GyroBot 700

[ErichF]

Hey everyone! Well, as some of you may have read, I am playing with a flybarless setup on my Caliber 6 50-size helicopter. I'm using the Outrage 50 universal flybarless rotor head kit, after reading about the success that user Heli-Cuzz was having with it. He has been flying his setup on a Caliber 5 with NO electronic stabilizer system, so I decided I would see what all the fuss was.



Installation of the head was pretty straight forward, except I had to fabricate my own pitch links. I used the links included with the head kit by cutting them in half and grinding barbs into each piece. I then measured and cut to length two carbon rods. The linkage ends were then epoxied into the ends of the carbon rod. As you can see by the photos, they came out just right!

The next day, I took the Caliber 6 out for some test flying without any stabilizing system. I left my pitch curves the same, but reduced my cyclic rates about 20% and set expo to about 33% to start with. On the first flight, the spool up was fine and the tracking ended up just perfect. Takeoff was OK, but I had to really fight a nose down tendency. Now, this bird was perfectly trimmed with the flybar head, but I did have the canopy off now. There was a slight 5mph wind on the nose, too. I then started to fly the bird around gently with moderate forward speed. I didn't notice any pitchiness, but there was now a good nose up tendency. Holding slight forward cyclic pressure held the nose down for Forward Flight wasn't too bad.

I took the bird upstairs and did some rolls. Wow, slower than I expected. I came back down and reset my cyclic rates back to where they were with the flybar head. After that, she really livened up. Roll was crisp and fast, and so was the flip rate. In fact, she back flipped within the diameter of the rotor disc faster than I thought could be possible. I thought the tail was about to come off. I flew around and did some more flips, rolls, and other steady state maneuvering to test the stability of the rotorhead. The only issues I had to work with were the constant tendency to drift around in hover, depending on which side the wind hit the airframe and the constant nose up during FFF. It was obvious to me that this setup would never trim out in a hover, and general flying around would take a little bit of active piloting. I could easily fly scale maneuvers and not have any issues, though.

It was also obvious to me that close-in hard 3D would be easily done with this setup. Bigger, technical 3D like sustained hurricanes and such would require some extra work. Big air sport flying maneuvers definitely took some proactive attention to keep straight. Essentially, the smaller, more radical you fly the better. Flips, tic-tocs, piro-flips, and other stationary maneuvers should be flown just fine.

So, after reporting my results about flying without a stabilizer, Danny Melnik at Esprit Models contacted me about comparing my results after using a GyroBot 700. I agreed, and he sent me the system the next day.

As you can see in the photos, this is a pretty compact system. Just one box about 1 .25" square contains the whole system. A nice, simple programming box is also included, as are all the servo lead jumpers and sticky-back foam tape.

The installation on the Caliber 6 was super easy. I positioned the gyro box on top of the radio enclosure with two layers of tape, as suggested in the manual. After all connections were made, I reset my radio for a new model using the standard "H-1" mechanical mixing mode.

The setup of the Gyrobot is pretty straight forward. There is one button on the LCD box that is used to accept values or menu headings and values are changed by moving the rudder stick on the radio left or right. So, obviously you need to have your radio powered up and running during any adjustments to the system. The manual guides you in Germanglish dialect through the entire process in proper order of operation. There wasn't any point at which either didn't make sense or made me feel lost. It's nice not to have to lug my laptop over to the bench; however there is that option with the USB cable. It really offers no advantage to setup and adjustment, though.

The installation and base setup process was completed in about an hour. I set all the Proportional (P) and integral (I) gains per the starting settings in the manual. I left my ATVs and Expos around 55% and 20% respectively. This system uses the radio end points or ATVs to set rotation rate for aileron and elevator, similar to how tail gyros do for pirouette rate. This is nice, since it doesn't require you to plug into the system to make rate changes...you just do it in your radio.

The first couple flights were rather mundane. I had flown without the system at all, so I wasn't worried if the system wasn't going to do much at the low gain presets. I was surprised at how slow my cyclic rates were, though. I landed and increased all my cyclic rates to about 70% and 18% expo. This felt just like my flybar head. However, the gains were not near enough to properly hold the attitude like I expected. So, several tuning adjustments were made, each increasing the P and I gains about 50 units at a time for both cyclic channels. I also had no issues getting the tail gyro to hold well with no overshoot. This system is easy to adjust at the flightline. You don't even have to kill the engine. Just plug in, enter program mode, make your changes, exit program mode, and disconnect. Then back to flying!

Another nice feature for using when making radical adjustments is that the system has two independent "phases." You can switch between phases in flight with a spare channel. You make the gross adjustment in say, phase two. You spool up and take off in phase one, then once stable in hover at altitude, you switch to phase two and test your adjustment. If it goes nutty or oscillates, switch back to the other phase and land for further adjustment.

Anyways, after about four flights of tuning, the system was pretty much solid. We turned up the control rates to around 80% and lowered the expo. Also, we tuned the tail a bit to eliminate high speed oscillation in FFF. The result is what you see in the video:

GyroBot 700 Demo Video

GyroBot 700 Demo Video

I will continue to tune the system and try different settings to explore the capabilities of the system. I am still fine-tuning the gains to get the bird to track straight and level during FFF. It's tough to totally eliminate any nose rise over time without the gains getting critically close to the oscillation point. I also found that, just like a tail gyro, you can't trim your heli's cyclic from the radio anymore. You can trim it out for that flight, but once you power off and recycle the system, it will be out of trim again. The key to trimming the model with this setup is proper CG, followed by swash plate centering adjustment. If your model constantly drifts backwards in a hover, go into the swashplate base menu and add a couple points of forward swashplate trim. This takes more trial and error, but works out very well in the end for near hands off hover upright and inverted.

Feel free to post your questions or comments here, or just PM me. I will have this helicopter and the system at the Jacksonville funfly, if you would like to test fly it.

Cheers for now,

[TomC]

Thanks for this very useful review Erich. What Ail/Elv P and I values are you currently using? Thanks.

Cheers,
TomC

[ErichF]

Thanks, Tom:

I will post my gains with the disclaimer that they will really mean nothing on your particular setup, unless it's a Caliber 6 with the same servos, servo wheels, head, and blades. According to Danny at Esprit, my gains ended up a bit higher than usual...which is a good thing I guess.

Rudder P = 450
Rudder I = 400
Elevator P = 500
Elevator I = 600
Aileron P = 520
Aileron I = 600
Feedback for both = 20
Piro Optimization = 0000

I think the key to a high gain setup is using the smallest servo wheels that still get you the full range of motion, accurate servos, zero slop linkages and bellcranks, and long pitch arms on the grips. My servo wheels are only 12.5mm from center. Most guys use 15-20mm.

[Heli-RND]

This definitely comprehensive, thank you so much

[ErichF]

When setting this system up, you may notice a new set of terms that you may be unfamiliar with...P and I gains...

I have looked up some "PID Loops for Dummies" type of resources for you to look over. Maybe after reading these you can "gain" a better understanding of what the Proportional and Integral gains do, what to expect from adjusting each, and how much to adjust them.

http://en.wikipedia.org/wiki/PID_controller

http://www.robotc.net/teachingmindst...odindex.htm#13

http://zone.ni.com/devzone/cda/tut/p/id/3782

Enjoy!

[F1 Rocket]

Great stuff Erich. The only thing that really looks odd with your settings is the feedback value of 20. If you are still seeing some ballooning in FFF I would reduce this value. I have found that high feedback feels better in the hover but lower values work better upstairs.


Danny

[ErichF]

Actually, after setting to 20, the nose-rise in FFF is almost gone to the point it's not an issue anymore. It was at 25.

Despite the high gains, this thing is more nimble than anything I have flown yet. I will have another video with the latest tweaks and rates set after tomorrow.

[F1 Rocket]

Must be the servo/linkage geometry in the Caliber 6. Not a bad thing, just different then what I'm use to. It will be interesting to see what setting the TDR ends up with. One of the things I like most about the GB 700 is how tunable it is.

Danny

[donwag]

Erich,
I am new to the FBL setup and only been flying for two years.(novice)

Danny sold me the GyroBot 700 and after reviewing the documentation and setup it looks really easy. I was one of the lucky winners at IRCHA and won a FBL head kit. One of the questions that I have is the comment about the radio setup. Starting a new model in the radio? Do we still make sure that any servo reversing is done?

I don't understand the "standard H-1" settings. I am using a JR9303. Do we still setup the radio as a 120 degree swash? I think I get the P and I thing going on as you explained(thank you). I am going to put this in a Align 600e. Any help you can provide would be helpful. Maybe Danny can shed some light on this as well.

Thanks,
Don

[Heli-RND]

Originally Posted by donwag

Erich,
...One of the questions that I have is the comment about the radio setup. Starting a new model in the radio? Do we still make sure that any servo reversing is done?

There is nothing you need to set in your radio, you should leave the servo reverse to the GB 700 where you set it there instead of the radio

Originally Posted by donwag

I don't understand the "standard H-1" settings. I am using a JR9303. Do we still setup the radio as a 120 degree swash?

H-1 is the heli mode without any CCPM mixing, you must again, leave it H-1 in the radio and set the GB to 120 degree swashplate setting, meaning, leaving the swash mixing job to the BG.