How does the auto pilot know that the plane is right-side up?

SixPapaCharlie

SixPapaCharlie

May the force be with you
Joined
Aug 8, 2013
Messages
16,572
Display Name
Display name:
Sixer
Really nothing to add here.
How does it know?
 
It knows the aircraft is right-side-up because the aircraft is not certified for inverted flight, e.g. the aircraft *must* be right-side-up.
 
Clark1961 said:
It knows the aircraft is right-side-up because the aircraft is not certified for inverted flight, e.g. the aircraft *must* be right-side-up.
Click to expand...

Yep but how does it know I am level?

If I have a gyroscope, and it tilt it, it fights me, however after a bit of time, tilted gradually becomes the new level.
 
SixPapaCharlie said:
Yep but how does it know I am level?

If I have a gyroscope, and it tilt it, it fights me, however after a bit of time, tilted gradually becomes the new level.
Click to expand...

That's 'cause ya gots the wrong autopilot. Rate based autopilots decide what to do based on what is changing. Attitude autopilots decide what to do based on what the current state is (and there a a couple inputs on that).
 
If I flew a plane in a forward slip for x number of minutes, could the AP gyro reorient itself so that it reacts as though banked is the new level?

I assume not but I don't know why not. Google is not helping much yet.
 
Bottom line, it's too stupid to know. If the airplane is flying the heading or track that you tell the autopilot you want, it will provide whatever roll inputs it needs to keep it there. If the gyros "tell" the autopilot that it's in a 20-degree bank to the left when the airplane is wings level, the autopilot will simply "think" that it takes a 20-degree bank to the right to hold heading.

I forget which (ancient) autopilot it was, but I used to fly one that would only hold heading...and the only heading it would hold was 360 degrees. You simply lied to the autopilot and adjusted the DG so that it showed 360 degrees when the airplane was on the desired heading.
 
Clark1961 said:
That's 'cause ya gots the wrong autopilot. Rate based autopilots decide what to do based on what is changing. Attitude autopilots decide what to do based on what the current state is (and there a a couple inputs on that).
Click to expand...


So one possible option might be what another poster alluded to which would be reading the AI. I sort of suspected maybe that was the case.

>>> Sub question. Why doesn't the AI precess?


But the mention of the gyro chips is interesting and I am trying to dig info up on this. I am not sure... Er actually I know I don't understand how they work and stay true.
 
SixPapaCharlie said:
If I flew a plane in a forward slip for x number of minutes, could the AP gyro reorient itself so that it reacts as though banked is the new level?
Click to expand...

Interesting question. I don't know the answer. I know the attitude based Century autopilot in the 'kota doesn't care what I did in the past, it'll follow its analogue programing, the attitude input, the altimeter input (if altitude hold is selected) and the nav input (if selected) and decide what to do based on now, not what has happened in the past. The attitude indicator is self correcting (as long as it isn't broke and has vacuum) so history doesn't matter.

In other words, attitude reference error is based on an assumed correct pitch with closed loop adjustment to level or operator specified pitched flight. Bank is closed loop adjusted to either wings level or heading/nav hold depending on mode selected.
 
SixPapaCharlie said:
So one possible option might be what another poster alluded to which would be reading the AI. I sort of suspected maybe that was the case.

>>> Sub question. Why doesn't the AI precess?


But the mention of the gyro chips is interesting and I am trying to dig info up on this. I am not sure... Er actually I know I don't understand how they work and stay true.
Click to expand...

Attitude based autopilots do have input from the AI. Rate based autopilots use the turn coordinator for gyro input. Either type of autopilot can have altimeter input. The early Century autopilots have a dedicated altimeter called an altitude chamber or can. They work but the operation is a bit strange from a digital world point of view. Of course the early Century autopilots are analogue so they make semi-perfect sense in that context.
 
MauleSkinner said:
According to the current Instrument Flying Handbook, it has an "erection mechanism". Some of the older versions were more descriptive.:rolleyes:
Click to expand...

Yup, they are "self-correcting" as long as everything is working right. The earlier descriptions were excellent primers on gyro behavior and practical engineering. Now, ya jus' as well say it's PFM.
 
Good question. All autopilots that I've been extremely familiar with in the past did not know if they were upside down or right side up, including some GA autopilots and military airplanes. The F-4 Phantom, for instance, would hold attitude in pitch and roll just fine using attitude gyro(s) as a reference and rate gyros to halt the motion, but yaw, in addition to a rate gyro, used an accelerometer (lateral) and it didn't work right upside down. Altitude Hold would also give opposite corrections inverted, so that mode was no good either.
 
right, how do you know that it knows? maybe it doesn't know about too many things but it knows what it knows if you know what I mean.
 
I assume whatever gyro is installed would have a baseline zero/level orientation, and roll differentials would be calculated from that zero point.

Also, I believe that even solid state gyros still use some form of movement (although miniaturized) to work.
 
SixPapaCharlie said:
Yep but how does it know I am level?

If I have a gyroscope, and it tilt it, it fights me, however after a bit of time, tilted gradually becomes the new level.
Click to expand...

Then it would depend on how sophisticated of an autopilot it was as to whether it just followed the AI or if the conflict between the AI and the changes occurring would kick it out and flag.
 
There is a tiny cat inside the instrument.
 
I have done an experiment with a typical 40yo turn coordinator.
Turn it off with the master (this one had a pullable breaker).
Wait 10 minutes.
Enter a std rate turn and allow the airplane to stabilize.
Turn the TC back on, and continue the turn til it spins up (5mins of turning) to be sure.
In my mind, the TC should think 'up' is about 15° 'off' the vertical following this.
Go about your flying.
In less than 5 minutes, it is reading correctly.
There is something in there which allows it to find the horizon, something I have not been taught yet.
 
Let'sgoflying! said:
I have done an experiment with a typical 40yo turn coordinator.
Turn it off with the master (this one had a pullable breaker).
Wait 10 minutes.
Enter a std rate turn and allow the airplane to stabilize.
Turn the TC back on, and continue the turn til it spins up (5mins of turning) to be sure.
In my mind, the TC should think 'up' is about 15° 'off' the vertical following this.
Go about your flying.
In less than 5 minutes, it is reading correctly.
There is something in there which allows it to find the horizon, something I have not been taught yet.
Click to expand...

That seems amazing to me.
I want to learn how.
 
SixPapaCharlie said:
That seems amazing to me.
I want to learn how.
Click to expand...

Mechanical attitude gyros have 4 vanes in them that are designed to pivot in response to gravity. If the gyro is tilted to one side relative to the pull of gravity, a vane will pivot and allow a jet of air to apply a correcting force (90 degrees from the direction the gyro needs to tilt) to erect the gyro.

A side effect to this is that centrifugal (yeah, I know...:rolleyes:) forces in a turn will cause the gyro to precess because the vanes sense "gravity" in some other direction than towards the center of the earth and try to "correct" the gyro. This precession will be worst at 180 degrees of turn, and will correct itself if the turn is continued to 360 degrees. This actually shows up as an error in pitch more than an error in bank.

If you fly in a prolonged slip, as you suggested earlier, the gyro will eventually erect itself relative to gravity and show an error in bank until you fly coordinated for a few minutes and it can right itself as designed.

As for AHRS, I only fly that...I don't understand it.:dunno:
 
MauleSkinner said:
Mechanical attitude gyros have 4 vanes in them that are designed to pivot in response to gravity. If the gyro is tilted to one side relative to the pull of gravity, a vane will pivot and allow a jet of air to apply a correcting force (90 degrees from the direction the gyro needs to tilt) to erect the gyro.

A side effect to this is that centrifugal (yeah, I know...:rolleyes:) forces in a turn will cause the gyro to precess because the vanes sense "gravity" in some other direction than towards the center of the earth and try to "correct" the gyro. This precession will be worst at 180 degrees of turn, and will correct itself if the turn is continued to 360 degrees. This actually shows up as an error in pitch more than an error in bank.

If you fly in a prolonged slip, as you suggested earlier, the gyro will eventually erect itself relative to gravity and show an error in bank until you fly coordinated for a few minutes and it can right itself as designed.

As for AHRS, I only fly that...I don't understand it.:dunno:
Click to expand...

The guy who invented the solid state gyro is the guy who taught me to operate a 421, "Chooch", good guy, also made the first ELTs at the behest of he USAF and manufactured many antennas.

Basically what you have is "pitchforks" in the three axis in a housing that measures the interactions between the waves produced. As Gravity is introduced to the pitchforks, the wave pattern they propagate changes, and the sensors pick up this change in the pattern of wave interactions and interpret it to the corresponding vector changes. The first production use of it IIRC was in the Patriot Missile system.
 
It's all sorcery and magic.

Fly high enough and drag will begin to actually increase your speed. Stay that high and higher throttle will slow you down. Fly higher and none of it will matter anyway.
 
Let'sgoflying! said:
I have done an experiment with a typical 40yo turn coordinator.
Turn it off with the master (this one had a pullable breaker).
Wait 10 minutes.
Enter a std rate turn and allow the airplane to stabilize.
Turn the TC back on, and continue the turn til it spins up (5mins of turning) to be sure.
In my mind, the TC should think 'up' is about 15° 'off' the vertical following this.
Go about your flying.
In less than 5 minutes, it is reading correctly.
There is something in there which allows it to find the horizon, something I have not been taught yet.
Click to expand...
Neat experiment. But are you saying that you have no knowledge of the inner workings of flight instruments to which you may depend your life and that of your pax?
 
Richard said:
Neat experiment. But are you saying that you have no knowledge of the inner workings of flight instruments to which you may depend your life and that of your pax?
Click to expand...

It's not particularly relevant to survival to know 'how' it works, it's not like you're gonna fix it in flight. So long as you recognize it has failed through your scan, you'll be ok.
 
Richard said:
Neat experiment. But are you saying that you have no knowledge of the inner workings of flight instruments to which you may depend your life and that of your pax?
Click to expand...

That is a pretty good stretch. "no knowledge of the inner workings of flight instruments"

I didn't see where he said or insinuated that. :dunno:

Seems like a bit of an unfair assessment.
 
Henning said:
The guy who invented the solid state gyro is the guy who taught me to operate a 421, "Chooch", good guy, also made the first ELTs at the behest of he USAF and manufactured many antennas.

Basically what you have is "pitchforks" in the three axis in a housing that measures the interactions between the waves produced. As Gravity is introduced to the pitchforks, the wave pattern they propagate changes, and the sensors pick up this change in the pattern of wave interactions and interpret it to the corresponding vector changes. The first production use of it IIRC was in the Patriot Missile system.
Click to expand...

You should change your name to Jesus. Everywhere at every time for every thing.
 
Am I the only one that looks forward to opening 6PC's threads? Each one gives me a great laugh.
 
You must log in or register to reply here.
Back
Top