[Flyingboats]

Hi. I've been doing a conversion on a FX71 helicopter (200 sized). This is on the back of a similar conversion on a V912 using a Tarot ZYX S2 flight controller. Thought the CC3D would be good to use having accelerometers and self levelling. Anyway it's flying but not well. A few questions:

• Has anyone fitted a CC3D or similar to a similar airframe?
• Is there a list anywhere which shows reasonable minimums and maximums for stabilisation settings?

Certainly have been some unusual things going on. I was playing with 2 similar Attitude flight modes. 1 gave almost like Toilet Bowl Effect and was near impossible to fly (as had some Rate modes I'd previously tried) but heli was smooth in the air. The other flew controllably, but had a serious wobble. I progressively changed settings until they were exactly the same, and the 1st was still smooth but with TBE type effect, and the 2nd was still flyable but wobbly! I've been going back to my smooth flight modes and progressively lifting Gains but it's still not great to fly (although I haven't got the shakes occurring yet on those modes).

One thing I noticed on my ZYX file is that I settings on Roll and Pitch are right up near maximum, hence question about reasonable limits for settings including Derivatives. The 912 with ZYX is flying really well. After grounding the 71 CC3D tuning I flew a battery in the 912 chucking it around in the violent breeze. Hope I can get the CC3D setup to match it....

Some video from yesterday. Best flight was mid video in the rain in Attitude and high cyclic gains but wobbling. Reduced the gains from that point in the video, then have been increasing gains on the smooth flying mode (2) today. Latest change is to drop the Accellerometer filtering from 0.2 to 0.08, haven't flown this yet. Otherwise intention would be to keep cranking the gains higher...

Flight modes are (1) Rate, (2) Smooth Attitiude (3) Wobbly Attitude. Had tried (2) and (3) with the same settings when (2) was smooth but TBE (3) flyable but wobbly.

Look forward to any comments or suggestions!

[daveapplemotors]

Flyingboats, Welcome to LP.

I am not sure what airframe you are using. But I have controllers on 250 sized helis and I saw a CC3D board on a nano--ncpx. It'll probably work on yours. 

There is no list of reasonable maximums or minimum. The best we do so far is to post our UAV numbers on google drive:https://docs.google.com/spreadsheets/d/1vbqxQkTHJAUkiKvjfJXQcSsrPBz_NNRFiDazssHaXwc/edit#gid=0

In your UAV:

 <field values="0,0.25,0.5,0.75,1" name="ThrottleCurve1"/>
            <field values="0,0.25,0.5,0.75,1" name="ThrottleCurve2"/>
            <field values="100" name="MixerValueRoll"/>
            <field values="100" name="MixerValuePitch"/>
            <field values="100" name="MixerValueYaw"/>
            <field values="0" name="RollDifferential"/>
            <field values="1" name="FirstRollServo"/>
            <field values="Throttle" name="Curve2Source"/>
           
Should look like this:

         <field values="0,0.25,0.5,0.75,1" name="ThrottleCurve1"/>
            <field values="-1,-0.5,0,0.5,1" name="ThrottleCurve2"/>
            <field values="50" name="MixerValueRoll"/>
            <field values="50" name="MixerValuePitch"/>
            <field values="50" name="MixerValueYaw"/>
            <field values="0" name="RollDifferential"/>
            <field values="0" name="FirstRollServo"/>
            <field values="Collective" name="Curve2Source"/>
             
Your SettingsBank1
    <field values="0.004,0.008,1e-05,0.3" name="RollRatePID"/>
            <field values="0.006,0.012,5e-06,0.3" name="PitchRatePID"/>
            <field values="0.002,0.004,1e-05,0.3" name="YawRatePID"/>
Your SettingsBank2
<field values="0.003,0.0065,3.3e-05,0.3" name="RollRatePID"/>
            <field values="0.0045,0.0065,3.3e-05,0.3" name="PitchRatePID"/>
            <field values="0.0034,0.01,3.5e-05,0.3" name="YawRatePID"/>
   
These are awfully large jumps .003 to .004 Roll, 0.006 to .0045 for the Pitch P number  and since you are changing I and D values at the same time it is very complicated.

Suggestions to you for tuning:

1)  P -- Start small and raise in small increments (.0001 to .001) it until you get oscillations and then back off. Change I along with P but don't mess with D yet.  Change D only if P and I don't do the job. Do this for each axis.  Too much P and I and it is unflyable--too squirrelly-- too little P and I and the controls are mushy.       

2) Use just Rate or Attutude to tune. Switching from Rate to Attutude is adding more variables.

3) Change only one variable at a time until you get it flying pretty well.
                 

[Flyingboats]

Hi Dave,

Thanks for the quick reply, much appreciate the help 

Maybe I should back track a little. The setup is fixed pitch rather than collective pitch, so given that I'd read the Heli wizard has been buggy I decided to set it up as a plane, because there is no need for servo/swash mixing (in fact I added a little mixing to tweak the response given that the stock model does this). I did start with lower P settings on Rate, but the handling was mushy so I started increasing them. After the initial flights were almost uncontrollable (and I fly normal 3 axis/Rate happily fast and furious on fixed and collective pitch) I decided to try Attitude, and was trialling changes one at a time between Flt Modes 2 & 3. Making further changes on Fl Mode 1 in Rate was just to experiment with cranking the gains up in order to see if the wobbles would kick in, as it was flying smoothly but mushily on that.

I'll dial everything back and start again, and the settings on the Google drive are a help thanks (had looked at some posted UAV's already for reference).

If I can query exactly what the code changes do you suggested:

1. Changing throttle Curve 2: I take it this is to add Idle Up and negative pitch? Because my heli is fixed pitch only 1 throttle curve 0-100% is required.
2. Changing Mixer Roll and Mixer Pitch values to 50 i/o 100: Is this so you have 50% travel for collective and 50% travel for cyclic? If so in a fixed pitch setup all the servo movement is used for cyclic. There is no Collective
3. Could you comment what changing the First Roll Servo from 1 to 0 will do?
4. Could you comment what the figure (0.3") is for after the Derivative in the following? <field values="0.004,0.008,1e-05,0.3" name="RollRatePID"/>

Again thanks for the input. Really hoping to get this setup working but it's quite a learning curve! Not sure how the different frame dynamics impact but I'm running 6 degree blades, full headspeed will be around 3700 rpm and hover is at around 50% throttle, motor driven fixed pitch tail rather than variable pitch/main motor driven. So hovering headspeed will be very low compared to the 250's. Vibrations will be very different and lower headspeed at hover means swash movements will result in smaller changes in response. I guess with all these differences the PID settings could end up looking quite different!

[daveapplemotors]

Hey Flyingboats,

I must backtrack because I didn't read your post right. I am sorry I gave you bad information yesterday. My brain hadn't absorbed enough caffeine to register FIXED PITCH.
Yes, we saw fixed a guy with fixed pitch on CC3D but sorry I didn't learn enough at the time.

You probably have the throttle curves right. I don't know.

First roll servo==0 or 1 doesn't apply to you because of fixed pitch. I believe it names which servo to start with in the collective math.

.3 ==That is a limit. I don't know what it limits. The total correction of P ? Anyone else remember what that limits? 

Try a lower headspeed at first and then when you get control raise the speed. There is 'Thrust PID scaling' which is designed to help quads stabilize when you chop the throttle. I believe that it could be used to increase PID numbers with throttle and in your case be really handy. I haven't figured how to use in on a CP heli yet, though.

hope it helps,
Dave

PID numbers look different when I change geometry on the head slightly. 

[Flyingboats]

Hi Dave,

Cool thanks for the feedback. Will let you know how I get on.

If the worst comes to the worst I can always put it onto the 250 kit that's been sitting in the cupboard for a while....

[Flyingboats]

Haven't done a lot more as it's been really windy, although had a great thrash around in the wind with my Tarot ZYX controlled V912 yesterday, then a superb flight today as the wind died away just before sunset today.

What's bugging me at present is the difference between the 2 flight controllers. I rigged a bench test system with another CC3D to check if the response was specific to the board in my FX71 (it's not), then shot this video to demonstrate the difference in servo control between the ZYX and CC3D.

In a nutshell when rotating and holding the helicopter the CC3D moves the servos to counter the rotation. That's all good but with the heli still in that attitude the servos immediately bounce back to the neutral position, even though the heli is still in the wrong attitude. In comparison the ZYX allows the servo to move slowly back into the neutral position. I can increase the servo response (additional amplitude) by increasing the P gains on Roll and Pitch, but I can't find any way of the slowing the servo from bouncing back to the neutral position.

This fits with my observations on flying the CC3D so far, where it's largely been a manual exercise except where I had the P gains way up. The ZYX V912 is a very similar setup so is a good benchmark for what is required to make it work.

Can anyone shed any light on this and whether it's achievable somewhere within the LP settings? Fingers crossed as LP has some really nice features...



Flying the 912 in the wind yesterday plus some footage of a stock PCB FX71 in Rate mode in the same conditions:

[daveapplemotors]

What's bugging me at present is the difference between the 2 flight controllers. I rigged a bench test system with another CC3D to check if the response was specific to the board in my FX71 (it's not), then shot this video to demonstrate the difference in servo control between the ZYX and CC3D.

In a nutshell when rotating and holding the helicopter the CC3D moves the servos to counter the rotation. That's all good but with the heli still in that attitude the servos immediately bounce back to the neutral position, even though the heli is still in the wrong attitude. In comparison the ZYX allows the servo to move slowly back into the neutral position. I can increase the servo response (additional amplitude) by increasing the P gains on Roll and Pitch, but I can't find any way of the slowing the servo from bouncing back to the neutral position.

Can anyone shed any light on this and whether it's achievable somewhere within the LP settings? Fingers crossed as LP has some really nice features...

Try changing GyroTau in System / StabilizationSettings. Sorry I am so long to reply...

hope it helps

[Flyingboats]

Thanks Dave will give it a try, it's been at 0.006 but will experiment with some different values.

Have to say I'm scratching my head a bit over the setup. Had been slowly increasing P settings but continue to get this low frequency oscillation (like 1-2 seconds frequency) that's worse on pitch. If I leave it the oscillation continues to get worse and I think if left would result in a crash. I would have assumed when I've tried high P's (and I've gone as high as .006 on Pitch P) it should have been such a strong stabilising force that the heli attitude would have been relatively locked, or had a high frequency oscillation. Tried high I's and even a range of D's....

Fingers crossed the Gyro tau helps...

[Flyingboats]

Changing the Gyro Tau has definitely helped. Dropped it back to 0.03 with some more moderate cyclic P gains and there was definitely an improvement, better with higher gains. Shot this video on second flight yesterday trialling some higher P settings and had a half decent flight. Definitely moving in the right direction now thanks 

[roger1994]

I put cc3d on my old corona lmh 120 - fixed pitch with flybar. The attitude mode on roll/pitch, axislock mode on tail flies fine, need much less attention comparing with manual on roll/pitch,  axislock mode on tail. Axis lock on roll/pitch/rud files a little bit wired. It can be hands off for 1-2 seconds. Attached is my configuration.

Thanks.

[Flyingboats]

Cool good to hear. just looking back and it's ages since I've been on the LP site, have been just plugging away and banging my head against the wall for a while. Just in the last few weeks I've made some serious progress, and really it's only been in the last couple of days that I'd say I'm close to a setup that I'm happy with. It's really starting to fly well, although there are definitely a few more tweaks required.

Some things I found included that on right rudder inputs the helicopter was being pushed left so needed some right roll mixed, plus the nose was lifting. This was also showing up in 3G/Rate where often the nose would drop on left hand turns. I've mixed in 15 Roll and I'm now mixxing 37 on the Pitch elevator to cure this, piros are now good and flying behaviour is way better.

Other major thing has been using higher response levels than what i first used and dropping the servo throws back on the Tx, plus other one was using the maxx roll angles to improve Attitude agility.

Anyway video from Friday (better again yesterday) plus my latest file attached. Please note for some reason the PID's work differently on the different Banks so my Bank 1 PID's are about double 2 & 3 despite flight performance being very similar. I've also attached my tuning notes with all the changes and what i found.

[roger1994]

After putting on fuselage on lmh 120, it flies much better, I can hands-off hover for quite a while. Since I put on li-po instead of original ni-mh battery, the heli might be a little bit head-light. The down wash on the front fuselage from main rotor might help the pitch balance. The major change is to increase GyroTau to 0.05, I kept rate PID almost untouched, and increase attitude Ps.

The hovering drift horizontally due to the side-thrust from tail rotor, I will tilt the CC3D horizon to compensate this drift. Anyway, I am also happy with the heli's flying now.     

[Flyingboats]

Cool! Really glad you've had success with it. 

[jcg1541]

I am having the toilet bowl with my blade 230s (but I use oxy2 210 rotor head and blades) . This is going to be an interesting read.

[TheOtherCliff]

Here at LP, toilet bowl usually describes a GPS problem with Revo class FC's.

Since you are using CC3D, it is probably just a PID issue.  PIDs are too high.  Are you in Rate mode or Attitude?