RunRyder RC
WATCH
 1 page 1280 views Post Reply
HomeRC & Power✈️Aircraft🚁HelicopterFlybarless Rotor Head SystemsFutaba CGY › CGY 750 V2 weird swash bump after start up init
06-01-2016 05:46 AM  4 years ago
Topic Vote0Post 1
BrianApp

rrVeteran

Grand Rapids, MI

MyPosts All Forum Topic
CGY 750 V2 weird swash bump after start up init
Just setting up V2 for the first time and now instead of the sharp up, down, up swash bump after initialization it will go up then do a slow up, down, up then come back down. The tail will also twitch just before the swash bump which I don't think it did before. Is it normal and is everyone else seeing the odd slow swash up, down, up on start up? I don't know why they would change that but I hope it normal for V2.
SHARE  PM  EMAIL  GALLERY  Attn:RR  Quote
06-01-2016 05:49 AM  4 years ago
Dr.Ben

rrMaster

Richmond, VA, USA

MyPosts All Forum Topic
Normal.

Ben Minor
Peak Aircraft/Team Minicopter Team Futaba Team Kontronik USA
SHARE  PM  EMAIL  GALLERY  Attn:RR  Quote
06-01-2016 06:21 AM  4 years ago
BrianApp

rrVeteran

Grand Rapids, MI

MyPosts All Forum Topic
Thanks,

I see that for most of the precomps, such as cyclic to tail and throttle, the default is 0. Is that the way it works best and should only be changed if there is a problem? I know the beast says to start with no tail precomps but vbar has them all over the place and both work very well.

In the videos Nick suggested the 3D mode added precomps while FC3 mode is like the old 1.4 but if all the precomps are zeroed what is the difference?
SHARE  PM  EMAIL  GALLERY  Attn:RR  Quote
06-01-2016 03:16 PM  4 years ago
nickmaxwell

rrApprentice

Huntsville, AL

MyPosts All Forum Topic
Brian,

In F3C mode the actual core algorithm for how the axis's interact is different from 3D mode. F3C mode is essentially 2 cyclic axis flying their respective input where 3D mode is basing one axis's correction/response off of what the opposite axis is doing to give a more natural feel.

While the tunable precomps are set to zero (Ele Comp, Rudder and Cyclic Comp, etc) there is a internal compensation that is applied to certain inputs and commands not present in F3C mode.

Unfortunately, I could not think of any better name to describe what is going on, but now I understand using the word precomp kind of meshes the two together. In reality in 3D mode, there are background items (precomps) going on as well as the user adjustable precomps that require tuning based on model application and other variables.

You'll notice other changes as well that allow the user to tune the model easier. In the past items like "I-Gain" were pure I-Gain adjustment where as in 3D mode, there is no raw I-Gain adjustment but rather a parameter called "Heading Hold Gain" which actually not only just adjusts the I-Gain, but the heading limit, and rate constant parameters as well so the user is effectively truly changing the HEADING HOLD action of the gyro, not just a gain that can have an effect on it. This is why simply mimicking previous I-Gain settings from 1.4 (or F3C mode) into 3D mode do not seem to correlate directly. They will however be close, and in most cases actually "magically" be better due to the subsequent other adjustments made with the new parameter.

Regarding the change in initialization as Ben said, this is normal but to answer your question, it was changed because in the past if the blades were not perfectly straight during boot up, the fast pitch pump would jump or rattle the model on the bench. It just seemed rough on stuff, so it was changed to a more elegant pump so that things don't jump around if the blades are not perfectly feathered.

Hopefully this helps!
Nick
SHARE  PM  EMAIL  GALLERY  Attn:RR  Quote
06-01-2016 05:09 PM  4 years ago
BrianApp

rrVeteran

Grand Rapids, MI

MyPosts All Forum Topic
Thanks for the response.

First, excellent job on the videos. I ran them step by step next to me on the bench while I did the setup. It was like having an expert next to me explain how to set up each setting and what it does. Cool. I never understood exactly what was going on or how to set the ele to pit, ali to pit, etc mixes but now I do. As it turns out I don't need any of those but now I know what to look for and why.

Internal interaction precomps running in the background in 3D mode not accessible to the user makes sense. I just missed that idea in the video.

I get why you start with the interaction mixes set to zero but why wouldn't you want to start with at least a little of the FF? When you add collective you are going to have a torque increase and need at least some rudder. Why default this to zero? When you use cyclic you are going to need a tiny throttle boost. Why default the cyclic to gov to zero?

You are exactly right on the previous violent swash init bump. I would always makes sure the blades were exactly straight otherwise it looked like it put a lot of stress on the servos/links. Getting a twitch/check of the tail gyro is a nice idea to show that axis had a good boot up as well.

Now I need to fly it so I can see the results of all the work put into V2.

It would be nice if futaba gave the option to do setup up on a PC. Using full words, having setup up tips on the setup on screen, and embedding the manual in the setup up screen would be helpful. With the videos explaining what each setting means, what it does, and how to set it I have even more information. Futaba should host a copy of the videos directly and link to them on their manual down load page.
SHARE  PM  EMAIL  GALLERY  Attn:RR  Quote
06-01-2016 06:52 PM  4 years ago
nickmaxwell

rrApprentice

Huntsville, AL

MyPosts All Forum Topic
Brian,

Great question.

The answer to this might be a little lengthy so I apologize in advance.

Ideally, the interaction parameters (in the swash detail) are used to correct for servo geometry and linkage geometry on gyro stabilized axis (so the gyro effectively has less work to do), but also corrects for interactions on NON gyro stabilized axis (collective pitch ex- aileron/ele to collective and maximum collective disturbance from cyclic).

I realize you understand that but just wanted to clarify incase someone reading had not yet dove into the "detail" menu of swashset.

Regarding why the FF mixes set to zero initially are ideal is due to variables in aerodynamics and also varying flight conditions. For example, the ELE COMP. When ELE COMP is used a forward elevator action is applied priority over the gyro command during positive pitch movements, and back elevator is applied during negative collective pitch movements. (opposite for a counter clockwise disk). This I believe is caused in a vertical ascent/decent by induced flow through the clockwise rotation of the disk which is leaning to the right slightly due to translating tendency needing an elevator correction due to gyroscopic procession (but I do not fully understand the vertical ascent/decent aspect of the trim shift), but in forward flight this same POS PIT->Forward ELE tendency that needs corrected for is caused by the combined effects of dissymmetry of lift and transverse flow.

Where this is a bit screwy on our model helicopters with inherently unstable rotor blades and incredibly fast servos and stabilization systems while that pre-compensation generally helps, it can in turn hurt situations when in high wind, or depending on the direction of motion (the amount of translational lift in effect) is changing. So what Futaba has found is that the faster and more reactive the control system is, the less precomp you can run yet get good angle correction when the system does in fact need it, but also not during the instances where the compensation isn't needed as much.

A perfect example is of two of my identical models with different CG'd rotor blades. Actually, the helicopter with a set of blades that is LESS stable (as they are naturally going to be more reactive to command) needs LESS precomp. Actually, none. But a set of incredibly stable set of F3C blades needs MORE precomp. This is simply a speed thing. In the case of the unstable blades it requires LESS incidence chance to create a result so the system can react and correct faster after the gyro sees the angle deviation from the sensor. With precomp, you are simply adding a correction before the gyro even has a chance to see a angle of attack change to the body in all situations regardless of what the gyro is seeing. Essentially, by running the minimum amount of precomp, you are trying to eliminate the chance of the precomp being wrong before the gyro takes in information of what the body is doing.

The same theory applies to the tail rotor and the governor. On the governor it is quite simple, the faster the motor can react the easier the gov can make corrections without error. In a situation where precomp is useful from TAIL-GOV or CYCLIC-GOV would be if it is a lower powered system or slower operating frequency ESC or slower throttle servo. The downside could be, the chance of over speed in situations where the precomp adds throttle based off an input, but the rotor or tail load does not increase on the motor and in turn adds throttle until the gov sensor notes the RPM increase. EX- an over speed upright drop where you actually spool the rotor disk while adding cyclic or tail with no additional load like it would if the model was climbing or stationary.

The tail rotor can lead to more bad things, for example corkscrewing rolls or elevator flips. For example if a pilot tunes a tail rotor FF so that during full collective to full collective punches the tail stays perfectly still and it requires like 30% Feed forward, when you go to do maneuvers where the tail is weathervane (horizontal rolls) the tail input induced by the collective will break the model off a clean roll axis. The GERAT news about this, Futaba has a solution that NO OTHER FBL system has at the moment. It is called feed forward acceleration (acc.gain in the tail menu). What this does is amplify the amount of tail rotor feed forward based off how quickly the collective stick input is moved. So for example once you find a FF amount in which the model flies true and the tail is solid through most general maneuvers, but yet when you pitch pump the collective really fast the tail still seems to need more feed forward, increasing the acceleration gain will change how much tail is briefly added to hold the tail still during those aggressive changes without messing up the smoother changes.

Again, sorry for the long winded post and hopefully an answer to why FF is not always a good thing but more of less just been accepted because it generally helps things. Futaba would just like the user to be able to find the best balance of that with this tuning method and idea to start at 0 and work from there.

Nick
SHARE  PM  EMAIL  GALLERY  Attn:RR  Quote
06-01-2016 09:27 PM  4 years ago
BrianApp

rrVeteran

Grand Rapids, MI

MyPosts All Forum Topic
I get the explanation. FF is good for MOST but not all situations and in some case it can hurt so if you can work without it you are better off. The easy example is the neg pitch upright descent. The rotor actually UNloads and torque is reduced but the FF expects an increase in torque so it would actually hurt. I should have thought of that one. On a vbar that uses FF I have actually seen a tail kick on upright neg descents probably for this very reason.

One more easy question-
In the Gov rpms. In the video you show how if the RPM displayed on the 750 don't match what is shown in a futaba radio, 14SG for me, you can use the + and - to get it right on. Great, but the weird thing is when I adjusted RPM 2 to match the radio then I adjusted RPM 3 to match the radio when I went back to RPM 2 it had changed because of what I did to RPM 3.? Why would adjustments I made to RPM3 in the 750 affect what I had as right on in RPM 2? I know the 750 is probably just looking for a pulse width on ch 7 to set its rpm but when "trimming" rpm 2 or 3 why does one affect the other? When turning on the gov in the 14SG it asked to reset the endpoints to 100? and the max to 150 when used with futaba governors. If the 14SG is trying to match its endpoints to work with futaba governors why are they off in the 750? I can't get the RPM in the 750 to be an exact match to the 14SG because at a certain point changing rpm 2 will affect rpm 3 and 3 to 2 and I don't get why.
SHARE  PM  EMAIL  GALLERY  Attn:RR  Quote
06-02-2016 03:03 PM  4 years ago
nickmaxwell

rrApprentice

Huntsville, AL

MyPosts All Forum Topic
Brian,

Great and exactly.

Regarding the GOV RPMS, there is a little bar you can highlight down at the bottom of the TX GOV screen that says 1000-2000 and 700-2500 or something like that I forgot the exact numbers are try switching that and see if the RPMs match then. Default should be the 1000-2000 range.

Generally speaking I just leave the GV functions endpoints at 100/100 and limits at 135/135 and they always match up.

You might reset the gov completely and try that as well using the RESET in the Expert menu if you've tried to get them to match a lot with poking it after tweaking on the endpoints, if it gets too far out of the span or the adjustment was made in RPM 1 (low RPM) vs 2/3 (Higher RPM value) it'll be hard to ever get it back lined up between RPM 1, 2, 3.

Thanks,
Nick
SHARE  PM  EMAIL  GALLERY  Attn:RR  Quote
06-03-2016 02:43 AM  4 years ago
Dr.Ben

rrMaster

Richmond, VA, USA

MyPosts All Forum Topic
Brian,

I helped a guy this weekend whose governor was doing exactly as yours. The issue was the selected rpm range was incorrect. Once I changed that and matched the 750 to the rpm range in the transmitter, it worked exactly as you would expect.

Ben Minor
Peak Aircraft/Team Minicopter Team Futaba Team Kontronik USA
SHARE  PM  EMAIL  GALLERY  Attn:RR  Quote
06-03-2016 06:31 AM  4 years ago
BrianApp

rrVeteran

Grand Rapids, MI

MyPosts All Forum Topic
I don't know why but a reset of the governor system fixed it. Now they all track directly with the 14SG

Now I see a drift problem and am not sure what, if anything, is failing. After about 20 min working with the gov and some other stuff I look up and the swash is about 45 degrees back elevator at full pos collective. When I bring the collective down the ele servo stays where it is until the aileron and pitch servos come down to match it then they all move down even together. At full neg pitch the same thing happens after about 30 seconds. Now when I bring the collective back up the aileron and pitch servos stay still until the elevator servo moves up even to them then they all three go up even together. In the middle something weird happens after about 30 seconds as well. I didn't look into it too much because the two other problems were enough. I was going to take a video but for the next three boots I couldn't get it to happen again. I do remember a few days ago something similar happening except it would go full forward elevator instead of back. At that time I did a quick reboot and since I didn't see it again I thought it was a one time glitch. Both of these problems started after I was looking over notes or the videos for 10 - 15 minutes and when I looked up I saw a tilted swash. I was going back and forth through most of the setup menus at the time. Could leaving at a certain setup menu cause this to start or do I need to send it in? It could be a bad servo but I wouldn't expect it to "wait" for the uneven servos to line up to a level swash then move again.
SHARE  PM  EMAIL  GALLERY  Attn:RR  Quote
WATCH
 1 page 1280 views Post Reply
HomeRC & Power✈️Aircraft🚁HelicopterFlybarless Rotor Head SystemsFutaba CGY › CGY 750 V2 weird swash bump after start up init
Print TOPIC

 3  Topic Subscribe

Monday, November 30 - 8:00 pm - Copyright © 2000-2020 RunRyder   EMAILEnable Cookies

Login Here
 New Subscriptions 
 Buddies Online