RC model yaw control B-2 Spirit stealth bomber

[Totalsolutions]

Hi everyone, My friend Nick is an avid radio control plane designer using his own CNC machine. Nick has asked me if I know anyone to help with completing a problem.


Quote from Nick
"The problem:
My B2 model has no fin or rudder so it is unstable in Yaw (direction left and right) but stable in pitch (Up and Down) and Roll (banking to the left or right)
In order to keep it stable in Yaw it has to be fitted with a digital (MEMS) gyroscope.
This senses any deviation from a straight ahead flight path.
The control surfaces on the model are the same as on the full sized aircraft. Ailerons control roll. Elevators control pitch. But the yaw is another matter because there is no rudder and fin on the aircraft. Control of yaw is therefore achieved by increasing the drag at one wing tip and this literally drags the aircraft around in that direction. The drag is created by a pair of "Clam shell" like surfaces at each wing tip (They're called drag rudders but I'll call them clam shells here). So if you wanted to turn right the right hand pair of clam shells would open up causing drag in that direction. However the clam shells on the other wing tip would remain closed.
The actuation of the clam shells is via the rudder stick just like on your helicopter transmitter. The receiver picks up the command and passes it through the gyro so that the gyro knows that a signal is coming from the pilot and knows not to neutralise it as it sees the aircraft veering off a straight ahead path.
The problem is that when the gyro is left to do it's job it opens the clam shells on one wing tip but also tries to close the clam shells on the other wing tip. As the clam shells can not be moved past closed any signal to close them stalls the motor in the servo and burns it out. The servos, as you probably know are basically a motor, a gearbox to multiply the torque and a feedback potentiometer on the output spindle. The radio receiver is always looking at the resistance feedback from the pot' and referring to the position of the sticks on the transmitter which essentially tell the servos via the receiver to move to a place where the resistance is a certain value.
This is what I need:
I need an interface circuit to fit between the gyro outputs (it has two outputs - one for each servo on the clam shell pairs) and the two servos.
The interface would recognise the position (resistance feedback) from the servo potentiometer when it is at rest (the closed clam shell position). It would also recognise the resistance feedback when the clam shell servo was at it's fully open position. But the interface would have to prevent any transmission of a signal (some kind of blocking device) that would take the servo outside the maximum and minimum resistance levels that it somehow has to recognise and store."

Sorry its a bit long but hopefully someone here knows what to do !

Thanks for reading.

[IO390]

There's a couple ways of doing this. If you want to make life easy, though, just mix it on your transmitter. If your R/C gear can't do mixing or anything like that, then you ought to upgrade.

RC Powers did a lot of tailless stuff with cheapo gyros. Look up their videos, it might help.

Edit: Are you sure this thing is even gonna fly? It looks way too small to be using scale sized intakes and exhausts. I've been flying models for the past 10 years, and even on large jets cheater holes are mandatory when scale intakes are used.

[David Hess]

Analog and digital limit circuits are pretty straightforward but details about the signaling are needed to make a recommendation.  I am used to RC servos which accept pulse width modulation and have local feedback to control position.

I guess you cannot just turn the servo to limit its travel in one direction?

[eneuro]

In order to keep it stable in Yaw it has to be fitted with a digital (MEMS) gyroscope.

Which gyro is used?
I played with this MPU6050 gyro & accelerometer sensor, but not in RC so far, but maybe at winter will be more time to make this great B-2 Spirit stealth bomber 

It is possible to connect additional I2c digital compass to MPU6050 and maybe it is a must in RC, while sensor fusion is needed and this chip can do it in hadrware, but it was a little bit tricky when played with this thing a few months ago.
However there is another one gyro with accelerometer and compass all in one, as well as open source tools to use it for flight controll.
Without compass such things can happen and this digital gyro is not as stable as one could expect 

More details and forum there:
 http://www.i2cdevlib.com/devices/mpu6050#source

[Nick Butler]

Yes, the model flies very well. Scale intakes and exhaust using 50mm EDF units. It will only fly with twin gyros on the drag rudders and the ones we have tried so far are unreliable hence the request for help.
Nick.
www.nick-butler.co.uk

[IO390]

Yes, the model flies very well. Scale intakes and exhaust using 50mm EDF units. It will only fly with twin gyros on the drag rudders and the ones we have tried so far are unreliable hence the request for help.
Nick.
www.nick-butler.co.uk

Could you perhaps use a single gyro with a servo reversing lead for the second rudder? Plug and play: http://hobbyking.com/hobbyking/store/__31875__Turnigy_Servo_Signal_Reverser_UK_Warehouse_.html

I'd recommend these gyros. Cheap as chips but I've not been able to notice ANY difference between this and a genuine Futaba GY401 on my 50-size heli:
http://hobbyking.com/hobbyking/store/__26075__Hobby_King_401B_AVCS_Digital_Head_Lock_Gyro_UK_Warehouse_.html

Also, your SR-71 looks incredible. Do you have any videos of it flying?

[eneuro]

I'd recommend these gyros.

Is it possible to detect "free fall" using this Hobby King 401B AVCS Digital Head Lock Gyro units?
I'd like to activate parachute for safety reasons (do not let fall this B-2 from the blue sky like a drone  ) and radio/buzzer beacon in the case of serious RC power or system failure when there is no pings from radio control and MPU6050 accelerometers detects free fall..

BTW: I've got paraglider license and without parachute it is very taff to fly high like the Ikar, but did not saw any RC models with this option 

[max_torque]

Most servo's run with a basic PWM signal to set there target position.  If this signal is reducing too much you either need to intercept that signal, and put in some logic that relays the signal but holds it at a minimum value to prevent the servo from travelling too far.

However, a much easier option might just be to use a mechanical link on the servo that has "lost travel" in the correct direction.  This will effectively allow the servo to continue to move to its internal end stops, but use a sping or similar to prevent that motion from forcing the clams shut further than they are

[kc9qvl]

The Drag/Split Tips on that B2 design like that  are meant to be always open a little bit. The plane should fly fine without a gyro.  If scale appearance is not super important add drag wedges.



A reference thread on rcgroups.:
http://www.rcgroups.com/forums/showthread.php?t=747046

[Nick Butler]

My SR71 Blackbird was the first own design fully moulded EDF jet which I built a few years back. This is the video showing some early events with this exciting project. Since then I have replaced the motors with Lehner motors and experimented with 5S LiPo batteries but reverted to a 4S set-up. Also the 'plane is now bungee launched from a wheeled dolly.
Let me know if you would like to buy one (I have two spare airframes) or if you need further details.
Nick.

[FreddyVictor]

The problem is that when the gyro is left to do it's job it opens the clam shells on one wing tip but also tries to close the clam shells on the other wing tip. As the clam shells can not be moved past closed any signal to close them stalls the motor in the servo and burns it out. The servos, as you probably know are basically a motor, a gearbox to multiply the torque and a feedback potentiometer on the output spindle. The radio receiver is always looking at the resistance feedback from the pot' and referring to the position of the sticks on the transmitter which essentially tell the servos via the receiver to move to a place where the resistance is a certain value.

just trying to understand the problem a little better and maybe offer an alternative ..

is it not simply that you want to ignore the PWM signals if they go beyond a certain amount ?
Getting a bit techy here but...
the Yaw signal will usually be PWM signal of ~1500us for  straight-ahead (range is roughly 1000us to 2000us), so, on a particular side say, pass thru signals >= 1500us but if any less keep sending 1500us (or only pass thru <= 1500us)

[kc9qvl]

Something else to try. Move the horns and linkages on the servo so that when the spit tips are closed the servo is already fully over. They can't possibly strip that way. The tips will still be open a bit though in flight. They have to be to counteract the roll. Even on a real full sized b2 they are open.

[sacherjj]

The issue is that an RC gyro works by hooking to the "rudder" output of the receiver and mixing with the rudder pulse length, based on yaw input.  So you need to feed two servos from a single yaw output.

Seems like the easiest would be to split the gyro output drive to two servos.  If you mount the servos as a mirror image, you will get "v-tail" like setup where one pushes when the other pulls.

Now our issue is that we only want them to open with half travel, and extended travel the other way stalls the servo.  The physical solution is to have a push linkage with a lightly spring pull section for over travel.  That is how we did it, back when we had channel numbers on our antenna. 

The electrical solution would be to put a micro inline and set it up to limit timing pulse maximum or limit timing pulse minimum, depending on which side the servo is jamming.  It would receive input from the gyro output split to both servos by toggling two pins.  You could also include things line "position at center" and some type of curve mapping for actual extension vs servo signal.

If you got a micro with ADC channels, you could just put some trim pots for trimming settings.  Per servo, you could have pots for Minimum Position at Center, Minimum Pulse Length and Exponential of Output. 

To control max throw, you should just be able to limit throw in the transmitter or linkages.

I would also use a Gyro with as much tuning as possible, as most are made for specific uses, such as helicopter tail rotor.
 

[Skimask]

Now our issue is that we only want them to open with half travel, and extended travel the other way stalls the servo.  The physical solution is to have a push linkage with a lightly spring pull section for over travel.  That is how we did it, back when we had channel numbers on our antenna. 

That's what I was going to say, something similar to spring loaded deals on a nose-wheel steering setup.
Either that or a hard control rod with a lock bushing on it...you can pull the arm because the lock bushing butts up against the servo arm, but when you push it, the rod just slides thru the servo arm.
But if it was me, I'd whip up something with a PIC that reads the servo PWM and acts accordingly, as was described above.  Lets the signals thru that are above 1500us.  "Clamps" the PWM signals that are below 1500us (or the other way around as needed).

[hamster_nz]

You could take a standard DIY servo mixer circuit (e.g. http://ilufa.orgfree.com/MIXER_E.htm) and rework the ASM (http://ilufa.orgfree.com/MIXER508.ASM) to either split a single rudder signal into left and right servos of the drag rudders,  or maybe take two channels in and then only limit and scale channels as required (using mixing in the TX to mirror the rudder channel)

It would still need careful setup to ensure that your servos are not under a constant load when the drag rudders are closed.


 

[David Hess]

The electrical solution would be to put a micro inline and set it up to limit timing pulse maximum or limit timing pulse minimum, depending on which side the servo is jamming.  It would receive input from the gyro output split to both servos by toggling two pins.  You could also include things line "position at center" and some type of curve mapping for actual extension vs servo signal.

This is how I would do it with the older style servos which use a pulse width modulated control signal.  Digital logic and microcontrollers are good for working with pulse width modulation although it is possible to do this completely in the analog domain as well.

[eneuro]

Digital logic and microcontrollers are good for working with pulse width modulation although it is possible to do this completely in the analog domain as well.

Maybe, but oryginal B2 has hundreds of MCUs 

I plan put into my 3m wing span B2 RC model tens of AVRs in pararell  providing support via custom I2C for  GPS, gyro & accelerometer & digital compas with sensors fusion library hardware accelerated in MPU6050

I found nice ultrasound sensors, now looking for wind speed sensors based on measurements of air pressure   

[sacherjj]

Digital logic and microcontrollers are good for working with pulse width modulation although it is possible to do this completely in the analog domain as well.

Maybe, but oryginal B2 has hundreds of MCUs 

I plan put into my 3m wing span B2 RC model tens of AVRs in pararell  providing support via custom I2C for  GPS, gyro & accelerometer & digital compas with sensors fusion library hardware accelerated in MPU6050

I found nice ultrasound sensors, now looking for wind speed sensors based on measurements of air pressure   

You sound like you have once held a government job. 

[eneuro]

You sound like you have once held a government job. 

It is not another military project like this Predator Drone Reportedly Spotted Over ISIS Stronghold, but I'd like this B2 be a little bit bigger than Golden Eagle 
There a few nice big hunters close to our yard, so watching them flying and reusing vertical hot air movements require bigger wings area  I guess, so bigger means better eficiency