Leap's going flybarless... GyroBot configuration

[leapfrog]

Okay...  I think I'll post the individual setups in a separate topic for each (to keep from having to scroll through a single long thread).

So, first up is the GyroBot configuration.  It's a three-axis stabilizer with CCPM mixer.  It uses three MEMS gyros and a three-axis accelerometer internally, and software that not only mixes the swash servo configurations - but also adds control feedback with differing levels of gain to provide separate flight "modes" (e.g. "beginner", "standard", and "advanced").  I have the latest release of the firmware - 1.05 - and this version only supports the "standard" and "advanced" modes of flight regime.

The manual says: "In the Beginner Mode each model can be controlled as easy as a coaxial helicopter.  The GyroBot will always navigate the helicopter back into the horizontal as soon as you loose the sticks. (for aileron and elevator only)  The Beginner Mode is not yet activated in the actual software V1.05 and will be offered as software update at a later date.  Therefore it is not yet described in this version."  In addition, the manual goes on to say: "In the in version V1.05 automatically activated "Standard Mode" you reach, compared to conventional paddle rotor heads, a fully adjustable, very precise flight control.  So to say, the model seems to fly "on rails".  Furthermore, you can even during the flight switch the models agility from very calm to extremely agile."

(that is a direct quote from the manual - including translation issues)  I'll upload the manual as well, so you can follow along.  *smile*

So... what's in the box?  The GyroBot system unit, the "Cockpit" interface, the USB interface, several servo jumpers (the interfaces to the RX and the Cockpit), software CD, and a *bunch* of stickers.

I ordered the GyroBot/Cockpit combo... and the USB cable. (wierd, that you can order the unit, "Cockpit" interface, and USB cable separately, as a combo with the unit and the "Cockpit" interface, or as a combo with the unit and USB interface cable - but *not* as a combo with all three.  Here's the deal... you *need* the USB cable to update the firmware and/or configure the unit using a laptop.  The "Cockpit" interface allows you to modify the settings at the field without a laptop... but if you get the combo of the unit and the interface and not the USB cable... how are you going to update the firmware later on down the road?)

So I just uploaded a pic of the GyroBot sitting next to a *very* familiar GY401 for scale reference, and the Cockpit interface.  Both the GyroBot and the Cockpit are incredibly well finished - each housed in a CNC'd aluminum housing and polished to a mirror finish.  These are just very pretty boxes.  We'll soon see if they perform as well as they look... *smile*

(the manual is the pdf link just below the last picture)

[harry]

Great stuff .. I might start the same review style for the SK360

[leapfrog]

Okay...  Update time.

I got some time to start the "configuration" portion of the GyroBot setup... and it appears that the unit is made for taking an existing setup and converting it to flybarless... the mechanical configuration is "limited" by the software...  and using a mechanically limited setup will simply require greater travel limits in the software...

The setup is really straight forward, with adjustments made to only one item (e.g. servo, mix, function, or axis) at a time.  That makes for quite a few menus, but they are ordered logically and navigation is really easy.  There is a "calibration" function that must be used each time something is changed from the stored configuration.  The menu navigation is managed by the single "mode" pushbutton on the "Cockpit" interface (or cockpit emulator software) and the TX sticks...  with menu selections made with the rudder stick, and "values" set with the cyclic stick.  Feedback is interactive - meaning what you select or adjust is actually shown on the model real time.  This is similar to other mixers like the Helitronix, etc...  and is far easier than the "adjust-load-test" mixer interfaces.

So far, so good.  The head is now set up with the correct "directions" and "mixing"...  so I'm gonna finish the mechanics reassembly...  and I am waiting on some little items that will make the head pretty stinkin' great...  stay tuned.

[BrockJ]

That's some fancy guide ties on the extensions.
 Will this mount under the rear tray like a gyro?
And is there a seperate gyro component or is it all internal?  It's small

[J-Heli]

Cool! Staying tuned!

[leapfrog]

Hehe... that is waxed lacing - what is used on full-scale (in fact, that would be a FAA-approved bundle as I had to learn Part 23 to work on punkin'...  ).

All three axis are internal to the box...  gyros for each of the axis, and a three-axis accelerometer - they mix the signals with various gains (if you are familiar with a closed control loop)... and it makes for a really precise little package.  I'm mounting it on top of the boom block - right where I put the gyros...   and the RX is either going on a "diving board" on the bottom frame - or on the RX tray (probably the tray) like my other 600 builds.  There is a specific orientation it has to be placed in... and if you notice, I left "play" in the bundle at the GyroBot end... the bundle will be laced to the frames between the boom block front and rear mounts in such a way as to not add a "bias" to the unit, and to remove any possibility of "induced" vibration due to the cable's inertia.  This thing is sensitive.  *smile*

Oh yeah...  I forgot to mention... the "Cockpit" interface has a nifty four-line backlit LCD...  it is a work of art.

[leapfrog]

Okie doke... the bird is ready for maiden (mechanically)... I am playing with the software to better understand how each menu works, and how the software does what it does.  I am waiting to receive some "bits" to finalize (even out) the mechanical setup - like the Align grip extensions that Harry posted...  but I couldn't wait to fly the bird (the weather is getting worse, and I want to get as much stuff in as I can...  no worries, though... I'll be sure to keep things totally even when the mechanics are solid).

There are three "Phases" that can (should be) programmed into the GyroBot, and are "switchable" depending on your radio setup.  These "phases" are different from the familiar "flight modes" that you usually program into a radio to change the throttle/pitch curves...  you can think of a phase as a "personality", depending on how the settings are configured.  Remember, we are talking a three-axis device - there is no external tail gyro... so the P and I summing feedback gains are managed entirely within the unit for *all* three axis on the bird... (aileron/elevator/rudder)... so the personalities that are possible are "diverse" by any stretch of the imagination.  If nothing is connected to the "phase switch" port (port 7 of the Bot)... then it will default to Phase 2 operation... and therefore the "personality" that is programmed there.  Use a two position switch, and it will use phases 1 and 3.  Use a three position switch, and it will open up all three phase options...

There is a "copy" function of the software devoted to migrating settings from one phase to another, and there are "global" settings (affect all phases) and "local" settings (only affect a particular phase)... so one should be pretty sure of what they are messing with if the multi-phase approach is chosen (one phase could be totally flyable, and another not so much).

So... what am I going to do?  First off, I am going to use the gear switch (two-pole switch) to play with personalities... and I'm first going to program one personality with a high "I" gain on rudder (rather like heading hold), and one with a low "I" gain (rather like rate mode)... with all else being equal (after I figure out what *they* should be).  I'm going to use phase two as the "default" phase settings (since you don't really store any configurations on the PC)...  I fly both modes with discrete tail gyros... why not try it here?

[MALC1]

Very nice indeed 
It does give me some ideas...........hmmmm but which one 

[harry]

That's some nice wiring job

Got my SK360 in the air and will post some updates soon.

[leapfrog]

Okie doke...  software all done... Phases all programmed, and stuff all ready for maiden.  I'll build a table of the settings I end up with - after I finish the flight tests/tuning (as recommended by the manual).

The shortened main shaft gives the bird a more aggressive look...  kinda low/mean... *smile*.  More to come.

[J-Heli]

Okie doke...  software all done... Phases all programmed, and stuff all ready for maiden.  I'll build a table of the settings I end up with - after I finish the flight tests/tuning (as recommended by the manual).

The shortened main shaft gives the bird a more aggressive look...  kinda low/mean... *smile*.  More to come.

Can't wait to hear what you think! 

BTW, you could make a fortune selling those amazing airbrushed canopies of yours,  Canomod stands no chance!

[leapfrog]

*smile*

Hehe... the *really* funny part is that my 500's all have "custom" canopies that look like the older stock canopies...    I love my "dress whites"...  *grin*

[J-Heli]

*smile*

Hehe... the *really* funny part is that my 500's all have "custom" canopies that look like the older stock canopies...    I love my "dress whites"...  *grin*

!  That's awesome!

[leapfrog]

Weeeeeeeeee!

She maidened this evening after work...  and while I had to rush a bit (it gets dark sooner now), I did get a full bat through her (with several start/stop flights).  The first flight was "interesting" - not quite a "perfect" flight, by any means.  It seemed to be going pretty well at first, a little jumpy on it's own (it's weird to see the bird do things "on it's own")... and I got the sense the gains were a little high.

The manual suggests either starting with relatively low gains and working up, or matching the settings from one of their models and adjusting as necessary.  I used the "approximate" settings of one of their models of "similar stature"...  the Roxxter 33 SE (it has a 600mm blade span).  Apparently, the 600E is either lighter or has more control authority.  There are both "P" and "I" gains, as well as feedback for each axis to configure... and there are methods to adjust after flight tests.

The hovering was going well (just a bit of a jump once in a while), and the tail was rock solid... so I decided to check the feedback settings (what helps keep the bird on track through translation to FFF).  So I spun the tail around, put in forward elevator, and added collective for a "crisp" translation... and quickly found myself flying a "crack" bird...  hehe.  It was pretty funny.  I could arrest the oscillations by doing a high-G maneuver, but as soon as I relaxed the force on the bird she would just start up again... so I translated to a hover again, and she just sat there shaking... (it was like she was scared... *smile*).  So I took her up a bit and auto'd her to a landing... and she auto'd like a dream (well, like a 600), but she doesn't like *any* aileron/elevator input once skids heavy.  The manual does warn not to put *any* input while on the ground under power, and you do have to lift off rather quickly once head speed is achieved.

I adjusted the "I" gains down for aileron and elevator, and reduced the feedback (really easy with the Cockpit interface - right there at the field), and spun her up again...  She was "better"...  but I got some more "crack" in FFF.  She did seem to calm down back in a hover, though...  so for the third flight I reduced both the "P" and "I" gains a bit more... (only on the aileron/elevator axis... the tail was fine from the jump - maybe even a bit slow compared to my other 600's...  but I'll get to that later).  I think I went too far on that one - while there was no "crack" at all, the bird seemed a bit lethargic... and I had to constantly fly her (doing my own corrections)...  but it got too dark for me to continue.

Oh, BTW... I had to let the gyros in the unit (they are MEMS parts) acclimate before flying because it was around 40°F (~4°C) with a little breeze... *sigh*  Cold weather is coming...  but the nice thing was that there was *no* bug splat to clean up afterwards, and the performance was *wonderful*.

More to come.

[leapfrog]

WoooooHoooooo!

Two more bats, and I am narrowing in on the "perfect" configuration.  The swash is pretty set for now. She "feels" like a flybarred bird... and I really strayed from some of the manual's recommendations - like expo... it said to use a *lot*, but it felt like I was playing a video game with those long stick throws...  so I all but removed it, and she "came to life".  I do have reduced "P" and "I" gains - just slightly higher than the "starting point" suggestions... and I will play with them a bit more later.

I had to come back on the feedback settings a little to the aileron/elevator, which actually made the bird "better" than her flybarred siblings... she'll lock in at an attitude in FFF, and just fly thumbs-off as long as you let her go... (once again, it's a little weird to see her "self-correct" for gust upsets...  but she does).  My first instinct was to correct her attitude myself... and I ended up "fighting" the bird more than necessary... if you chill out a bit and "let" her fly, she'll take care of the little upsets on her own.  It's an awful lot like flying with AP on...

Once I got that, I started on the tail gains/settings... I kept getting a little "wag" under load, so I reduced the "P" gain a little and "I" gain a bit more than that... dang... that is one *snappy* tail with no bounce-back.  I'd say it is easily on par with all of the "bigger" tail gyros I've played with for the 600.  The lethargic nature was the expo I had in...  and once I got rid of that, she piros like a maniac and stops on a dime.  I can easily do full-speed sideways and backwards flight...  and it will not even flinch... not even a little bit.  You can certainly "hear" it working (interesting tail screams)...  but you can't see it.  It is just simply "solid".

Now, I need to work on the "rate" mode configuration (remember, I'm one of those old guys that also flies in rate mode)... that is basically removing the "I" gain from the rudder gains... and while I copied the settings between the two "phases" and only removed the "I" gain on the rate mode configuration...  I noticed a "bump" when I switched... (it may be due to the fact that the entire system settings are modified between "phases"... and even though the settings are the same - other than the "I" gain on the rudder... it may have to reload all of the summing biases and basically "start over" in mid flight).

More later... it got dark on me again.  But it was nicer... 58°F (~14°C), and about a 5-10mph (8-20kph) breeze, and pretty steady... and CAVU!