HobbyKing.com New Products Flash Sale
Reply
Thread Tools
Old Mar 26, 2012, 11:49 AM
Registered User
mike00top's Avatar
Launceston UK
Joined Mar 2010
33 Posts
Top marks John NZ for software, and copterx for cx-3x1000 as these teamed together is a winner,

i for one would not have got a cx-3x1000 otherwise, as my pet hate is programming using sticks, bleeps and led flashes

just been out and test flown my 450 pro, i used all stock settings ( 3D) except for taking roll and pitch gains to 30 right from the start, (most people say the 40 is too high) tail gain on my TX is +35 HH, advanced settings left where they are, and 20% expo, 30% might be better, as its sensitive around centre

This gyro flew BETTER than my Align FL760, the tail is maybe 2 clicks too much gain otherwise perfect.

Copter X ought to be contacting John NZ for a Job, as using the pair together is the best FBL experience i have had.

i have had to do the Static mod with the FL760, so i left it on the model, so i cant comment on if its more or less static sensitive.
mike00top is offline Find More Posts by mike00top
Reply With Quote
Sign up now
to remove ads between posts
Old Mar 26, 2012, 11:53 AM
just gotta mess with it!
2Doggs's Avatar
North West Louisiana
Joined Nov 2009
4,710 Posts
Quote:
Originally Posted by Dawk View Post
Oh my 14 mm seems to be waaaay to much then !
That's probably why I have so much shaking.
Thank you very much, I'll try to reduce to near 8mm.
The thing is that by doing this, the "arm to swash" link will no more be vertical. But I suppose it does not matter ?
It does matter a little - all the 450 linkages I've looked at seem to be optimised for 1/2" (12.7mm) servo arms, which result in more or less vertical links to the swash. Any angle will result in some differential movement - but that will be a lesser effect than the wild oscillations you're currently getting.

The big problem with the CopterX FBL heads is that their blade grip arms are too short. It's a bit like Goldylocks and the Three Bears..... The RJX head has arms that are too long, the CX head has arms that are too short....but the Tarot head has arms that are just right..... at about 18mm to the feathering spindle!

What matters is that you should be able to use a collective range setting in the FBL controller of around 60, to get -12 to +12 pitch. If it's much less than that, you need to use shorter servo arms, and if it's much more, longer ones.
2Doggs is offline Find More Posts by 2Doggs
RCG Plus Member
Latest blog entry: Nano Tech 1800mah 4S 25-50c
Reply With Quote
Old Mar 26, 2012, 11:58 AM
Registered User
mike00top's Avatar
Launceston UK
Joined Mar 2010
33 Posts
Quote:
Originally Posted by Dawk View Post
Oh my 14 mm seems to be waaaay to much then !
That's probably why I have so much shaking.
Thank you very much, I'll try to reduce to near 8mm.
The thing is that by doing this, the "arm to swash" link will no more be vertical. But I suppose it does not matter ?
shorter arms is better for the servo too ! they will deliver more torque that way, and and any slop in gear train will be reduced to the swash i think, and less force will be exerted on the gears.

the only negative is that too short a CCPM arms will show up interactions worse at full pitch and full neg pitch, when using ail and elev.

i think its been said before, but you cant go far wrong, 60 to 70 on collective travel (John NZ software) and find the right arm length to match, most models about 9mm - 13mm


Sorry 2doggs posted a few mins before
mike00top is offline Find More Posts by mike00top
Last edited by mike00top; Mar 26, 2012 at 12:15 PM.
Reply With Quote
Old Mar 26, 2012, 01:08 PM
Registered User
France, IdF, Paris
Joined Mar 2012
44 Posts
Thank you again John (and 2Doggs and mike for your explanations).
It worked. No more "wobble of death" !
Dawk is offline Find More Posts by Dawk
Reply With Quote
Old Mar 26, 2012, 01:17 PM
Registered User
France, IdF, Paris
Joined Mar 2012
44 Posts
Quote:
Originally Posted by tdriver View Post
new tarot unit.... link
It's a gyro + a receiver ?
Is it like AR7200BX ?
But I don't think so.
I don't understand the "now it will work also with satellite" function.
Dawk is offline Find More Posts by Dawk
Reply With Quote
Old Mar 26, 2012, 02:25 PM
Registered User
DerClown's Avatar
Lithuania
Joined Mar 2009
220 Posts
Quote:
Originally Posted by Dawk View Post
It's a gyro + a receiver ?
Is it like AR7200BX ?
But I don't think so.
I don't understand the "now it will work also with satellite" function.
I believe it would be same as Align 3GX. Just plug in satellite receiver and you dont need the main DSM2 receiver. Sweet !!!! Correct me if I'm wrong
DerClown is offline Find More Posts by DerClown
Reply With Quote
Old Mar 26, 2012, 02:36 PM
Registered User
France, IdF, Paris
Joined Mar 2012
44 Posts
So you plug directly the servos and power into the gyro ? Seems nice.
Dawk is offline Find More Posts by Dawk
Last edited by Dawk; Mar 26, 2012 at 02:53 PM.
Reply With Quote
Old Mar 27, 2012, 12:49 AM
as much as I can
beenflying's Avatar
NZ
Joined Jan 2010
4,912 Posts
It looks like you're going to need a Y connector if you want to use an external BEC. There will be no spare socket, with power and ground, with all servos, ESC and sat connected.
beenflying is offline Find More Posts by beenflying
Reply With Quote
Old Mar 27, 2012, 02:03 AM
Registered User
United States, CA, Berkeley
Joined Jul 2011
120 Posts
Quote:
Originally Posted by mike00top View Post
shorter arms is better for the servo too ! they will deliver more torque that way, and and any slop in gear train will be reduced to the swash i think, and less force will be exerted on the gears.
....
not really. it is actually the opposite. the servo and the gears will see more load if the "moment arm" is shorter (less leverage). The only positive is that when full range of servo travel is not utilized, shorter arm results in servo having to move more so you will have better resolution.

In any case, these are minor relative to the importance of having vertical links and orthogonal configuration the best you can.

Frankly, I am surprised and reluctant to believe shorter arm had positive effect on oscillations since they could also occur when in "perfect hands-free hover" (if there is such a thing) where servos are not even fully engaged. If oscillation is observed while in transition, then it most likely has to do with PID adjustments rather than servo arm length.
asasan is offline Find More Posts by asasan
Last edited by asasan; Mar 27, 2012 at 06:00 AM.
Reply With Quote
Old Mar 27, 2012, 03:35 AM
Registered User
New Zealand, Taranaki, New Plymouth
Joined Jan 2012
333 Posts
Quote:
Originally Posted by asasan View Post
.

Frankly, I am surprised and reluctant to believe shorter arm had positive effect on oscillations since they could also occur when in "perfect hands-free hover" (if there is such a thing) where servos are not even fully engaged. If oscillation is observed while in transition, then it most likely has to do with PID adjustments rather than servo arm length.
My reasoning is this. The gyro must be hard programmed for a particular amount of aileron or elevation change per degree movement of the servo arms. This must be the reason that Tarot quote the required length for the arm. So when the gyro senses a movement away from the desired position, it responds with the input to the servos that it thinks it needs to correct the situation. If the servo arm is too long then it gets more correction than it bargained for and over corrects, then responds with a bigger correction in the opposite direction making it worse, so producing the wobble of death. Of course the parameters of the PID model can be changed to compensate for this, but only to a point. The experience has been, no matter how far you turn them down the WOD haunts you. But shortening the arms to give the same movement that the Tarot heli has puts it back into the right ballpark and then any mismatch is easily handled by the PID parameters.
John_NZ is offline Find More Posts by John_NZ
Reply With Quote
Old Mar 27, 2012, 04:56 AM
Team Mulikow 3D
Oxfordshire, UK
Joined Jan 2007
4,118 Posts
Quote:
Originally Posted by John_NZ View Post
My reasoning is this. The gyro must be hard programmed for a particular amount of aileron or elevation change per degree movement of the servo arms. This must be the reason that Tarot quote the required length for the arm. So when the gyro senses a movement away from the desired position, it responds with the input to the servos that it thinks it needs to correct the situation. If the servo arm is too long then it gets more correction than it bargained for and over corrects, then responds with a bigger correction in the opposite direction making it worse, so producing the wobble of death. Of course the parameters of the PID model can be changed to compensate for this, but only to a point. The experience has been, no matter how far you turn them down the WOD haunts you. But shortening the arms to give the same movement that the Tarot heli has puts it back into the right ballpark and then any mismatch is easily handled by the PID parameters.
This sort-of matches what Beastx gets you to do - you have to show it 6 degrees of cyclic and it feeds back on whether or not the geometry is optimised - it is looking for a certain mount of servo movement = 6 degrees.
MikeHRC is online now Find More Posts by MikeHRC
Reply With Quote
Old Mar 27, 2012, 05:15 AM
Registered User
United States, CA, Berkeley
Joined Jul 2011
120 Posts
Quote:
Originally Posted by John_NZ View Post
My reasoning is this. The gyro must be hard programmed for a particular amount of aileron or elevation change per degree movement of the servo arms. ...
For the gyro to have any engineering value in performing it's primary function, it should understand and accept the parameters during the setup where the limits of movement is defined to bring scale or scope to the function.

One factor is the servo's range of movement and utilizing close to full range is optimal. For that, if the demand is small, say max collective of +/- 12 or 14 can be achieved using less than 2/3 of what servo can deliver, then shorter horn corresponding to bigger required movement would help. Now, the more important factor is the scope of movement of the swash. Obviously, since this gyro can be used on any size heli, edge movement required by a 250 swash to produce +/- 8 degrees of cyclic deflection is a fraction of what a 700 or 90 swash needs. Even on the same 450 size setup, I may prefer only 10 collective and 5 cyclic where others like 12 and 9. These are all set in your software and define the range and scale of total movement available to the gyro. It better not be "hard programmed" to deal with say only 7.5mm horn or else overcompensation, confusion and WOD or WMD. In other words, the correction should be relative in context of each specific heli because that move per degree of servo rotation is clearly defined during setup.
asasan is offline Find More Posts by asasan
Last edited by asasan; Mar 27, 2012 at 05:22 AM.
Reply With Quote
Old Mar 27, 2012, 05:55 AM
Registered User
United States, CA, Berkeley
Joined Jul 2011
120 Posts
Quote:
Originally Posted by 2Doggs View Post
It does matter a little - all the 450 linkages I've looked at seem to be optimised for 1/2" (12.7mm) servo arms, which result in more or less vertical links to the swash. Any angle will result in some differential movement - but that will be a lesser effect than the wild oscillations you're currently getting.

The big problem with the CopterX FBL heads is that their blade grip arms are too short. It's a bit like Goldylocks and the Three Bears..... The RJX head has arms that are too long, the CX head has arms that are too short....but the Tarot head has arms that are just right..... at about 18mm to the feathering spindle!

What matters is that you should be able to use a collective range setting in the FBL controller of around 60, to get -12 to +12 pitch. If it's much less than that, you need to use shorter servo arms, and if it's much more, longer ones.
I agree with everything you said with the exception of the whole middle paragraph.

Imagine the grips being four inches long grabbing one third of the blade. It would just be an overkill of a hold and not make any difference since all they do is rotate the blade around an axis that is parallel to their own long length. In engineering terms, the moment of inertia (I) of the grip's cross section remains the same with any length for that rotation.

Edit: Just realized that you are probably not talking about the length as you stated, but how far the linkages are from the center in the transverse direction (wide/narrow vs. short/long). If that is what you mean, then you are right on again. Excuse the babbling. Nevertheless, the gyro should take that into account at setup (what does it take to rotate 12 degrees given what we have) and it shouldn't pose a problem.
asasan is offline Find More Posts by asasan
Last edited by asasan; Mar 27, 2012 at 06:39 AM.
Reply With Quote
Old Mar 27, 2012, 11:28 AM
Registered User
New Zealand, Taranaki, New Plymouth
Joined Jan 2012
333 Posts
Quote:
Originally Posted by asasan View Post
For the gyro to have any engineering value in performing it's primary function, it should understand and accept the parameters during the setup where the limits of movement is defined to bring scale or scope to the function...
I still stand by my thinking as the gyro still gets feedback from the applied input to the servos as a response from the accelerometers as the heli changes attitude.
The so called hard programmed parameter would effectively be 'scaled' proportionally by the size of the helicopter that is it fitted to. This would mean as long as the same servo movement produces the same change in attitude at the same rate, there is no reason to make it a user changeable parameter for different scale Helis.
Mike's comment that the BeastX has a setting for this exact parameter surely must infer that since the Tarot has no such external parameter that it must have an assumed internal one.
John_NZ is offline Find More Posts by John_NZ
Reply With Quote
Old Mar 27, 2012, 12:23 PM
Registered User
Sunnyvale, CA, USA
Joined Apr 2000
1,783 Posts
Asasan - your comments indicate some familiarity with control theory.

Do not forget, however, that the servos have a dead band and differing small and large signal response characteristics. These are constant at the servo arm in angular terms and therefore have varying effect as you change the mechanical gain (servo angle to blade angle). This effect shows up in terms of a "sweet spot" with respect to the overall system loop response. If you have way too much mechanical gain and compensate by turning down the electrical gain, the servo dead band and small signal response will become more and more critical - fast, tight digital servos are always better than poor analogs.

As a general comment on the way the units are "hard coded" if you increase the overall gain and reduce the PID gains to compensate it can be seen the feed forward term (immediate response of the swash to stick movement) will reflect the increase in response to the increase in overall gain.

The process is the equivalent of changing the bell/hiller mixing ratio and should have similar effects on handling. I have not yet had time to play with this in the air. I will say that this ability is explicit with the Skookum and is a very valuable feature. I suspect that the design point in the Tarot is optimized for 450 size 3D but could be better for larger models set up for scale/precision flying.
teamdavey is offline Find More Posts by teamdavey
Reply With Quote
Reply


Thread Tools

Similar Threads
Category Thread Thread Starter Forum Replies Last Post
Sold Darthdrk Drone CX Body for coax helis like the Blade CX etc. Twmaster Aircraft - Electric - Helis (FS/W) 0 Oct 08, 2011 07:31 PM
Multiplex mc V2 servo programming PC software Thomas B Radios 3 Jul 23, 2006 09:49 AM