between center and the ball is now 21mm already. This was 17.5mm with the JR servo and its original servo-arm ... we definetely can conclude that these MKS servo's WILL have less movement for the same pulswidth ! strange ...

The issue with the trim is also a tough one because with these servo-arms it's just not possible to get is better, also not shifting only one spline or changing to either of the other 3 servo-arms ....
Maybe I have to order another brand of servo-arms, hoping to get the trim right....mmm... any ideas are welcome !

to be continued...
regards
Sebastien

I could solve the issue!

I removed all servoarms again and found another to be better suitable for the nickservo (1), it still needs significant trim, but instead of +52 I got to -39 now. Also reduced cyclic ring from 100 to 90. Extensive benchcheck (all maximum values, roll, nick ,pitch ) appeared to be ok, the pulselength seems not to be on the edge of what the MKS wants to have as valid value anymore. It keeps on working properly.

Then I did several testflights and no issue anymore for flying the figure.

Thanks for helping out Rainer.

However it's still a mystery to me why the MKS HBL950 servo's do have significant and noticable less rotation for a given stickpinput, I've had many different servobrands for the last 10 years and this is the first time that I see something like this ...

Anyway, case closed !
regards
Sebastien
ps. keep up the good support on this forum it's outstanding !

Well done!

regards (EN) / Gruss (DE)

Rainer Vetter

_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
Posting and answering - please use a PC and a standard keyboard. Answer all questions fully, take the needed time, don't forget probably requested attachments.

Für Posts und Antworten bitte einen PC und eine richtige Taststaur benutzen sowie alle Fragen vollständig abarbeiten. Nehmt Euch Zeit, beantwortet Rückfragen in Ruhe und vollständig.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
Für Einzelkommunikation bitte ausschließlich Email verwenden! - PN, Messenger, Chat usw. werden nicht bearbeitet.
For single user communication please send a email only! PM, Messenger, Chat etc. won't be read.

Hello Sebastien,

if you want the trim values to be zero, you can give the KST-servo-arms a try.

stepless servo arms

These arms have four M-2-threads in a radius form 11 to 20 mm´s.

You can complete them with the following Mikado parts:

-) 01570
-) 01939
-) 04804
-) 02070

Or you drill one of the holes, in your case the 20-mm-radius, to 2,5 mm and cut a M-3-thread. In this size you can find ball-heads that will fit.

Gruß, Wolfram


PS: here you can find the dimensions: KST-arms



Edited 2 time(s). Last edit at 03/29/2019 01:04AM by Wolfram.