PDA

View Full Version : AP question


N941WR
11-14-2008, 10:45 AM
For those of you with TT, Trio, or other autopilots...

Can you engage them while in unusual attitudes? If you can engage them, what do they do?

For example, if you are nose high and in a steep bank, how does the autopilot behave if you (try to) engage it? Does it continue to fly on in the unusual attitude or does it roll wings level and capture your heading?

scard
11-14-2008, 10:52 AM
TT VSGV It will engage, roll wings level, capture current heading (actually track if it has GPS input), and attempt to maintain current vertical speed (airspeed limited, it won't stall the airplane).

N941WR
11-14-2008, 12:32 PM
Scott,

Thanks, that is the kind of info I was looking for. Is that true for all the TT autopilots?

One more question, what happens if the AP is engaged while you are inverted?

scard
11-14-2008, 12:38 PM
Scott,

Thanks, that is the kind of info I was looking for. Is that true for all the TT autopilots? Probably, except for the vertical part.


One more question, what happens if the AP is engaged while you are inverted? Yeah, right, like I'm going to comment on that :-).

N941WR
11-14-2008, 12:44 PM
...One more question, what happens if the AP is engaged while you are inverted?
Yeah, right, like I'm going to comment on that :-).
I'm trying to find out if these things can save your backside if a VFR pilot flies into a cloud, loses control, and then engages the AP.

SteinAir
11-14-2008, 02:13 PM
I'm trying to find out if these things can save your backside if a VFR pilot flies into a cloud, loses control, and then engages the AP.

Come on Bill.....be honest here....you're a Beta tester for Dynon and their Autopilot?!?! :)

Most AP's with good gyros in them will right the airplane. Garmin has the Blue "Level Button", and others are doing the same thing now to allow a customer to hit the "Blue Panic" button to right the aircraft from unusual attitude and have it right the aircraft.

Anyway, I assume this is something you (or they) have tested with the Dynon AP? Any comprehensive development testing of an Autopilot would include unusual attitudes, so I assume Dynon has this data somewhere?

Cheers,
Stein

dynonsupport
11-14-2008, 02:32 PM
Most AP's with good gyros in them will right the airplane. Garmin has the Blue "Level Button",

Actually a rate based autopilot will roll the wrong way when inverted or in some uncoordinated maneuvers. There's a reason the Garmin/Cirrus blue button only works to 75 degrees of roll and 50 degrees of pitch.

Anyway, we didn't ask Bill to ask the question or even bring up the topic. It's his own personal curiosity. How else is he supposed to find out how all the other AP's in the market behave?

We may eventually advertise really extreme unusual attitude recovery as part of our ADAHRS based AP, but for now that will require more testing. We have of course done quite a bit of testing already and are confident that it compares favorably to other AP's on the market.

SteinAir
11-14-2008, 02:57 PM
Actually a rate based autopilot will roll the wrong way when inverted or in some uncoordinated maneuvers. There's a reason the Garmin/Cirrus blue button only works to 75 degrees of roll and 50 degrees of pitch.

Anyway, we didn't ask Bill to ask the question or even bring up the topic. It's his own personal curiosity. How else is he supposed to find out how all the other AP's in the market behave?

We may eventually advertise really extreme unusual attitude recovery as part of our ADAHRS based AP, but for now that will require more testing. We have of course done quite a bit of testing already.

Don't worry, Bill R is a good guy and I think we get along well - I've been getting helpfull advice from him on his taildragger -9 for awhile...sometimes I just have to rib him a little bit and we do a bit of back/forth.

Anyway, absolutely good points....I deleted some of what I had typed in my original post for sake of length. I was going to say that if you're completely inverted, knife edge or other crazy/extreme or unusual attitude in IMC then you are in DEEP doo-doo anyway and there is not much anyone/anything can do! The whole goal of any system is to keep that from happening to begin with....which is naturally everyones goal. Also, please do be careful though about drawing conclusions to my quotes by making the leap between what I actually said and your response. I didn't say anything about the inverted part, you did. I was speaking in general of unusual attitudes....so please don't make my response into something it wasn't.

Good to know you won't advertise it before you do the testing on it....quick trivia question - If you are inverted, which is the "right" way to roll upright? Have you verified that the Garmin only works at the degrees stated above while flying in a Cirrus (and therefore by your implication doesn't work outside those paramenters)? Sometimes what is written on a paper to cover a legal point isn't always 100% the limit of a system....

Cheers,
Stein

BTW....why don't you guys sign your posts? All of the other mfgrs do? None of them remain completely anonymous?! Just personal curiosity, that's all! :)

airguy
11-14-2008, 03:03 PM
quick trivia question - If you are inverted, which is the "right" way to roll upright?

Cheers,
Stein



Simple - the direction the roll rate is already progessing in, if truly exactly inverted. If inverted and stable, I would assume roll into engine torque, not that it matters much at that point...

N941WR
11-14-2008, 03:18 PM
Come on Bill.....be honest here....you're a Beta tester for Dynon and their Autopilot?!?! :)...
Stein,

I am a Beta tester for Dynon but I have never flown with any other autopilot, so I don't know how they work, thus the questions.

..Anyway, we didn't ask Bill to ask the question or even bring up the topic. It's his own personal curiosity. How else is he supposed to find out how all the other AP's in the market behave?
Dynon did not know I posted this question, it was for my own education only. A number of people have asked me to compare the Dynon AP to other AP and I can't answer when I don't know how the others work and I don't like giving out BS answers.

Don't worry, Bill R is a good guy and I think we get along well - I've been getting helpfull advice from him on his taildragger -9 for awhile...sometimes I just have to rib him a little bit and we do a bit of back/forth.
Stein, no ill will on my part. Your answers are honest and truthful, which is what I like, even if I don't like the answers. BTW, I've been meaning to email you and ask how the -9 is coming.

As for the radical unusual attitudes, read what Dynon said. These guys really have created a very good autopilot but as I said above, I have only flown with their unit, so my experience base is limited and I'm trying to learn more.

dynonsupport
11-14-2008, 03:58 PM
BTW....why don't you guys sign your posts? All of the other mfgrs do? None of them remain completely anonymous?! Just personal curiosity, that's all! :)

Interesting. We thought it had the opposite effect. We think it's an easy way to make it really clear that the response is on behalf of the company. Also, there are a few of us that post under the banner of dynonsupport, so we find it easier to manage the single account and search for posts we've collectively made. Some of us do maintain our own accounts too, and make it a point to ID our employment when voicing our personal opinions.

Rainier Lamers
11-14-2008, 10:55 PM
For those of you with TT, Trio, or other autopilots...

Can you engage them while in unusual attitudes? If you can engage them, what do they do?

For example, if you are nose high and in a steep bank, how does the autopilot behave if you (try to) engage it? Does it continue to fly on in the unusual attitude or does it roll wings level and capture your heading?

Good question.
In case of ours - it depends and I don't have a straight answer (no pun intended). The answer I am giving is for a typical 2 axis system where one axis is typically connected to the ailerons and the other to the elevator. I am qualifying this as you can connect up to 6 servos to our system and thus can theoretically have a full authority system that will even drop your gear before touchdown (autoland if you have a radar altimeter). But let's stay with a normal 2 axis system as this is what most of our customers will be using (at least I think so).

If you do not have a current navigation solution (i.e. not an active GPS "goto", route, ILS, GLS, VOR, etc, etc), the autopilot will engage and do whatever it takes to hold the heading as well as the altitude at time of engagement (i.e. it automatically sets and switches to the altimeter and heading bug). The autopilot will cause the servos to engage at their current position and will then control them to achieve straight and level, subject to low and high airspeed limit restrictions.
If you have an active navigation solution prepared, as you engage the autopilot will then latch onto the heading (and altitude as required) from the navigation source correcting any unusual attitudes as required.

As interesting point, should you engage while you are upside down, it will attempt to level in upside down attitude first, before rolling you the right side up, all the while being aware of both lower and upper airspeed limits.
Obviously, some aircraft can do this better than others so this may be a little academic (you can tell I'm having fun with this thing :D). The exact reaction also depends highly on the type of servos fitted (mainly servo speed) and how the autopilot system is characterised (and this is decided by our users).

As a footnote, our autopilot has not yet been released so final operation may differ slightly (but not much) and, as usual will likely develop with time due to user input and special requests - so no telling what it will do one day.

To be fair, autopilots that form part of EFIS systems tend to be able to do much more compared to stand-alone autopilots. The reason for this is that only the EFIS "knows" everything - actual attitude, rate of turn around all axis, magnetic heading, GPS track and heading, airspeed, vertical speed, climb or descent angle, G-forces, even terrain proximity. If you think about it - it's really the only place for an autopilot to be.

Many stand-alone autopilots simply do not have the luxury of all this data. Nevertheless, for simple flight stabilization and keeping track they can be just fine. Many are based on just a simple, single gyro mounted such that it will detect both roll and turn (but actually cannot distinguish between these). Needless to say, these types would not know if they are upside down.

Rainier
CEO MGL Avionics

P.S. Off to the airfield in a few minutes to do some more testing on our autopilot system. I'm having LOT's of fun...

Alschief
11-15-2008, 06:27 AM
Does anyone have available the Roll Servo mounting under the seat of an RV6available for posting?

dynonsupport
11-15-2008, 10:12 AM
We have all of our mounting kit install instructions on our web site. They don't have photographs, but they do have very detailed 3D CAD drawings.

The RV-6 Roll is here:

http://dynonavionics.com/downloads/Install_Guides/RV6%20Roll%20Kit%20Instructions.pdf

rvmills
11-15-2008, 12:08 PM
Bill,

Interested to hear if you've done any testing on engagement in unusual attitudes, and how the Dynon AP reacts.

Once I have mine up and running, I can try to emulate what you've done and get some more data points, but would like to hear about your experience with it going in.

It might be nice to know what'd happen if I was to do a roll with one of my boys and they reached over and said, "hey Dad, what does this big AP button do?" as we transited the inverted. (I'm kiddin' around here of course!) :)

Still owe ya that call on wiring!

Oh, and Al, will post some pix of the roll servo install when I wrap it up next week. Would be sooner, but gotta head back to work tomorrow after the SJSU-UNR game today (go Spartans!) (I know, not a big game on everyone's radar...but the alma-mater is in town!)

Cheers,
Bob

N941WR
11-15-2008, 12:32 PM
Bob,

The results of the Dynon autopilot beta testing belong to Dynon, not me, even though I may have performed some of the tests.

It is up to them to release the results, not me.

So, I do not feel comfortable discussing the results of a specific test.

rvmills
11-15-2008, 12:44 PM
Roger that Bill, no worries!

Cheers,
Bob

Ron Lee
11-15-2008, 01:46 PM
As for the original question, it makes sense to engage the autopilot well before you leave controlled flight. My procedure for this event is to engage the roll autopilot (if not already engaged) to ensure roll stability. Pitch control "should" be ok just by leaving the stick alone and using instruments as required. I don't have a pitch autopilot now.

Activate the 180 degree function if it is the best solution.

I should practice this with another pilot using goggles.

49clipper
11-16-2008, 09:47 AM
For those of you with TT ADI II, the manaul says on the ground check, the movement should be smooth when moving the control to roll left or right. Mine is anything but smooth. Its a 200 step motor, and it just steps with a clunk till it reaches the roll limit. Is that correct?

SteinAir
11-16-2008, 11:29 AM
For those of you with TT ADI II, the manaul says on the ground check, the movement should be smooth when moving the control to roll left or right. Mine is anything but smooth. Its a 200 step motor, and it just steps with a clunk till it reaches the roll limit. Is that correct?

Is your AP engaged? - if so, then yes. The "ground check" should be performed with the servos dis-engaged.

Cheers,
Stein

RecVehicle123
11-22-2008, 04:38 PM
[QUOTE=N941WR;271907]Scott,

Thanks, that is the kind of info I was looking for. Is that true for all the TT autopilots?

Its true of all TT's with vertical.

kentb
11-24-2008, 12:43 PM
So if there can be a "Blue" button to save you from an upset condition, can a "Red" button be added to make the AP do a loop or a roll?:D

Kent

N941WR
11-24-2008, 12:55 PM
So if there can be a "Blue" button to save you from an upset condition, can a "Red" button be added to make the AP do a loop or a roll?:D

Kent
Kent,

I'm sure the Dynon could be so programmed, as long as the maneuver didn't require the use of the rudders.

After all, it does know your attitude, airspeed, G's, etc.

That would be cool but you could bet someone would engage the red "Easy Button" close to the ground and then the attorneys would get involved and put the company out of business.