Jump to content


450 FBL Heli - CC on board :)


  • Please log in to reply
39 replies to this topic

#21 Harbormaster

Harbormaster

    Key Member

  • Members
  • PipPipPip
  • 184 posts
  • LocationSeattle
  • Country: flag of United States United States


Posted 19 January 2012 - 09:26 PM

View PostRicoAlonso, on 19 January 2012 - 08:44 PM, said:

wow! just did a quick hover at the driveway and this 13mph wind is just too much for this small heli. I guess I'd better stop now while still ahead :) I almost dumped it. :P

That's the spirit! :)

#22 SargeNZ

SargeNZ

    Frequent Prop Purchaser

  • Members
  • PipPipPip
  • 89 posts
  • Country: flag of New Zealand New Zealand


Posted 19 January 2012 - 09:27 PM

This is promising stuff. Harbormaster, what FBL head are you using? I just got donated an HK450GT airframe and am considering doing the DIY flybarless conversion from stock parts as described here: http://rc.runryder.c...0_for_SK360.pdf and then flying on CC.I might have to put the training skids back on :)

#23 Harbormaster

Harbormaster

    Key Member

  • Members
  • PipPipPip
  • 184 posts
  • LocationSeattle
  • Country: flag of United States United States


Posted 19 January 2012 - 10:05 PM

View PostSargeNZ, on 19 January 2012 - 09:27 PM, said:

This is promising stuff. Harbormaster, what FBL head are you using? I just got donated an HK450GT airframe and am considering doing the DIY flybarless conversion from stock parts as described here: http://rc.runryder.c...0_for_SK360.pdf and then flying on CC.I might have to put the training skids back on :)

On my FBL T-Rex 450, I'm using the stock Align FBL head, which works well except the ball links on the head are so long that they have a long lever arm and can be broken out of the soft aluminum threaded holes when bad things happen. I would buy something more robust if I had to do it again. The adjustable phasing on the Align head always seems to take some fussing with to get right, as well. A head with the driver arms locked to it seems like it would be easier to set up. Many people seem to like the RJX. I suspect that which head you chose may not make a big difference from a performance standpoint.

My 250 has a stock Align head with a Mikado FBL conversion kit. It seems to work well.

I'm loving the OpenPilot project, but if you're doing your first flybarless build you might want to consider a BeastX, which is pretty much plug-and-play--people seem to love them--that is unless you enjoy fussing with the settings like the rest of the heligeeks here and watching the CopterControl platform evolve; then by all means please join us. :) Please feel free to PM me if you want to chat...

#24 wrench76

wrench76

    Advanced Member

  • Members
  • PipPipPip
  • 77 posts
  • LocationNew York
  • Country: flag of United States United States


Posted 20 January 2012 - 01:17 AM

The RJX head is the best I have found http://www.helifligh...it_p/fl450u.htm  ;) it is the one I'm going to buy.

@ SargeNZ I am useing that type of head setup on my T-Rex 450 at the moment, it works very well. It's not ideal but it works well.

#25 Pellicle

Pellicle

    Advanced Member

  • Members
  • PipPipPip
  • 79 posts
  • LocationNew York, US
  • Country: flag of United States United States


Posted 20 January 2012 - 05:22 AM

Ki in rate should give you stability in windy conditions and help hold the attitude of the copter in flight.  I think this is why Wrench mentioned it.
It sums up the difference of deviation over time from desired attitude which you have manually chosen and applies this towards the output.  This helps it hold desired target since Kp only may not hold the target and in some conditions can drift from desired attitude.  In stable conditions like however you can do fine without Ki.  Kd reacts to the rate change in difference.  So it reacts more strongly for high rates of change.  So it the difference is changing more rapidly it applies more correction. This acts as a damping factor and can help quicker reaction to upsets.

I ran without Ki and it was fine till I went out on a windy day.  Then it was very squirrelly.  Not as bad as when I accidentally flew in manual mode because I forgot I had set it to manual :wub: for some mechanical linkage adjustment.  The  wind just blew it around.   But hovering in the shop it flew fine in both cases.

#26 RicoAlonso

RicoAlonso

    Key Member

  • Members
  • PipPipPip
  • 165 posts
  • LocationDerby, KS
  • Country: flag of United States United States


Posted 20 January 2012 - 10:19 AM

Pellicle,
Thanks for correcting my understanding of Kd.

For now, I am tuning for Rate-Rate-Rate stabilization mode so yes, stable in no wind but gets blown around when windy is the kind of stability that I am targeting or looking for now. I just want to soften the sudden change to give me time to react.  That is also what I observe if I fly my other flybarred heli and put the GY401 tail gyro in rate mode, although it's a bigger heli and usually can handle the wind better than the 450. So far, the CC seems to be doing good with my tuning.

When I'm satisfied with Rate-Rate-Rate mode tuning then I will turn to the AxisLock (or as Heli guys call HH) and Attitude (which I think is equivalent to self-level) tuning. In the AxisLock and Attitude mode I'd start tuning the Ki in the OuterLoop. And with that, I will be expecting that AxisLock will hold the desired axis position and that Attitude will perform auto-level the cyclic even during windy condition.

I know I was able to fly this heli in Attitude-Attitude-AxisLock mode the first time I did my test last December or early January and I liked it. However, I was rushing things and keep changing things aroung without any plan of actions and that started to mess up the AxisLock of the Tail. It got 'jumpy'. I suspect it was the settings in the InnerLoop fighting with the setting in the OuterLoop. So this time, I'll do it slowly and tune it one mode at a time. :)

#27 RicoAlonso

RicoAlonso

    Key Member

  • Members
  • PipPipPip
  • 165 posts
  • LocationDerby, KS
  • Country: flag of United States United States


Posted 21 January 2012 - 09:16 PM

Alright, did additional tuning today and did it inside the garage to make sure no wind interference and I think (and I'm satisfied) with the cyclic PID's both in the Inner Loop and Outer Loop. Did Rate-Rate-Rate and AxisLock-AxisLock-AxisLock hovering with very stable results.

Then, I level my heli and did the Attitude calibration on the bench and did Level. However, as I hovered in Attitude-Attitude-AxisLock mode oscillation starts. Will continue tuning. Still on the cyclic. Will work on the tail when cyclic is all good.

I guess I was wrong (again). After I tuned for Rate, I should have done Attitude tuning and not worry about AxisLock as it come in automatically if Attitude stabilization is tuned.

Got it all done for the cyclic except I still need to tune the AccelBias. I see forward and left side drift when hovering at Attitude mode for the cyclic.

#28 Harbormaster

Harbormaster

    Key Member

  • Members
  • PipPipPip
  • 184 posts
  • LocationSeattle
  • Country: flag of United States United States


Posted 24 January 2012 - 10:40 PM

View PostHarbormaster, on 19 January 2012 - 05:38 PM, said:

That's the table below that I added to the Stabilization Panel topic. I think it will be useful (it helped me to finally wrap my head around the settings) but I wanted to have people take a look at it to make sure it's correct (especially for the non-heli airframes).


Update just FYI

Posted Image

#29 RicoAlonso

RicoAlonso

    Key Member

  • Members
  • PipPipPip
  • 165 posts
  • LocationDerby, KS
  • Country: flag of United States United States


Posted 24 January 2012 - 10:45 PM

Thanks. Got it :)

BTW, I restarted my tuning. All the while I thought I had tuned the Rate mode well and it was doing good. However, while tuning the Attitude mode, I noticed it does not have much authority... it can go level but onced disturbed, it can't recover fast enough...so I increased the Maximum rate in attitude mode. After that, the heli oscillated like crazy when in attitude mode. I lovered the Kp even to 1.0 (50% of the default) and still I see oscillation when some gust of wind would hit it. Maybe my Rate mode was not tuned well after all, so I am restarting and logging things in a separate note. Maybe when I'm done with everything I can share it.

#30 wrench76

wrench76

    Advanced Member

  • Members
  • PipPipPip
  • 77 posts
  • LocationNew York
  • Country: flag of United States United States


Posted 25 January 2012 - 02:41 AM

Hey Ricco
I'm curious what your inner rate mode settings were.  How was it in rate mode? I am only tuning for rate mode on my cyclic. I had a chance to fly two days ago, it was quite windy and cold, so no real good testing was done. I didn't have too much trouble with the wind, unless I tried to let it hover without input for some time (3-5 seconds), whichever side of the rotor disc was headed into the wind would raise up and then would continue in that direction. But it was cold and I didn't have any way of making any adjustments at the field. It was manageable though.
I did see some strange behavior on the bench in regards to the Full Stick Rate and Maximum Rate In Attitude Mode. If these two settings are not the same for a given axis, I would see a slower response of the swash plate at the limits. If my roll was set to 300 FSR and 360 MRAM and a full stick input was given, the swash would move quickly to about 80% and then slow down the rest of the way. If I switched them around, 360 FSR and 300 MRAM the same thing would happen. Only when both FSR and MRAM were set equal would the swash movement be consistent throughout the full range.

#31 RicoAlonso

RicoAlonso

    Key Member

  • Members
  • PipPipPip
  • 165 posts
  • LocationDerby, KS
  • Country: flag of United States United States


Posted 25 January 2012 - 03:17 AM

wrench,
Before I changed back everything to default just this evening, I had these settings:

GyroTau = 0.05

Rate Stabilization (Inner Loop):

Kp = 0.0025; Ki = 0.0000; ILimit = 0.000; Kd = 0.00005

Kp = 0.0025; Ki = 0.0000; ILimit = 0.000; Kd = 0.00005

Kp = 0.0035; Ki = 0.0035; ILimit = 0.300; Kd = 0.00000


Attitude Stabilization (Outer Loop):

Kp = 1.0; Ki  = 1.0; ILimit = 10.0

Kp = 1.0; Ki  = 1.0; ILimit = 10.0

Kp = 2.0; Ki  = 0.0; ILimit = 50.0


Stick Range and Limits

Full Stick angle = 100, 100, 100

Full Stick Rate = 400, 400, 300

Maximum rate in attitude mode = 400, 400, 300


I was just working on the cyclic and was planning on working at the tail after I had the Rate and Attitude for the cyclic properly tuned. I was tuning the auto-level (from Attitude mode) yesterday when I found that no matter how much I change the AccelBias, it seems that it wont level. That's when I remember to increase my MRAM to 500 and suddenly the heli went into violent oscillation. I lowered it to 400 and it was more manageable and the model seemed to level better. However,  I also need to lower the Kp in Attitude Stabilization because I can notice some cyclic pitch oscillation when a gust would hit the heli, while in Attitude mode.

I'm not satisifed so I'm going back to sqaure one... back to default values and start with Rate again.

#32 RicoAlonso

RicoAlonso

    Key Member

  • Members
  • PipPipPip
  • 165 posts
  • LocationDerby, KS
  • Country: flag of United States United States


Posted 25 January 2012 - 10:31 PM

You know what, I missed the AccelKp which was discussed in some thread. When I noticed that the heli won't go back to level and keep on drifting after an initial gust, I should have tried and increased the AccelKp instead of increasing my MRAM.

Well, I'll make sure to remember that next time.

#33 RicoAlonso

RicoAlonso

    Key Member

  • Members
  • PipPipPip
  • 165 posts
  • LocationDerby, KS
  • Country: flag of United States United States


Posted 28 January 2012 - 11:40 PM

Did another set of tuning today. I have to do it inside the garage for no-wind condition. Did the Rate and Attitude tuning. Without wind, the heli flies so good in Rate and Attitude. AxisLock doesn't seem to lock the tail at all. I even did more mechanical tuning on the linkage.

When in Attitude, I just can not keep the Heli fly with no drift. It keeps drifting to the left. I even did the Level procedure with almost half and inch shim under the left side of the skid to force it to tilt to the right. I also increased the AccelKp. I think it's really difficult for the Heli due to the torque from the main blade and the force from the tail blade.

I tried flying it outside with gusts of about 10mph and the heli will be all over the place. Of course, I'm trying to confine it in a small front yard. One more thing, my GyroTau is at 0.25 right now and so far, I haven't notice any oscillation on the pitch this time.

Time to rest. Had been spending much time on this, maybe I should fly my other Heli with the Co-Pilot2 or maybe my flat foamie to get some break :) and have fun in the wind.

Attached Files



#34 dankers

dankers

    Janitor

  • Administrators
  • 5125 posts
  • Country: flag of Australia Australia


Posted 29 January 2012 - 03:50 AM

Quote

I just can not keep the Heli fly with no drift. It keeps drifting to the left.

Ah, well this is normal!!

It's your tail rotor, for every action there is an equal and opposite reaction. The tail rotor counter acts the spin of the main rotor but also pushes the heli to the to the left (on a heli with clockwise main blade rotation).

So when you fly in rate mode and you hover in one spot, you will have right skid low to counter act this, most pilots do this without thinking about it. However, the CC will level the heli and hence it will get blown to the left by the tail rotor as not enough right cyclic is being added.

Rather than shim it to get it level, play with the accel bias in the UAVObect browser.

Posted Image

For the FBL heli set up, we need a setting that allows this off-set to be activated in self level mode.

But for now, mess with the accel bias in the Objectbrowser. You will also be surprised at how much angle you need, it is more than you think but take it slowly.

#35 RicoAlonso

RicoAlonso

    Key Member

  • Members
  • PipPipPip
  • 165 posts
  • LocationDerby, KS
  • Country: flag of United States United States


Posted 29 January 2012 - 01:44 PM

View Postdankers, on 29 January 2012 - 03:50 AM, said:

Ah, well this is normal!!

It's your tail rotor, for every action there is an equal and opposite reaction. The tail rotor counter acts the spin of the main rotor but also pushes the heli to the to the left (on a heli with clockwise main blade rotation).
...

Yes, that's what it is... but I never thought it would take that much value in the AccelBias-Y for the tilt needed. Right now the AccelBias-Y is already at -45. Since you had confirmed it,  then I'll keep going. For a while there, I thought I was not doing it right. I hope someday we can equate this number to how much angle in degress tilt to the left or right... would be easier to visualize :)

Thanks.
Rico.

#36 dankers

dankers

    Janitor

  • Administrators
  • 5125 posts
  • Country: flag of Australia Australia


Posted 29 January 2012 - 03:03 PM

Quote

I hope someday we can equate this number to how much angle in degress tilt to the left or right... would be easier to visualize

In the Heli set up screen the plan is to do exactly that. We really don't want "normal" people to mess with the UAVObjects directly but sorry, right now only way. That number relates to the bias offset in the firmware, I have no idea what it related to in deg.

So what I would do is have the heli connected, have it level and change the value while watching the PFD, as you change it you will see the PFD tilt and that will give to an indication of how many deg you are off setting.

#37 RicoAlonso

RicoAlonso

    Key Member

  • Members
  • PipPipPip
  • 165 posts
  • LocationDerby, KS
  • Country: flag of United States United States


Posted 29 January 2012 - 05:48 PM

View Postdankers, on 29 January 2012 - 03:03 PM, said:

...
So what I would do is have the heli connected, have it level and change the value while watching the PFD, as you change it you will see the PFD tilt and that will give to an indication of how many deg you are off setting.

Oh thank you for mentioning :) I will do that now.  Why didn't I thought of that? :(  oh well... so may times, I get overwhelmed, I forget to use common sense :unsure:

#38 RicoAlonso

RicoAlonso

    Key Member

  • Members
  • PipPipPip
  • 165 posts
  • LocationDerby, KS
  • Country: flag of United States United States


Posted 05 February 2012 - 02:39 AM

Well, I guess that's it for now. My day job got so demanding lately that I can no longer have enough time to play around. And so, I'll close this thread with the last video of me still doing the Rate tuning. Maybe some other time when work is not too heavy, I will have time again to play.



Thanks a lot everyone. I had fun.
Rico.

#39 Harbormaster

Harbormaster

    Key Member

  • Members
  • PipPipPip
  • 184 posts
  • LocationSeattle
  • Country: flag of United States United States


Posted 07 February 2012 - 02:44 AM

Hey Rico -- Looking on the bright side, it's good to be employed in this economy!

Come back and play soon! :)

#40 RicoAlonso

RicoAlonso

    Key Member

  • Members
  • PipPipPip
  • 165 posts
  • LocationDerby, KS
  • Country: flag of United States United States


Posted 07 February 2012 - 03:55 AM

True. Yup, I'll get myself updated from time to time when I have a chance. You guys have fun. :)