KAP140 behaviour - is this correct?

[peter-h]

Have a look here.

They are basically saying that the KAP140 will drive the trim servo in an uncontrolled manner as soon as the AP button is pressed, when on the ground.

The KFC225 certainly doesn't do that even though it is a similar system.

On mine, the only way to make the pitch trim servo run is to either subject the pitch servo to a "torque" (in flight, or by pushing/pulling on the yoke either on the ground or in flight for several seconds), or of course by operating the electric trim switch.

 

[Henning]

Strange that he couldn't overpower the AP. I haven't flown a plane with a KAP 140 where it wasn't either placquarded INOP or didn't make uncommanded major deviations randomly in flight ( last one was a 172 that would go hard left every now and again). People keep telling I should have an Autopilot, in 20+ years and 2500+ hrs of fling long days across the country, I'm not yet convinced where I get my $25ks worth out of it. I'd rather have. Kitfox as a spare plane than an AP for the same money, and I bet it would be cheaper to maintain as well.

 

[gprellwitz]

Henning, I've a couple hundred hours in a similar plane to the accident plane. The AAIB wasn't claiming that the pilot was unable to overpower the autopilot, but that the autopilot had run the trim to full nose up position prior to takeoff and that the pilot may have been unable to exert the needed force in a timely manner.

Moreover, tests on an intact aircraft indicated that an autopilot malfunction was not necessary in order to produce maximum trim. A simple process, which insidiously applied full trim with no associated audio or configuration warning, was demonstrated repeatedly. All that was required was one inadvertent press on one of the autopilot AP, HDG or ALT buttons after completing item 10 of the BEFORE TAKEOFF checklist. To some extent, the layout of the checklist and the relative positions of the engine and avionic controls contributed to the probability of inadvertent autopilot engagement. Whichever button was pressed, the autopilot would engage in a pitch mode that demanded level flight. If the pilot noticed that the autopilot was engaged before take-off and disengaged it, there was no checklist prompt to re-check the elevator trim position. If full nose-up trim had been set before take-off, the push force on the control wheel necessary to control the aircraft's tendency to pitch up at initial climb speed could have been too much for the pilot to exert and hold. There was no indication that the pilot had trimmed forward, but even if he had, the trim tab gearing was such that he would probably not have been able to re-trim in time to prevent the aircraft stalling, whether he used the MET or the trim wheel.

 

[peter-h]

What I don't understand is why the autopilot would trim UP and not down.

If you engage an autopilot during the ground roll, and then do the takeoff i.e. pull back, it would drive the trim DOWN.

But maybe the C182 installation from that era is different.

 

[Henning]

Henning, I've a couple hundred hours in a similar plane to the accident plane. The AAIB wasn't claiming that the pilot was unable to overpower the autopilot, but that the autopilot had run the trim to full nose up position prior to takeoff and that the pilot may have been unable to exert the needed force in a timely manner.

You can't stop the trim by grabbing the spinning wheel?

 

[TMetzinger]

Have a look here.

They are basically saying that the KAP140 will drive the trim servo in an uncontrolled manner as soon as the AP button is pressed, when on the ground.

The KFC225 certainly doesn't do that even though it is a similar system.

On mine, the only way to make the pitch trim servo run is to either subject the pitch servo to a "torque" (in flight, or by pushing/pulling on the yoke either on the ground or in flight for several seconds), or of course by operating the electric trim switch.

The KAP140s I've flown with don't drive the pitch trim unless they sense a force they need to trim away, on the ground or in flight. Nor does the documentation for the KAP140 suggest that this happens.

Now, perhaps there was an error/fault in the KAP140 in the accident airplane, but that's not the same thing.

 

[ronnieh]

.....

 

[TMetzinger]

I have no time with a 140 but a KFC 200 and a Collins 65 will do that. To check proper operation in either while on the ground, engage AP and pull yoke full aft and the trim should trim down, allow yoke to go full forward and it should trim up. This is normal operation and I suspect the same is true in the 140 or any AP with altitude hold. I am sure there is an AP expert out there that will confirm this.

That's true. But it's NOT uncontrolled trimming.

Related question. For autopilots with electric trim with the split switches, does everybody else test it the way the AFM supplement says (one side of the split switch engages the clutch, and the other runs the motor)? I got a funny look once from a CFI until I pulled out the documentation and explained why I was doing it that way. Sometimes I feel like I'm the only one who RTFM.

On another airplane I found that the "clutch switch" only worked in the forward position, so it was possible to trim the nose down but not up, and the guy who squawked it thought it was a problem with the motor. I put the clutch switch forward and the trim switch backwards and the trim ran nose-up. 20 minutes to resolder a broken wire on the switch rather than dig into the tail to look at the motor.

 

[ronnieh]

delete

 

[peter-h]

Why would the autopilot drive the pitch trim if there is no torque on the pitch servo?

 

[TMetzinger]

Yes on the split switch and check to see that captain's switch will overide FO.
Not sure what you mean by uncontrolled but, if you are taxiing out and not holding the yoke it will normally be full forward resulting in a trim up until it reaches the trim stop if left alone

The scenerio described sounds very likely to me. If the AP is engaged while taxxing or doing a run up or watever so that the yoke is full forward (elevator in the down position) on a KFC 200 and a Collins 65 it WILL result in full up trim if no intervention takes place. I suspect that is true in all AP's of 2 axis and above. Again I have no operational experience with the 140

In a typical GA single, the yoke will be neutral during taxi, not full forward. And the pitch trim relieves torque/pressure (as Peter notes), it doesn't care about absolute elevator position. The preflight test procedure (from memory so may not be perfect) is to engage the autopilot (which usually results in roll-only mode in a KAP140) and then engage the altitude hold mode, which will engage the pitch servo and "freeze" the yoke in position. Then apply yoke pressure forward and validate that the trim runs aft. Then apply pressure aft and validate that the trim runs forward.

Now some more sophisticated 2-axis autopilots will go into "attitude hold" mode when initially engaged (usually the ones with flight directors) and in those, the pitch servo is active when the AP is engaged. But I haven't seen ANY GA autopilot that runs the trim when the autopilot is engaged on the ground except in response to pressure on the yoke.

 

[Henning]

In a typical GA single, the yoke will be neutral during taxi, not full forward.

Most every plane I've flown unless I was holding the yoke back, default was full forward at taxi speeds with it coming to neutral at about 10 knots shy of flying speed.

 

[TMetzinger]

Most every plane I've flown unless I was holding the yoke back, default was full forward at taxi speeds with it coming to neutral at about 10 knots shy of flying speed.

Hmm... you might be right about that for Cessnas. Not in diamonds though.. I think there's a spring in there or something as you have to push the stick forward.

But I still stand by my statement that the autopilot doesn't respond to position, only to the forces it senses.

 

[peter-h]

The last 3 posts sum up what I thought - except that on the more "modern" autopilots just pressing the AP button gives you pitch and roll control mode; to get altitude hold you have to press ALT.

When taxiing, the yoke is free in both axes, obviously (except on spring loaded roll axis systems like the Cirrus). During takeoff, the airflow drives the yoke to a central-ish position, very quickly.

But none of this presents the pitch servo with a significant torque, so I can't see what would make the pitch servo run.

The report mentions pressing ALT to start with, but I don't see that would make any great difference to what should happen.

Possibly, if the AP or ALT button got pressed when one goes to full-throttle on takeoff, the autopilot would grab the yoke at that moment, and if let's say the yoke was fully back, the airflow "centralising" would present the pitch servo with a torque which would cause the trim to start running backwards, after a few seconds. But how long is the average takeoff run in a C182? Would it be long enough to end up with the pitch trim all the way back?

 

[ronnieh]

delete

 

[TMetzinger]

First, the KAP140 has no flight director (hence the KAP instead of the KFC).

Second I've flown the KFC 150 and KFC 250 and while those do have flight directors, they do NOT behave as you state. If you leave the yoke alone, activating the autopilot does NOT cause the trim to run in either direction. I don't recall if I've flown a KFC200 but I'm 99.9% certain that they do not try and move any control surfaces including trim unless you give them a goal.

Now, if you're on the ground, and you give the autopilot a 10000 ft target altitude and any sort of climb rate, THEN they'll move the trim, because the goal is a climb. But merely turning on an autopilot generally invokes a goal of "hold what you've got" in attitude, if it invokes a goal at all.

 

[Clark1961]

Kick this around all you want to but, first in most airplanes, all that I have flown, the default position of the yoke is full forward. Even at take off speed you will have to apply pressure to even get the elevator to neutral position.

Never flown a Cherokee or other stabilator equipped aircraft?

As for the second sentence, I'll call BS. If trimmed to neutral, the airflow will push the elevator to the neutral.

 

[ronnieh]

You are all correct. You would think I would learn by now that facts just are not appreciated here. Even without a FD the computer will sync and want the elevator in neutral, the computer syncs the FD, not the other way around. Heck don't take my word for it we have multi hundred hour pilots and students, many students with years of experience tens of hours of experience. Many of these pilots have read about AP's in Private Pilot magazine, thus making them experts. I defer to them.

The few thousand hours I have with King auto pilots soon slip the memory of this old and tired pilot. I agree the auto trim feature of the King series of AP up through the KFC 300 does not work on the ground, while taxiing or where ever else you say.

I sincerely apologise for interjecting facts into a thread. It will not happen again.

 

[Clark1961]

You are all correct. You would think I would learn by now that facts just are not appreciated here. Even without a FD the computer will sync and want the elevator in neutral, the computer syncs the FD, not the other way around. Heck don't take my word for it we have multi hundred hour pilots and students, many students with years of experience tens of hours of experience. Many of these pilots have read about AP's in Private Pilot magazine, thus making them experts. I defer to them.

The few thousand hours I have with King auto pilots soon slip the memory of this old and tired pilot. I agree the auto trim feature of the King series of AP up through the KFC 300 does not work on the ground, while taxiing or where ever else you say.

I sincerely apologise for interjecting facts into a thread. It will not happen again.

Going bonkers doesn't help...

 

[TMetzinger]

You are all correct. You would think I would learn by now that facts just are not appreciated here. Even without a FD the computer will sync and want the elevator in neutral, the computer syncs the FD, not the other way around. Heck don't take my word for it we have multi hundred hour pilots and students, many students with years of experience tens of hours of experience. Many of these pilots have read about AP's in Private Pilot magazine, thus making them experts. I defer to them.

The few thousand hours I have with King auto pilots soon slip the memory of this old and tired pilot. I agree the auto trim feature of the King series of AP up through the KFC 300 does not work on the ground, while taxiing or where ever else you say.

I sincerely apologise for interjecting facts into a thread. It will not happen again.

OK... Seriously, if you've got the AFM supplement that describes how the trim just "runs" when you engage the AP on the ground, please quote it.

I just pulled the supplement from the KAP140 2 axis autopilot in the C172S.
Turning on the AP engages the roll mode to wings level and pitch to VS (whatever the VS is at the moment - on the ground it should be zero). If the pitch trim runs for more than 5 seconds there is an aural "Trim in Motion" cue.

During the pre-flight checks, one tests the manual electric pitch trim, and then engages the autopilot. There is no mention of the trim running when the autopilot is engaged. All the pilot does is verify that he can override the autopilot in roll and in pitch.

There's also no mention of any sensor to detect elevator position in the system description.

We don't need to turn this into a credential contest. The autopilot doesnt care how many hours (or thousands) we have when we press the button. I'll be going out to fly a DA40 with a KAP140 later this month and will videotape the autopilot self-test and the preflight checks.


I did some looking on the KFC300, which has the following note in the Pilot Guide.:

CAUTION:

Overpowering the Autopilot in the pitch axis in flight for periods of 3

seconds or more will result in the autotrim system operating in the direction to
oppose the pilot and will, therefore, cause an increase in the pitch overpower forces,
and ​

​

if Autopilot is disengaged, will result in a pitch transient control force. Operation​

of the Autopilot on the ground may cause the autotrim to run because of
backforce generated by elevator downsprings or pilot induced forces.​

​

​

Note that the autotrim "may" run, not "will" run, and runs in response to forces induced by springs or the pilot on the elevator postion, not in response to the position itself. So I still think that even with the KFC300, if you are sitting motionless on the ground and engage the FD and AP that the trim will not run UNLESS you move the pitch control.

​

 

[flyingron]

Engaging the STEC-55X in ALT ground and holding even moderate pressure on the yoke will spin the trim wheel. After doing the autopilot checkout checklist from the flight manual, unless I'm careful to neutralize the wheel in the checks will result in the trim being set way off from the usual takeoff position (which in the Navion is the same as the landing position if you'd had properly trimmed her on short final). I always check that after running the autopilot on the ground.

 

[TMetzinger]

Ron, if you put no pressure on the yoke, does the trim wheel spin? I'm guessing that it doesnt.

 

[flyingron]

Ron, if you put no pressure on the yoke, does the trim wheel spin? I'm guessing that it doesnt.

Yes in certain cases it does. The weight of the elevator without any airload on tricks the thing into wanting to trim nose up. If I hold it in a roughly neutral position it will stop.

 

[TMetzinger]

OK the test I'm going to perform, and would like others to try is:

1. With the airplane idling and stopped and parking brake on, take hands off controls.
2. Engage the autopilot in roll (wings level) and altitude hold mode.
3. Don't touch anything.
4. Report if the trim wheel moves.

I predict that with the KAP140 and the GFC700 the trim wheel will not move unless the load on the elevator changes.

 

[flyingron]

Tim...get behind the wheel of an idling 182 and let go of the wheel and tell me what happens to the yoke. I bet it's up near the front of the panel if not on the stops. You engage altitude hold the thing will move the wheel back to you but it is countering the weight of the controls commanding nose up trim (which has no effect on the ground).

 

[TMetzinger]

In a 182 if I engage the autopilot in alt hold mode with the yoke full forward.. the trim doesn't move, nor does the wheel. Only if I try to pull the yoke back or increase the throttle (which "blows" the yoke back) does anything happen, and the trim at that point runs nose down to compensate for my pulling the elevator nose up.

Now, If I am HOLDING the yoke in neutral, THEN engage the AP, and then LET GO of the yoke, the autopilot senses the weight of the elevator as nose-down force, and runs the trim nose-up to compensate (and runs out of trim).

 

[gprellwitz]

OK the test I'm going to perform, and would like others to try is:

1. With the airplane idling and stopped and parking brake on, take hands off controls.
2. Engage the autopilot in roll (wings level) and altitude hold mode.
3. Don't touch anything.
4. Report if the trim wheel moves.

I predict that with the KAP140 and the GFC700 the trim wheel will not move unless the load on the elevator changes.

Just a note that a 1998 182S like in the original post does not have a G1000 or a GFC700. I'll try it on the 1999 182S with factory stack next time I get a chance. I honestly don't remember under what circumstances it will spin the pitch trim on the ground.

 

[TMetzinger]

Just a note that a 1998 182S like in the original post does not have a G1000 or a GFC700. I'll try it on the 1999 182S with factory stack next time I get a chance. I honestly don't remember under what circumstances it will spin the pitch trim on the ground.

Correct - it should have a KAP140 if it came with one from the factory.

 

[gprellwitz]

Correct - it should have a KAP140 if it came with one from the factory.

It does and it did .

I think I misunderstood your earlier post. I read it as if you expected the plane to have both a KAP140 and a GFC700, and those two are pretty much mutually exclusive AFAIK. There were some Cessna birds that shipped with a G1000 and the KAP140, but then they switched to the integrated GFC700, a much nicer solution.

 

[TMetzinger]

It does and it did .

I think I misunderstood your earlier post. I read it as if you expected the plane to have both a KAP140 and a GFC700, and those two are pretty much mutually exclusive AFAIK. There were some Cessna birds that shipped with a G1000 and the KAP140, but then they switched to the integrated GFC700, a much nicer solution.

To quote Vincent... Correctamundo!

The KAP140 was a real bandaid for the early G1000 airplanes. Set the altimeter three times, please, on the G1000, the standby, and the KAP140. Set the altitude bug twice on the G1000 and the KAP140. The only thing that WAS interconnected was the heading bug and the CDI.

Still, the KAP140 is a fine autopilot, it's just that the full garmin is a whole order of magnitude better. Anything that puts a TO/GA button in a single-engine propeller-driven airplane is by definition awesome :wink2:.

 

[denverpilot]

One of the CAP birds has the King here and it's interesting to compare and contrast it with the GFC.

 

[peter-h]

I did a quick test yesterday on my KFC225.

Pressing ALT only does nothing with the servos; it just engages the flight director in ALT hold mode.

The only way I could get the trim to run on the ground was as follows: when the autopilot is engaged with servos (pressing the AP button) the pitch servo is driven very slowly, and eventually reaches the stop, and at that point the trim servo is actuated.

 

[TMetzinger]

I did a quick test yesterday on my KFC225.

Pressing ALT only does nothing with the servos; it just engages the flight director in ALT hold mode.

The only way I could get the trim to run on the ground was as follows: when the autopilot is engaged with servos (pressing the AP button) the pitch servo is driven very slowly, and eventually reaches the stop, and at that point the trim servo is actuated.

Was the AI spun up when you did your test?

 

[peter-h]

That's a good question.

It was definitely erect, because if it isn't the AP won't start up.

My KFC225 has no "vacuum valid" input (I have checked) and no way of knowing if the engine is running, and no "gear up" etc input.

 

[TMetzinger]

Thanks Peter. I just wondered why it was trying to increase pitch if in altitude hold mode. Were the command bars on the FD higher than the pitch attitude? That's the only reason I can think for it trying to increase pitch.

In the KAP140, where there is no pitch attitude sensor (it's rate driven), the only way the autopilot knows what pitch is is by measuring the change in baro pressure. And if that's not changing, the autopilot should assume you're level, pitch wise.

 

[peter-h]

I agree but there is always some small drift in these systems.

For example of you accidentally select VS=0 instead of altitude hold you get a slow creep up or down.

In fact the KFC225 had a nasty little bug in that if you pressed some buttons the wrong way it was in VS=0 instead of ALT hold, but displaying ALT. They fixed it c. 2003, under the guise of some innocent SB

 

[TMetzinger]

Yes the in some ways the KAP140 (which if I recall correctly is digital) is more "stable" than older analog units, particularly the attitude-based ones.

OK I'll report back in a couple of weeks after my tests withe KAP140.