3DR and PX4 Announces Pixhawk. Next generation flight controller.

[R_Lefebvre]

3DR have just announced what will be the successor to the Ardupilot Mega platform. We are completely abandoning all ties to Arduino, and forging ahead with a state of the art processor running a real-time operating system. Excellent software engineering has allowed the 8-bit AVR2560 processor in the APM to deliver stabilization and autopilot performance that rivals the most expensive systems in the hobby and pro-sumer market, but we have taken it as far as possible as the processor resources are completely maxed out. The 32-bit STM32F4 processor in the PX4 system offers well more than 10 times the power of the AVR2560, and will open up a lot more capability.

Flight stabilization, RTH, auto-takeoff and landing, unlimited* waypoint flight, Follow-Me and geofencing is already functioning. Along with extras like gimbal control, navigation lighting control and telemetry to PC or Android device, it already delivers a complete autopilot system. But the Pixhawk allows room for growth, with proposed features such as on-board terrain data for intelligent RTH, optical flow navigation enhancement, Lidar and automatic collision avoidance, and mini-computer input for really high level functionality (vision systems?). For helicopters I'd like to add governor control, Piro-Comp and Acceleration Cross-Coupling, servo linearization, and multiple flight-conditions with full gain-scheduling.

I'm really happy with the performance of the system now, and can't wait to see all the neat things coming in the future.

Here is the press release:

Almost exactly one year after the first PX4 announcement, we would like to introduce our newest member of the family, Pixhawk! For those familiar with the existing PX4 electronics, it is the all-in-one board combining PX4FMU + PX4IO, combined with a processor and sensor update and a number of new features. The current board revisions will however remain in full service and active development and are fully compatible. Pixhawk is designed for improved ease of use and reliability while offering unprecedented safety features compared to existing solutions.

Pixhawk is designed by the PX4 open hardware project and manufactured by 3D Robotics. It features the latest processor and sensor technology from ST Microelectronics which delivers incredible performance and reliability at low price points.

The flexible PX4 middleware running on the NuttX Real-Time Operating System brings multithreading and the convenience of a Unix / Linux like programming environment to the open source autopilot domain, while the custom PX4 driver layer ensures tight timing. These facilities and additional headroom on RAM and flash will allow Pixhawk the addition of completely new functionalities like programmatic scripting of autopilot operations.

The PX4 project offers its own complete flight control stack, and projects such as APM:Copter and APMlane have ported their software to run as flight control applications. This allows existing APM users to seamlessly transition to the new Pixhawk hardware and lowers the barriers to entry for new users to participate in the exciting world of autonomous vehicles.

The flagship Pixhawk module will be accompanied by new peripheral options, including a digital airspeed sensor, support for an external multi-color LED indicator and an external magnetometer. All peripherals are automatically detected and configured.

Features:
-32 bit ARM Cortex M4 Processor running NuttX RTOS
-14 PWM / Servo outputs (8 with failsafe and manual override, 6 auxiliary,
high-power compatible)
-Abundant connectivity options for additional peripherals (UART, I2C, CAN)
-Integrated backup system for in-flight recovery and manual override with
dedicated processor and stand-alone power supply
-Backup system integrates mixing, providing consistent autopilot and manual
override mixing modes
-Redundant power supply inputs and automatic failover
-External safety switch
-Multicolor LED main visual indicator
-High-power, multi-tone piezo audio indicator
microSD card for long-time high-rate logging

Specifications:
-32bit STM32F427 Cortex M4 core with FPU
-168 MHz
-256 KB RAM
-2 MB Flash
-32 bit STM32F103 failsafe co-processor
-ST Micro L3GD20H 16 bit gyroscope
-ST Micro LSM303D 14 bit accelerometer / magnetometer
-MEAS MS5611 barometer

Interfaces:
-5x UART (serial ports), one high-power capable, 2x with HW flow control
-2xCAN
-Spektrum DSM / DSM2 / DSM-X® Satellite compatible input
-Futaba S.BUS® compatible input and output
-PPM sum signal
-RSSI (PWM or voltage) input
-I2C®
-SPI
-3.3 and 6.6V ADC inputs
-External microUSB port

Power System and Protection:
-Ideal diode controller with automatic failover
-Servo rail high-power (up to 10V) and high-current ready (10A +)
-All peripheral outputs over-current protected, all inputs ESD protected
-Monitoring of system and servo rails, over current status monitoring of peripherals

Dimensions:
-Weight: 38g (1.31oz)
-Width: 50mm (1.96")
-Thickness: 15.5mm (.613")
-Length: 81.5mm (3.21")

Availability:
This announcement is a service to our users and developers to allow them to plan their hardware roadmaps in time, and to show what we're currently working on. The board will not be immediately available, but 3D Robotics is taking pre-orders for Pixhawk now, and will begin shipping in late October. The price is $199.99.

[onepiece]

Rob,

I'm going to expose my ignorance for ALL to ridicule, deride, desparage and scorn. That said, I'd still like your comment.

The press release you've posted is LOADED with hot specs and names sensors out the wazooo. But, exactly what are the features of the unit? What will this baby do that other stabs won't?

For example, if I had to describe what a BD-3SX does for the helicopter flying experience... I would say:

- Provides all "flybarless" electronic control functions including rear rotor gyro duties.
- Provides a rescue or 'bail out' features to avoid a crash in dangerous or near crash manuvers
- Provides user variable assistance to maintain a stable airship attitude inspite of poor pilot input or windy conditions.
- Does NOT provide motor speed governer function
- Does NOT incorporate any GPS navigation/hold in position or Return To Launch nor failsafe functions if a Tx signal is lost.

Can you make a very descriptive list of functions this unit is likely to do?

Thank you,

OP
.

[R_Lefebvre]

Hmmm... OK, I will try, though it's tricky. First of all... it must be understood that the Pixhawk in the immediate future will run the Arducopter program. The exact same program that is running now on APM2.x, and you would expect no difference in performance. That is because the hardware is so new, that the program has not even been developed for it yet. I don't even have a unit to play with yet.

The other thing that needs to be understood, is that the program is open source, and continuing development. It's never "done". Anybody can come along and add a feature. Or sponsor/fund development for somebody to add a feature.

Anyway, so here's what I would say is working now:

- Works on Flybar or Flybarless
- Acrobatic mode that is as least good enough for a learner, maybe not IRCHA level though...
- Stabilization mode that allows hands-off hovering (drifts in the wind). I have flown a 450 heli at 400 yards in this mode (think about it).
- Stabilization mode can be used as a non-automatic bail-out. Just flick the mode switch from Acro to Stab, and it will roll to level pretty quick.
- Stabilization is an excellent mode for sport flying, scale, or learning to hover with much less stress and pilot load than a FBL controller.
- Stabilization is augmented with GPS data fusion into the IMU. Systems without GPS input can never fly as well at high speeds.
- Altitude Hold mode that stabilizes the altitude, but otherwise flies like Stab mode. This is sort of like putting a heading-lock gyro on your collective. It resists altitude changes very strongly, you just push the collective up and down for a pre-defined climb/descent rate.
- Loiter Mode, which is the same as GPS position hold mode on other systems. Hands off hover, and remaining in position fighting wind automatically. If you move the cyclic stick, it commands the heli to fly at a predetermined speed, precisely in the direction it is pushed.
- Simple Mode: accessed in any mode. Removes need to worry about orientation. If you "arm" the heli with the tail in, then no matter which direction the heli is facing, "up cyclic" will push the heli away from you.
- RTL. Flick a switch, it climbs to a predetermined altitude, and flies home, and then descends to a predetermined altitude. Can be programmed to remain at that altitude indefinitely, or to land and shut off the motor.
- Waypoint flight. Full programmable waypoint flying. Stop and hover at waypoints, or fly right through. Change altitudes. Take pictures. Almost endless options. Waypoint number is limited only by available memory, not a marketing department.
- Waypoint flight speeds up to 100 km/h tested. Higher than other systems, even professional UAV controllers.
- Circle Mode, automatically orbits a point, yawing nose-in. Great for photography.
- Radio Failsafe. Programmable, can RTL if radio link is lost.
- External lighting control. Control LED lights for vision aid, as well as status monitoring (flashes when battery is low, etc).
- Telemetry. Radio link is only $80. FULL telemetry, 100 pieces of data. Battery, altitude, ground speed, # of GPS Satellites... almost anything you could think of.
- PC based "Mission Planner" is the main program. Also 2 available Android Apps, work on phone or tablet. Apple app reportedly coming.
- Camera gimbal control. Stabilize servo gimbals, or interface with Brushless Gimbals. Can include automatic aiming at GPS targets.

Here are some things I want to work on in the future, these will likely only work on Pixhawk and not APM.

- Piro Comp. Exactly what you think. Can't run on APM2.x, will run on Pixhawk.
- Cross-Axis Coupling Compensation. Dynamically solves for the variable cyclic phase errors that occur in every heli. This is the common cause of "head bobbing" when running low head speeds. Again, Pixhawk only.
- Internal Governor. Control head speed by RPM setpoint instead of mysterious percentages or PWM values. RPM and Load reporting via telemetry. Error warnings.
- Follow-me Mode. Heli can follow ground target and automatically shoot video.
- 4 Servo Swash plates.

Long-term stuff:

- Automatic Auto-Rotation on head speed failure. Maybe not a smooth landing, but at least a softened crash. This is actually a big deal for UAV applications.
- Tandem Rotor Heli

[onepiece]

Rob – this is EXACTLY the kind of description non-engineers like me need to hear! Thank you!

My first thoughts? HOLY SMOKES – somebody designed this from my dreams…
My second thoughts? This is MUCH too good to be true.

Hope I’m 100% wrong on my second thoughts! I’ve dissected your answers…and have asked for clarifications on a few items. Thanks in advance for your truly helpful answers.

Couple more comments first:

- The size! The length is HUGE. Hope retailers will sell a mounting plate that fits between the skids….as that’s the only place I can see something this big fitting.

- Will set up procedures be written for Doctoral candidates in Computer Science – or can you make set up protocols, bench tests and procedures understandable for mere mortals with borderline dyslexia like me?? No Chinglish…right? Can the marketing folks release ‘set up’ videos at the time of release…or do we have to do ‘crowd testing and debugging’ like MS products?

- How many radio channels do you need for the features you mention? For example, does the governor function need a dedicated channel?

- Works on Flybar or Flybarless

Yahoo! You didn’t forget OLD SCHOOL! The FLB is such a classic look and heritage, it will never completely disappear!

- Acrobatic mode that is as least good enough for a learner, maybe not IRCHA level though...

How is flight stabilization different when in Acrobatic mode? What would happen when I flicked the switch to Acrobatic mode? In what way is flight & stick control affected?

- Stabilization mode that allows hands-off hovering (drifts in the wind). I have flown a 450 heli at 400 yards in this mode (think about it).

Will the stabilization mode hold altitude at one level as well as stop drifting on the horizontal plane? Is this done by the barometer, the GPS sensor, optically, a combination or none of the above?

- Stabilization mode can be used as a non-automatic bail-out. Just flick the mode switch from Acro to Stab, and it will roll to level pretty quick.

Here’s the $64M question on bail out… If my heli HAS momentum and is headed into a tree 40 feet away at high speed and oriented such that ‘rotor disk is PERPENDICULAR with the ground ’… will the bail-out feature change the orientation so that the heli becomes level, PARALLEL with the ground and still headed into the tree at high speed? OR… will it shoot skyward and level itself avoiding the sideways motion and subsequent crash into the tree altogether?

- Stabilization is augmented with GPS data fusion into the IMU. Systems without GPS input can never fly as well at high speeds.

Is the IMU part of the main unit in the above photograph or is the IMU an add on device? Whew…if its an add on…where do I put it? How about the GPS sensor….is that mounted remotely from the main unit?

- Altitude Hold mode that stabilizes the altitude, but otherwise flies like Stab mode. This is sort of like putting a heading-lock gyro on your collective. It resists altitude changes very strongly, you just push the collective up and down for a pre-defined climb/descent rate.

Is this altitude HOLD feat done by a barometer, the GPS sensor, optically, a combination or none of the above?

- Loiter Mode, which is the same as GPS position hold mode on other systems. Hands off hover, and remaining in position fighting wind automatically. If you move the cyclic stick, it commands the heli to fly at a predetermined speed, precisely in the direction it is pushed.

Is this Loiter Mode feat done by a barometer, the GPS sensor, optically, a combination or none of the above?

- Simple Mode: accessed in any mode. Removes need to worry about orientation. If you "arm" the heli with the tail in, then no matter which direction the heli is facing, "up cyclic" will push the heli away from you.

By ‘up cyclic’ do you mean push the cyclic stick forward? As in using the elevator for forward and upward flight?

- RTL. Flick a switch, it climbs to a predetermined altitude, and flies home, and then descends to a predetermined altitude. Can be programmed to remain at that altitude indefinitely, or to land and shut off the motor.

Sign me up NOW!!

- Waypoint flight. Full programmable waypoint flying. Stop and hover at waypoints, or fly right through. Change altitudes. Take pictures. Almost endless options. Waypoint number is limited only by available memory, not a marketing department.
Yahoo!
Here are some things I want to work on in the future, these will likely only work on Pixhawk and not APM.
- Piro Comp. Exactly what you think. Can't run on APM2.x, will run on Pixhawk.

Better explain this in simple English….not sure what you mean.

- Cross-Axis Coupling Compensation. Dynamically solves for the variable cyclic phase errors that occur in every heli. This is the common cause of "head bobbing" when running low head speeds. Again, Pixhawk only.

Better explain this in simple English….not sure what you mean. Describe how the flight is changed or affected.

- Internal Governor. Control head speed by RPM setpoint instead of mysterious percentages or PWM values. RPM and Load reporting via telemetry. Error warnings.

Do you need a separate channel for governor?

- Follow-me Mode. Heli can follow ground target and automatically shoot video.

Is this like ‘facial recognition’ software? Can it stalk people? Seriously…how does it track and follow? Do you need to buy a camera for this function or is a camera built in??

[R_Lefebvre]

This is MUCH too good to be true.

Bah.

The size! The length is HUGE. Hope retailers will sell a mounting plate that fits between the skids….as that’s the only place I can see something this big fitting.

Are you basing that on actual dimensions, or just appearance? It's actually not that bad... Ok, it is larger than most systems. I made a little subframe for my 450 to hold it, and there's tons of room. The thing to know is this is a UAV controller, not a flybarless controller. It would have no problem fitting inside a 450 scale body, for example. And it is not hard to fit onto a 550+ frame with a little customization.

- Will set up procedures be written for Doctoral candidates in Computer Science – or can you make set up protocols, bench tests and procedures understandable for mere mortals with borderline dyslexia like me?? No Chinglish…right? Can the marketing folks release ‘set up’ videos at the time of release…or do we have to do ‘crowd testing and debugging’ like MS products?

I'll let you judge for yourself.

http://copter.ardupilot.com/

No Chinglish, that's for sure. The wiki pages for Helicopter are a little weak, hopefully I'll be able to upgrade them soon.

- How many radio channels do you need for the features you mention? For example, does the governor function need a dedicated channel?

5 channels is the bare minimum, but you really should be using 8. If you don't have an 8 channel radio in 2013, I think you should be looking hard at your choice in radio supplier. I'm surprised anybody even makes radios with less than 8 channels anymore. IMO, companies selling 6 channel radios in 2013 are indicating that they make junk.

Sorry for the strong opinion on that but uh.... channels are truly "free". Companies who sell 6 channel radios so that they can charge more for 8, are taking advantage of their customers.

But I digress.

Ch 8 is typically used for motor control in this case. The flight controller ultimately needs to know the status of the motor, and control it. It can't do it's job properly otherwise.

Oh, and this reminds me, I forgot to mention that the code also supports a second ESC output for direct drive tail motor.

Yahoo! You didn’t forget OLD SCHOOL! The FLB is such a classic look and heritage, it will never completely disappear!

I have to caution though, that it's been a while since I tested FB on this system. I assume it still works but... it really needs to be confirmed.

How is flight stabilization different when in Acrobatic mode? What would happen when I flicked the switch to Acrobatic mode? In what way is flight & stick control affected?

Basically, Acro mode is exactly like any other FBL controller. It's pure rate control. It is like a 3-axis heading lock gyro.

However, we also have some really innovative features here. We have what's called "Acro Trainer", which you can turn on and off. What Acro Trainer does, is that if you center the cyclic stick, it will automatically self-level, gently. The strength of the effect is programmable. This could help a new pilot transition from flying in Stabilize, to flying in Acro/Rate control.

Will the stabilization mode hold altitude at one level as well as stop drifting on the horizontal plane? Is this done by the barometer, the GPS sensor, optically, a combination or none of the above?

No. Stabilization mode is manual collective, and it does not stop drift. It just controls the attitude. So if you push and hold the cyclic full "up", then that signals 45° nose down. It will move to that angle, and then hold. That's how I flew a 450 at 400 yards. You simply push the stick forward, maybe 20°, and hold it, and the heli will fly at that attitude. Then just use collective to keep it off the ground.

However, Alt Hold mode does manage altitude, exactly as you want. And "Loiter" also stops drifting.

Here’s the $64M question on bail out… If my heli HAS momentum and is headed into a tree 40 feet away at high speed and oriented such that ‘rotor disk is PERPENDICULAR with the ground ’… will the bail-out feature change the orientation so that the heli becomes level, PARALLEL with the ground and still headed into the tree at high speed? OR… will it shoot skyward and level itself avoiding the sideways motion and subsequent crash into the tree altogether?

At this point, it's fairly dumb in this regard. If you flicked to Stab mode, it would immediately level, but the climb rate would depend on where you have the collective. If you push it up, it will shoot up. However, if you push it up BEFORE it rolled level, then it will shoot towards the tree.

Automatic tree avoidance would be tricky.

If you flicked into Loiter mode instead (Position Hold), it's possible that it would come to a nice stop and avoid the tree, but nobody that I know of has tried going to Loiter from an upset position like that.

Is the IMU part of the main unit in the above photograph or is the IMU an add on device? Whew…if its an add on…where do I put it? How about the GPS sensor….is that mounted remotely from the main unit?

The IMU is internal to that device. This is both good and bad. Good for simplicity. But it's bad for vibrations. You do have to pay attention to how much vibration the system is subjected to. But I have found this is easily managed.

The GPS sensor is external, it's a little puck, like 1.5" square.

Is this altitude HOLD feat done by a barometer, the GPS sensor, optically, a combination or none of the above?

Alt Hold is controlled by the fusion of Baro data and Z-axis acceleration. It manages altitude extremely well. Usually within a few inches, even in heavy winds.

Is this Loiter Mode feat done by a barometer, the GPS sensor, optically, a combination or none of the above?

Loiter Mode is done by fusion of GPS data and X/Y axis acceleration. At the same time, it uses the Altitude Hold function to hold altitude.

By ‘up cyclic’ do you mean push the cyclic stick forward? As in using the elevator for forward and upward flight?

Yes, that's what I mean. If the heli is facing East when you "arm it", and then you take off and put it in Simple Mode, if you rotate the heli, "up elevator" always leans it East. If you rotated it so the nose was facing North, then it would lean to the right, to fly East.

Better explain this in simple English….not sure what you mean.

Piro Comp is simply that... well it's hard to explain what it does simply. But let me try this. If you fly the heli fast forward, and then attempt to spin it quickly (a Pirouette) without Piro Comp, it will wobble badly as it goes around. With Piro Comp, it will not wobble.

Most FBL controllers do this already.

Better explain this in simple English….not sure what you mean. Describe how the flight is changed or affected.

Again, very complicated. But basically this allows you to run a low headspeed without experiencing "elevator bounce" or "head bobbing" as it's typically described. Most people talk about head dampers controlling this, but the real fact is it's actually due to the way the controller works, and head dampers just mask it.

Do you need a separate channel for governor?

Yes. Ch8 is used for all motor control functions. If Ch8 is "low", then it's like Throttle Hold. There are several different modes you can select that control what happens when you start the motor. You can use an external governor at a fixed speed, ext with variable speed, control the tail motor, etc.

Is this like ‘facial recognition’ software? Can it stalk people? Seriously…how does it track and follow? Do you need to buy a camera for this function or is a camera built in??

No. You need a device on the ground (or on the target) with a GPS on it, and a radio link. Any smart phone will work. It simply sends GPS positions to the Heli controller and tells it to follow the GPS breadcrumbs. Simple as that.

This is really just a toy feature at this point. I've never used it. I wouldn't even think about trying it on anything but a small quadcopter.

And that is true for all of this. Technically, you could fly a 700 helicopter with nothing more than a tablet. No RC Tx at all. But even though the system is capable of it, I do not recommend using it for flying helicopters without manual control options. Nor should you fly beyond visual range, etc.

[Jgutierrez]

Another cool low level feature could be device pairing, that is the user would pair his Pixhawk to his mission planner and Tx. If the UAV was lost or stolen or wouldn't work unles the pairing was removed using the original PC and Tx. If plugged into a different PC it coul display a warning message with contact info for a reward or something.

A feature like this could deter theft

[PilotSmith]

Wow, very nice! I have not had time to fly my heli in a long time and put together a cheap quad to have something I can fly and not break from rusty thumbs. I was marvelling at how stress free it was to fly the quad and was wishing there were flight controller boards that would do the same for a heli. Looks like there is! Technology sure is moving fast.

[StraightUp]

This is great news!!! I remember reading about some of the limitations on memory and CPU power of the APM which prevented programmers form adding same very cool features.

now this is on my buy list

One of the features I was talking about almost a year ago with Rob (re a DJI thread) was intelligent auto home. basically the unit marks GPS waypoints (including height) as you fly, then if you hit the return to home button or your lose signal the heli automatically retraces the route you took to get back home (avoiding any obstacles you avoided ). This could then be coupled with battery monitors to make sure it has enough power to make it back and if not it will move say 500 meters back, and if signal is not regained then hover until battery is low (giving pilot a chance to reach heli) and then land very very slowly giving any people a chance to get out of the way, maybe even triggering a emergency decent alarm.

so many possibilities

[GerdS]

As the Pixhawk is quite wide with 50mm and doesn't fit into most helicopter chasis, is there a chance to mount it 90° tilted, with its upper side oriented to the left?

Adn did somebody already have a chance here ot test it within a helicopter?

Regards, Gerd

[onepiece]

Hi GerdS,

I've been waiting with baited breath on PIXHAWK news. More in another extension of this tread. PIXHAWK may be unwilling/unable to support traditional helis.

On mounting...since the PIXHAWK doesn't enjoy widespread popularity ...yet....I've been doing some mental gymnastics on best ways to mount in in a 450 sized heli..

JAPAN DRONES sells the CF plate mentioned in this webpage and the dimensions look good for use with the new PIXHAWK:
http://code.google.com/p/arducopter/...dHeli_Assembly

What I'm thinking : buy a landing strut assembly for a 500 or 550 sized heli and "creatively mount" the above noted CF plate approx halfway between the landing skids and the lower frame (horizontal component) of my 450 sized heli. Mounting points are TBD somewhere on the 450's lower CF frame component.

The larger 500/550 struts SHOULD give me all the clearance I need. (Will have to shop around with various skid makers for a nice looking solution). Yes the "stance" will be a bit broad for a 450 heli...but the CF plate will give me all the room needed to wire and adjust.

Hope others contribute ideas for all aspects of PIXHAWK and traditional heli use....

Now..for a bigger problem...read on to the next concern for PIXHAWK.

[R_Lefebvre]

Originally Posted by GerdS

As the Pixhawk is quite wide with 50mm and doesn't fit into most helicopter chasis, is there a chance to mount it 90° tilted, with its upper side oriented to the left?

Adn did somebody already have a chance here ot test it within a helicopter?

Regards, Gerd

Yes, it should be absolutely possible to do this. There is a setting to rotate the controller in any orientation you could want. The only complication is when you also have the external compass, you have to set it's orientation too. And it's confusing because what matters is it's orientation relative to the controller. It's not that hard, but something to keep in mind when setting it up.

Another option is to use the PX4 board, which is actually very small. This is my preferred solution for helis in fact. It has a number of other advantages as well. It's only 31mm x 50mm. You have to buy the PX4 FMU, and the IO board, and they stack.

https://pixhawk.ethz.ch/px4/modules/px4fmu

https://pixhawk.ethz.ch/px4/modules/px4io

You can even rotate these on their side, which would make for a very compact installation in most frames.

[GerdS]

Hi Rob,

I have also seen the PX4 board but also have read that this was just an interim solution on way to PixHawk and will probably not be supported very much in future.

If using the GPS/compass combo, how can the orientation of that external compass get set?
And can the orientation settings be done via the configuration software or must the code get recompiled for it?

I have also looked at the few videos around showing the 3.x version in action with a helicopter and found two drawbacks:

- In loiter mode the helicopter is drifting a lot around and the pilot thinks this is because of "bad GPS weather", I have a multicopter with proprietary autopilot and never seens such behaviour, even in very worse weather conditions.

- There is a video about flying a route, with camera mounted on helicopter. The speed is quite high and the behaviour at the waypoints looks very artificial, with abrupt stopping and acceleration. Nobody would fly this way by intention.

Are there activities for improving these points for helicopters actually? I could imagine that there is no real Kalman filter within the system as the APM hardware was too weak for it yet. Instead there might be some simple fusion algorithm which could be responsible for the loiter drifting. Or has this been imrpoved already for the PixHawk hardware?

Regards, Gerd

[R_Lefebvre]

I haven't seen anything to say that the PX4 was "interim" or that it would not be supported. There's not much reason why that would be. The hardware is almost identical (even more so now that the Pixhawk has had the MPU6000 added back on). And the PX4 will continue to be the hardware of choice for more "academic" pursuits. It's possible that the hardware manufacture could be stopped (but I doubt it), but I don't think the software support would be dropped. We still attempt to support APM1.0 which is nearly 5 years old now.

The orientation of the compass can be set with a parameter in Mission Planner. Very easy. It's the same for the orientation of the control board itself. This used to be a compile-only options, but now is an easy parameter. The only difficult at all is figuring out which orientation you need.

The videos you are talking about are probably mine. The first one, where it's wandering around on GPS, it may have actually been due to Interference from the servos on that heli. I recently discovered the servos were outputting RFI, which was completely preventing GPS lock. I'm not sure how I ever flew with these servos before. I did solve it by adding ferrite beads on the servo wires.

It may have been suffering in that video, hard to say.

I will say that I have a hard time believing that the GPS position never drifts on any hobby or commercial autopilot. It's just the nature of GPS, and there is no magic here. Not without RTK assistance anyway.

If somebody had some way to demonstrate that, I'd like to see evidence.

The second issue, yes, has to do with the way the autopilot functions work. I am curious to see comparable video from other systems, I've never been able to benchmark this, as not many people post videos of this sort of thing.

In any case, I agree that the performance here could be better. It is coming. David Dewey has been working on "Spline Nav" which basically makes the copter follow a spline curve rather than using fixed waypoints. This will probably be included on the next release, and may be a PX4/Pixhawk only thing, as it may not run on APM. I think you'll find this performance much more appealing.

[R_Lefebvre]

Oh, and no, there is no Kalman on APM. And so far, the Pixhawk is just using the APM program. Yes, there's a simple fusion between accelerometers and GPS data. I know that people are already working on a Kalman implementation for the Pixhawk, no idea how long it will take or how much better it will be.

[R_Lefebvre]

This video demonstrates a more typical loiter performance:

[ThirdLaw]

This all sound great but the PixHawk is a unicorn. You might as well say it will re grow hair and make you more attractive to women. Sorry for being so salty but the thing is just so interesting and I can see how much potential it has. Then they give deadline after deadline ( 6 of them before I gave up on them ) hard not to harbor some resentment.

[R_Lefebvre]

Well... OK, but 6 deadlines in 2 months. It's only delayed 2 months, not actually as long as "6 deadlines" would make it seem.

They did have an unexpected problem which is what caused the delay. They could have done what many commercial systems do: Sell it anyway, blame the problem on something unrelated, and then fix it with V2.0.

I believe it is or will be shipping soon. I got one, and now they have to fulfill backorders, and try to keep up with new orders. It's going to be on short supply. The cost of success I guess.

In the meantime, APM works great!

[GerdS]

Hi Rob,

The above videos look good but represent multicopters, not helicopters.
I am fearing a bit that the development focus goes just to multicopters and "normal" helicopters" are not tested that deeply. In fact if searching by Google for user experience there is almost no hit.

Another question, is there a reliable "bailout" function? E.g. if the helicopter is flown in manual mode, gets out of control and I then will switch to loiter mode and release the sticks, what will happen?
I lost a scale helicopter with DJI Wookong H this way, as the Wookong did take actual stick settings at the point of switching to GPS+Atti mode as new stick center positions and made the helicopter uncontrollable.

The PX4 board may better fit into helicopters from size, but is lacking a housing which is a must for all-day usability. Also I have no idea actually about its height all together with base board and connectors attached.
But a collegue has ordered a PixHawk to be used on his multicopter and so I will see it hopefully soon and can check if it would fit 90° rotated into my helicopter chasis.

Regards, Gerd

[R_Lefebvre]

IMO, the Loiter of a Heli is even better than a multirotor. More stable.

If you are flying in Acro, and need to Bail Out, then I recommend flipping to Stabilize. This will bring the helicopter upright instantly. I use it all the time and it works well. I don't know what would happen if you flipped to Loiter, that might not go well, I don't know but I've never tried it.

There's no silliness where flipping to Stabilize would take a new center position.

Yeah, if I had the money to invest, I'd do a run of CNC machined cases for the PX4.

[careyba]

I am looking at using this on my Raptor E620 and have a few questions:

1) I am assuming that this unit has a tail gyro (heading hold style) so I can discard my existing gyro.
2) How far from the motor do I typically need to mount the GPS sensor (motor is at the rear of the E620)
3) It is mentioned in the sales info above that this is not a FBL controller but a UAV controller. I am only interested in sports flying (not 3D) so will the Pixhawk allow me to still fly my chopper (via mode switches) if I like in full cyclic and pitch control.

Cheers, Brad