Spirit System
https://www.spirit-system.com/phpBB3/

Extended signal processing
https://www.spirit-system.com/phpBB3/viewtopic.php?f=20&t=537		 Page 1 of 1
Author:  Gtamas [ Tue 17. Mar 2015 22:26:09 ]
Post subject:  Extended signal processing
Hi!
Im using spirit for a few weeks now, and its working really good :)

But having some trouble with stabilization, only when im doing tic-tocs in the actual flight.
If i do a few elevator tictoc, then the stabilization starts to drift. Without tic-tocs, it is working good.
The heli is a 700E DFC, brand new, 1950 headspeed. motor is 750mx, with ice2hv120, and RJX fs 0521 servos at 333Hz, with lifepo4 receiver battery, and 2 spectrum satellites.

I didnt enable Extended processing yet, because first i'd like to know what it is doing exactly technically. Is it doing things with stabilization? or it is change the behaviour in normal flying too? is it some type of filtering?
please explain it if its not a secret :)

another question: 700 DFC has a high tail ratio, so the tail shoud work fine in low head speeds. But i have some tail kick to the right in high collective. I use a RJX fs0521THV servo with 0,03s time. delay at 0. dynamic at 4. piro consistency on 185, rudder gain 1,22, gyro at 80% (no wag at all)

You can see the tail kick in this vid at 5:40-5:42 time. tail is out about 5 degrees.
https://www.youtube.com/watch?v=Me9iJplDgKM

Sorry for the long post :) And to tell the truth, this is the best FBL i tried, and im very happy with it! ( i tried 3GX, beastx, and ikon - ikon was the best after the spirit ) And extra thanks for supporting OS X!
Author:  ZeXx86 [ Tue 17. Mar 2015 22:42:36 ]
Post subject:  Re: Extended signal processing
Hi!

from your video it seems to me, that Piro Optimization is in opposite direction then it should. Stationary pirouettes should be on the spot without any drift.

This should help with flight behavior and also stabilisation precision a lot.

What is your end point limits on the rudder? If they are too high, then lowering ball linkage to the closer hole should do the trick.
Then you have to just update limits and increase gyro gain little bit and it should be fine.
Also how you configured Rudder Neutral position? You can check this when Subtrim (tuning) is enabled.

Extended signal processing is just different filtering algorithm, but in most cases it is not needed. It can also improve precision globally, but if your helicopter is in good condition or excess vibrations can be fixed, then it is not recommended.

Many thanks for nice comments!
Author:  Gtamas [ Tue 17. Mar 2015 23:24:55 ]
Post subject:  Re: Extended signal processing
Pirouette optimization is correct, i check it every single time i change a something, because i know it is very important.

The stationary piruette drift you see in the vid is fixed, the swash wasnt 100% level at the time, now that is perfect. the vid is 2 weeks old, i havent got a new one yet. but the tail kick and the stabilization issue after tic tocs is still there. (without tictocs it is perfect)

Rudder endpoints are around 130. At 90 degrees at servo it is 0 degrees at the tail.
What is better for the tail algorithm? if the endpoint are low ( servo ball farther) or if they are high (servo ball closer) ?

Thank you for your time!
Author:  ZeXx86 [ Tue 17. Mar 2015 23:29:27 ]
Post subject:  Re: Extended signal processing
Ah, good.

You may try to set 3-4° instead of 0° for your helicopter. Some mechanics works better with this little blades pitch, but don't forget to update your rudder limits.
This should solve the issue, your current limits are fine.

It is good if limits are approximately in the middle.

If you are not sure about vibration level, you can measure it without blades at least.
Author:  Gtamas [ Tue 17. Mar 2015 23:52:28 ]
Post subject:  Re: Extended signal processing
Thanks for the infos! i try the 3-4deg pitch next time i can fly, and make a vibration analisys, and report back!
(which will be after the next weekend unfortunately... (this weekend i go for a quadcopter training :( )

Some little things for the "future v2" hardware it would be nice to see: i dont mind if the unit will be bigger if it would have a second satellite port, and i think it would be great if the will be a direct power port with some rigid connectors, not to power with servo ports, and Y servo wires, etc. I hope you dont think that i want to be the smart guy here, i just want to help you with some customer advices if i can :)

Greetings from Hungary :)
Author:  ZeXx86 [ Tue 17. Mar 2015 23:54:55 ]
Post subject:  Re: Extended signal processing
You are welcome and thanks for the suggestion!
Author:  Gtamas [ Sun 05. Apr 2015 14:56:58 ]
Post subject:  Re: Extended signal processing
Hi!

All problems are solved now!

The bailout instability was caused by the Trex 700 tail hum (a harmonic resonance in the tail on certain headspeeds).

Tail kick: as you suggested, i tried the 3-4 pos deegrees preset on the tail (when the servo horn is 90.) It' working much better, it is almost perfect now. Thanks !

When i was doing pitch pumps, there was a cyclic bobble both end, i raised cyclic gain, now it is on 73%, it's almost gone now. What is the max cyclic gain that i can go with a Trex 700E DFC, 2000-2100RPM?
Author:  ZeXx86 [ Sun 05. Apr 2015 16:48:57 ]
Post subject:  Re: Extended signal processing
Thanks for great news!

You can increase Gyro Gain until it will shake (especially during slow descending). On the bigger helicopters you can achieve even 85% gains. Sometimes even 100% is possible.

But mostly 65% is optimal.
Page 1 of 1 All times are UTC + 1 hour
Powered by phpBB® Forum Software © phpBB Group
http://www.phpbb.com/