Espritmodel.com Telemetry Radio
Reply
Thread Tools
Old Dec 06, 2013, 05:04 PM
Registered User
The Willamette Valley, Oregon
Joined Dec 2008
1,079 Posts
Discussion
Let's make an artificial horizon display from an RC autopilot

Let's make an artificial horizon display from an RC autopilot. Display it on an Android phone or something. Let's talk practical. Small and compact. Something you might carry in your pocket or flight bag to pull out for emergency assistance in a light plane or sailplane. Should boot up quickly with a minimum of fuss. The display should be easy to see and should have just the artificial horizon display, not a bunch of superfluous stuff. Maybe a heading indicator display as well. Decent battery life, several hours minimum, with good warning of when batteries are getting low.

Where would you start? Which components?

Has this been done already?

Steve
aeronaut999 is offline Find More Posts by aeronaut999
Last edited by aeronaut999; Dec 06, 2013 at 05:12 PM.
Reply With Quote
Sign up now
to remove ads between posts
Old Dec 06, 2013, 10:17 PM
Registered User
Near Austin
Joined Dec 2001
5,354 Posts
Yep, been done. Had it on my old phone and it was fun to use during business trips on commercial flights at night. Sure you can find on on Google Play
CenTexFlyer is online now Find More Posts by CenTexFlyer
Reply With Quote
Old Dec 06, 2013, 11:10 PM
Registered User
The Willamette Valley, Oregon
Joined Dec 2008
1,079 Posts
Quote:
Originally Posted by CenTexFlyer View Post
Yep, been done. Had it on my old phone and it was fun to use during business trips on commercial flights at night. Sure you can find on on Google Play
Really? Are you talking about a phone app with no extra hardware involved? I'm a bit skeptical (but would love to be corrected). I'm talking about the real deal, something that works well for prolonged circling at bank angles up to at least 45 degrees, as well as for serious pitch excursions. My understanding of the phone apps was that they were getting inputs only from the linear accelerometers built into the phone, which as far as I can see would not really be a suitable input for a full-functioning artificial horizon display.

Steve
aeronaut999 is offline Find More Posts by aeronaut999
Reply With Quote
Old Dec 07, 2013, 10:23 PM
Registered User
Near Austin
Joined Dec 2001
5,354 Posts
Well, I could only talk the Southwest pilot into one turn of about 30 degrees and that was it. Some Androids have a gyro in them as well as accels.

https://play.google.com/store/apps/d....ds.gyro&hl=en
CenTexFlyer is online now Find More Posts by CenTexFlyer
Reply With Quote
Old Dec 09, 2013, 08:21 PM
Registered User
HELModels's Avatar
Joined Aug 2003
2,616 Posts
Interesting. I've read a few things that always say that fused gyro/accel needs airspeed in an aerial application to get rid of centripedal effects. I've tried putting my own IMU code together, using parts of other programs. I can get a display of attitude on my computer, but using it to control a flight is still a work in progress - rudder input showing up on elevator and elevator on aileron isn't very good and confuses the hell out of me and I always put it to the side.

I know you only want a display. Post what you can about what you use. .
HELModels is offline Find More Posts by HELModels
Reply With Quote
Old Dec 09, 2013, 08:53 PM
Old fart. VAAATSP
d_wheel's Avatar
Deep in the Heart of Texas
Joined Jun 2004
1,931 Posts
I found 13 for the iphone/ipad. Prices vary from free to $100.00. Not sure if they work in an airplane but I can't make the free one I downloaded show false indications when moved around rapidly.

Later;

D.W.
d_wheel is offline Find More Posts by d_wheel
Reply With Quote
Old Dec 10, 2013, 09:11 AM
MultiRotor Guru
Norrad's Avatar
Thailand, Bangkok
Joined Feb 2009
110 Posts
I think easier said than done. The Artificial Horizon is based on an Earth Gyro which is spun up while the aircraft is level on the ground, the gyro maintains a fixed position relative to the Earth's Centre and the aircraft moves around it. <-What I remember from AT and G training during my PPL. If I was going to design an electronic one, I would probably use an Infrared Horizon sensor. Something that could simply detect the ambient temperature difference between land or sea, and air and then technically display the real horizon.
Norrad is online now Find More Posts by Norrad
Reply With Quote
Old Dec 10, 2013, 01:40 PM
Registered User
HELModels's Avatar
Joined Aug 2003
2,616 Posts
This is kind of a fun little program that runs in processing 2.0 which can run on android or unlocked iphones. I'm going to play with it and see if I can feed my IMU to it. As is it only reads mouse movements, so it will take some work.

http://70.179.172.144:8080/code/
HELModels is offline Find More Posts by HELModels
Last edited by HELModels; Dec 10, 2013 at 02:02 PM. Reason: link was broken
Reply With Quote
Old Dec 11, 2013, 11:38 AM
Registered User
The Willamette Valley, Oregon
Joined Dec 2008
1,079 Posts
I believe that a rotational accelerometer for each axis would be essential. I highly doubt that any phone app would cut the mustard. I'm imagining taking an autopilot intended for use with r.c. aircraft and displaying the output on a phone or small tablet. Should be simple enough I think. Most of the work might be in writing the software to generate a nice display with an artificial horizon and a heading indicator.

Beware-- some so-called artificial horizon displays generate the pitch information based on the aircraft's GPS-derived vertical velocity vector. This is likely to be the case with some of the phone apps. I definitely don't want that. For example, I don't want the output to be different in a 1000 fpm updraft vs a 1000 fpm downdraft, for any given actual pitch attitude of the aircraft.

Nor do I want the bank information to be derived from the GPS-derived ground track. That could be problematic when circling in very high wind where the groundspeed may approach zero. Plus, GPS reception isn't always very good inside a light plane. GPS information shouldn't be used at all, except possibly to help compensate for sensor "drift" over a long period of operation.

I'm sure the phone apps aren't going to cut it, but the autopilot technology that is abundantly available for use in RC aircraft should work very well. To take one random example:

http://www.eagletreesystems.com/imag...shot-large.jpg

http://www.eagletreesystems.com/imag...gers-large.jpg

http://www.eagletreesystems.com/inde...&product_id=50

So what hardware is the BEST choice to build this system around?

Let's do this!

Steve
aeronaut999 is offline Find More Posts by aeronaut999
Last edited by aeronaut999; Dec 11, 2013 at 11:43 AM.
Reply With Quote
Old Dec 11, 2013, 11:47 AM
Registered User
The Willamette Valley, Oregon
Joined Dec 2008
1,079 Posts
Quote:
Originally Posted by Norrad View Post
I think easier said than done. The Artificial Horizon is based on an Earth Gyro which is spun up while the aircraft is level on the ground,
That is a relevant point to consider-- see below. Steve
aeronaut999 is offline Find More Posts by aeronaut999
Last edited by aeronaut999; Dec 11, 2013 at 01:07 PM.
Reply With Quote
Old Dec 11, 2013, 01:09 PM
Registered User
The Willamette Valley, Oregon
Joined Dec 2008
1,079 Posts
Quote:
Originally Posted by Norrad View Post
I think easier said than done. The Artificial Horizon is based on an Earth Gyro which is spun up while the aircraft is level on the ground, the gyro maintains a fixed position relative to the Earth's Centre and the aircraft moves around it.
No problem. It's ok with me if we have to let it calibrate for a few seconds in wings-level flight. I don't expect to be able to boot it up in the midst of a spiral dive...

But that brings to mind one more feature that we could tack on, besides an artificial horizon and a heading indicator. A turn rate indicator built around a one-axis gyro (piezoelectric rotation rate sensor) has the unique advantage that it can be up and running with no calibration phase, as soon as it is switched on, even in the midst of a spiral dive etc. Any error in calibration will just cause the aircraft to be in a very slight turn when the pilot thinks it is wings-level. In the context of emergency recovery from an unusual attitude, this slight error is trivial. So it would be a very good idea to include a turn rate display, driven by the yaw rate sensor only. Or better yet, driven in the manner of a "turn coordinator", with a single gyro canted about 35 degrees from horizontal to mix in some roll sensing as well as yaw sensing. (Couldn't this also be done by mixing the output from a yaw rate sensor with the output from a roll rate sensor?) Either way, I would prefer the output of this feature be in the form of the old style turn rate indicator needle, or something analogous to that, not something that attempts to represent the bank angle or horizon line. To avoid any possible confusion with the artificial horizon display.

The idea is to give the pilot something that will work even when the system is switched on the in the midst of a spiral dive etc, or if the artificial horizon crashes due to hardware or software glitches or for any other reason. As noted by Norrad it seems likely that the full 3-axis artificial horizon display will need to be calibrated in level flight before functioning, unless we have some sort of extremely sophisticated system of calibration based on GPS-generated flight track data. I'm not looking to go to that level of complication.

More on the turn rate feature-- the turn rate output on the main display screen could also be replicated by a pair of LED's on the main hardware package, that flash to indicate a left turn or right turn, with the rate of flashing proportional to the turn rate. So now we have some redundancy even if the main display (Android phone or whatever) becomes inoperative for some reason.

An alternative philosophy would be to relegate the turn rate indicator feature to a completely separate piece of hardware, for better redundancy. As noted above I already have such a thing. But in real life this whole thing is meant as an emergency backup only, and we won't likely be carrying a backup for the backup, so why not go ahead and provide the turn rate output, both on the main display screen and also by means of a couple of LED's piggybacked right on to the main hardware unit. The idea of being able to pull something out of your pocket and switch it on as a spiral dive is starting to develop, and be immediately provided with useful indications on how to recover, is too good to pass up. In a different situation where you have not yet penetrated the clouds, and can hold the aircraft level as the instrument boots up, you'll also be provided with a fully-functioning artificial horizon display, and a heading indicator.

I don't know whether it will likely be convenient to drive the heading indicator from a magnetic sensor, or whether the heading will need to be set manually, possibly by means of a simple knob on the main hardware package. Either would be acceptable. It might well be simplest just to have the pilot set the heading manually, as is currently the case with the old-style heading indicator with actual mechanical gyro.

I'd rather simply give the pilot the option to calibrate/ recalibrate the instruments as needed, than involve a magnetometer in the system, if the magnetometer ends up causing problems when metal parts (e.g. the control yoke or stick) are moved around in close proximity to the firmware package. I don't know much about these systems but I suspect it would be best not rely on the output from a magnetometer.

No part of the system should be dependent on good reception of GPS signals (except of course for the ground track display.)

Summary of desired features--

* Turn coordinator (mixed yaw rate/ roll rate) display or turn rate (yaw rate only) display, designed to function well regardless of the attitude of the aircraft during start-up. For redundancy, output should include both a simple LED-based output built right onto the main hardware package, and a virtual turn-rate-indicator needle, or something like that, on the display screen. There may be a provision to calibrate the display in non-turning flight for better accuracy, but the default mode of operation should not include a calibration phase. To save space, the turn rate indication could be depicted on the display screen in the form of a simple rectangular bar underneath the atificial horizon display something like this-- (..*.....I........) - this is a depiction of left turn. The turn rate display should not be incorporated inside of the artificial horizon display, because the pilot should be able to concentrate on the turn rate display alone if he suspects that the artificial horizon display is malfunctioning.

* Full-functioning artificial horizon display on the display screen. Presumably a straight-and-level calibration phase will be required for this feature to work. Display should be flagged inoperative until calibration is achieved.

* Heading indicator on the display screen. Not just a numerical display, but a virtual depiction of an aircraft heading indicator, or something like that. To save space the display could somehow be combined with or superimposed on the artificial horizon display, as both are reliant on the same hardware and both will likely either be operative or inoperative at the same time. The heading indicator may be driven by a magnetometer, or may simply be manually set by the pilot, by means of a knob on the main firmware package. In real life we would expect that the pilot would be making some reference to a GPS display, which would provide a reference for setting our gyro-driven heading indicator. Again, a straight-and-level calibration phase may be required for the heading indicator to work. The heading indicator should NOT be driven by GPS-derived ground track information. For example in circling flight in strong wind, at a constant bank angle, we want to see a steady rotation rate of the dial, rather than the surges and lags that we would see in the rate of change of a heading display based purely on the GPS-derived ground track.

* Numerical display of GPS-derived heading (ground track), IF that information is readily available from our firmware anyway.

Steve
aeronaut999 is offline Find More Posts by aeronaut999
Last edited by aeronaut999; Dec 11, 2013 at 01:27 PM.
Reply With Quote
Old Dec 11, 2013, 06:24 PM
Registered User
HELModels's Avatar
Joined Aug 2003
2,616 Posts
I don't know how it does it but the MPU6000 running FreeIMU will boot in any attitude and show that attitude. It takes a few seconds to settle in and then it will respond correctly. It will drift in yaw because gravity is aligned with the rotation axis and something is undefined when the math is done. A magnetometer fixes that.

APM/ArduIMU uses "phone sensors" like the MPU6000 which combines gyro/accel and a seperate magnetometer. There's even a new invensense chip that adds the magnetometer on the chip.

This proposed instrument sounds more like a display problem than a find the right sensor problem.
HELModels is offline Find More Posts by HELModels
Reply With Quote
Old Dec 11, 2013, 07:38 PM
Registered User
The Willamette Valley, Oregon
Joined Dec 2008
1,079 Posts
Thanks for posting...

Amazing things can be done with a magnetometer...

Years ago was invented an instrument called the "Bohli compass" which featured an unweighted needle free to pivot into full 3-dimensional alignment with the local magnetic field. One end of the needle had a brightly colored "pipper" which the pilot was supposed to watch as it circled round and round inside a spherical or hemispherical transparent housing. I suspect it was rather unintuitive to use in practice, but in theory it provided pitch attitude information as well as turn rate and bank angle information.

Let's do this!

Steve
aeronaut999 is offline Find More Posts by aeronaut999
Reply With Quote
Old Dec 12, 2013, 04:39 AM
Registered User
Joined Apr 2013
65 Posts
I did a blog post on IMU maths. It covers all the basics for a quaternion based system.

http://www.camelsoftware.com/firetail/blog/c/imu-maths/

And an actual implementation running on Arduino

http://www.camelsoftware.com/firetai...mu-10-arduino/
SuperCamel is offline Find More Posts by SuperCamel
Reply With Quote
Old Dec 12, 2013, 01:20 PM
Registered User
The Willamette Valley, Oregon
Joined Dec 2008
1,079 Posts
Quote:
Originally Posted by HELModels View Post
I don't know how it does it but the MPU6000 running FreeIMU will boot in any attitude and show that attitude. It takes a few seconds to settle in and then it will respond correctly.
I wonder if it would read correctly if booted up while the aircraft was performing a banked turn? It would have to being doing something sophisticated with a magnetometer or GPS to pull that off. That sounds ideal, if it really works!

If need be, I'd be willing to settle for something that needs to calibrated (or self-calibrates) in straight-line flight. I do like the idea of also having a yaw rate display (or combined yaw/roll "turn coordinator" display) that is ready to go instantly regardless of the aircraft situation at start-up, as noted above.

Steve
aeronaut999 is offline Find More Posts by aeronaut999
Last edited by aeronaut999; Dec 12, 2013 at 01:31 PM.
Reply With Quote
Reply


Thread Tools

Similar Threads
Category Thread Thread Starter Forum Replies Last Post
Data How does artificial horizon display work? davensocal Eagle Tree Systems 32 Jun 07, 2012 03:51 AM
Discussion Guys lets do it! Lets make an official rc television show!!!!!!!!!!!!! lightspeeddud Fuel Plane Talk 0 Jan 28, 2009 02:07 PM