What's new
Van's Air Force

Don't miss anything! Register now for full access to the definitive RV support community.

Pan, pan

Ed_Wischmeyer

Well Known Member
?Approach, experimental XYZ, pan, pan.? That got their attention!

I?d used mayday two or three times in the past, but this was the first time using pan. I was flying with a friend, and when he gave me the controls back in the RV-9A after doing a turn, the stick was really heavy. The elevator trim indicator showed way, way nose down ? uncommanded. And the elevator trim switch was ineffective. After talking it over with my friend, also a very experienced CFI, we decided this was an appropriate PAN situation and going home right away was the best course.

We were about 8 minutes out, about 120 knots, and what to do? ATC offered closer airports, but home base seemed pretty doable. But it took a LOT of pull on the stick.

The first decision was that I had him fly so I wouldn?t be landing an out of trim airplane with really tired arms.

The second decision was what speed to fly. With nose down trim, flying faster would probably have meant less stick force, but the decision I made was that the plane was controllable as it was, so let?s not screw with anything we don?t have to. We didn?t discuss this, however, but he didn?t suggest anything different. We did discuss that this would be a no flap landing to avoid any more nose down pitching tendency.

What was a bit of a surprise to me was that when I wasn?t concentrating on a two handed pull on the stick, I suddenly had a lot more relaxed brain cycles to deal with the situation. On my plane, the autopilot pitch servo changes the speed of the electric pitch trim, depending on airspeed. With the autopilot pitch servo powered off, the elevator trim switch goes directly to the trim servo. Tried that and it restored normal trim function, so I told approach control that we had resolved the situation. But I did not explicitly cancel the pan.

Meanwhile, ATC had apparently escalated the PAN to a Mayday for us as we found out when tower referred to us as the emergency aircraft. I told them that we had canceled the emergency, we landed, put the plane away, debriefed and took a break.

The decisions made:
* Go back to the airport right away;
* Pass up the closer airports as the plane was flyable and we had two pilots. An okay decision, maybe not the best. There was an Army air field we flew over with a long runway, but the situation was resolved by then. A short runway would not have been a good choice. But if I had been flying solo way out in the boonies, I might have chosen to land it wherever before my arms gave out;
* Discuss the situation with the other pilot;
* Declare a Pan;
* Get the other pilot to fly the plane for a bit.

But there were good decisions made in advance:
* The autopilot servos were on a circuit breaker switch and were easy to find as they were at the end of that group of switches. I didn?t have to read any label to know which switch to flip;
* I?d spent lots of time learning the details of this installation and how all the parts interconnected and worked. I knew everything that pertained to the elevator trim without having to look anything up;

When we upgraded the avionics, the elevator trim stayed on its own circuit breaker in the middle of a row, supplying power to the electric trim switch. It?s easy enough to find, but not nearly as easy as the autopilot servos. If I were designing a panel from scratch, I?d have disconnect switches all next to each other as circuit breaker toggle switches.

The last step was to call the tower and thank them for good handling. A five second call, but they appreciated it. And no paperwork to fill out.
 
AP Engaged on accident?

Ed,
Any possibility you or your passenger had bumped the autopilot engage? If that was the case, then as your co-pilot continued to pull back pressure in the turn, the autopilot may have mistaken it for an out of trim aircraft, and tried to autotrim for you (driving the trim servo to a nose down position).

I had a similar experience in a G3XT equipped RV where the AP engaged (without my knowledge) during slow flight. The AP ran the trim all the way forward, which I hardly noticed while flying slow. However, as soon as I pointed the nose down to build airspeed, whooaah that nose-down trim got my attention quick :eek: .
Fortunately, altitude was plentiful enough to recognize, re-trim, and avoid lawn darting, but I came away from that a firm supporter of aural annunciation for autonomous flight systems!
 
Got a very prompt reply from Garmin in the wee hours of the morning. Basically, pilot error... the system worked as designed, but I didn't know all the details and didn't recognize what was going on. Extra credit to Garmin on this one, stay after school for me.

"The event I believe you are referencing starts at row 1909 in the datalog. It is at this point that you had ESP engage because your airspeed fell below your configured ESP target. ESP for the next 10 seconds proceeded to push on the stick at up to 66% torque trying to get you back within your configured speed envelope. After 10 seconds, the system engaged the autopilot in level mode. This is a configuration option to engage level mode out of ESP. The idea is that if the pilot is incapacitated and the airplane is starting to lose control, it will automatically go to level mode after 10 seconds. Once the autopilot engaged in level mode, the pilot continued to pull on the stick. The aircraft because it can't tell you are pulling on the stick starts trimming nose down. This continues until it hits the down trim stop. At the 1945 line in the datalog, someone hit the manual trim button. I believe it is at this point you had taken back over the airplane. The AP disconnect tone would have played. [My comment: if it did sound, I didn't hear it. Then again, in times of stress, aural cues are among the first sensory inputs dropped.]

"I cannot explain why it appeared that the trim was not working until you pulled the autopilot breaker. However, with speed scheduled trim (which you have configured), it is going to run the trim slower than it would run once you pulled power on the autopilot. My theory is that it takes a long time to move from full down to a more neutral position and it appeared as if it wasn't working because it was slowed down to the speed scheduling.

"I would classify this event as pilot attempting to override autopilot vs runaway trim. I do not believe there to be any issues with your aircraft. I would recommend that if you are intentionally planning to fly outside of the ESP envelope that you disable ESP. This can be done quickly using the onscreen AFCS mode controller. You could also consider turning off the autopilot engagement out of ESP if you do not like that feature."
 
Last edited:
Ed, thanks for posting this. I've been thinking about enabling ESP feature and it's good to know that it works as it was intended, the pilot just has to recognize that.
In the 20 years of that I've been flying, I've only had an autopilot for the past 2 year and one of the things that I've changed in my checklist is both take off and landing checklists I have the Autopilot switch to off position. For me, the switch remains off at or below 500 ft. AGL.
 
Buttonology training required - Thanks Ed

Thanks for posting, Ed. As I embarked on renewing my IFR skills, 55 hours since first flight, I found operating the panel to be a very steep learning curve. I noted your 100 hrs of learning required from another post.

This is a good reminder that some training to really learn the buttonology and all the details of AP operation are required of ALL G3X pilots not only the IFR. I have spent many hours studying the G3X AP section then making some check flights to test how it acts and understand all the 5 levels of nuances (it seems like that many) for what it does and how. Only to find something new and go learn some more. I thought it was just me.

Recap:
Lesson #1 - Know the equipment operation inside and out, then test it
Lesson #2 - Look at the data files to see what was happening

Note: I even had my pitch-roll servo wiring swapped. Excitement.
 
I also learned this lesson about auto-trim the hard way. While doing some IFR work with AP engaged CFI spots some traffic, takes control for a brief time without disengaging the AP. He pulled up, the AP trimmed down and then the "hey whats going on here with all the stick pressure". It also appeared at the time that the trim wasn't working when trying to re-trim but I'm pretty sure it was just the slow trim speed made it seem like it wasn't working.

I talked to Garmin when this happened with basically the same result that there was no problem and it was "pilot error".
I know we don't have any stick sensors but it would be nice if somehow the AP could know the difference between pilot stick input vs out of trim input from the elevator so it would auto disconnect with pilot input.
 
Ed- I’m grateful for the benign outcome after your experience. Your story made me cringe and immediatly brought to mind the rather light duty(re:lightweight) plastic linkages used between the actuator and the trim control tab on many RV’s. A real concern with any form of runaway down trim is the possibility of abrupt failure of these linkages at full deflection, which can(and I believe in at least one instance, may have done) fold the wings on an RV. This is one reason it is important that travel be limited within the actuator and not by external physical stops.

Two responses worth training oneself for in this situation are 1- slow down to a speed well below Va and 2- lock both elbows to the sides of your body and keep fists tight when exerting abnormal aft pressure on the stick for sustained periods. This can help limit aft travel of the stick should the force you are fighting suddenly disappear.

One of the best designs I’ve seen for elevator trim actuation, though I’m not sure how it would be adapted to RVs, has a reversible motor and cycles continuously through the full range of up and down trim in a stuck switch or runaway condition so that the pilot can wait for a relatively neutral point trim point to come around and pull the breaker there.

Anyhow, congratulations on your thoughtful handling of the situation!- Otis
 
Last edited:
We are always here to help! If you have questions about how the G3X system works, please contact us!

One key takeaway here is that if you configure a setting on your aircraft, please make sure you understand how it works. ESP and AFCS engagement from ESP are both features that are not on by default. If you bought the aircraft or if you have a shop do the install, it is prudent to sit down and understand how your system is configured.

If you let someone else fly or even while you are flying, please pay attention to your system. If there are flashing items and audio tones, they are trying to get your attention.

While MCAS and ESP are both designed to keep the aircraft within its designed envelope, they fundamentally operate differently. NOTE: ESP WILL NOT trim the airplane. ESP is also not a stick pusher and should able to be easily overcome by the pilot.

If you would like to know more about ESP. Section 8.12 of the G3X Touch pilot's guide has more information on ESP. You can download the Pilot's guide from our website: https://support.garmin.com/support/manuals/manuals.htm?partNo=010-00G3X-00

Thanks,
Levi Self

Team X Project Manager
 
Last edited:
Is it just me or does this sound like something Airbus or Boeing would build into their system?

Unlike MCAS on the 737 Max, Garmin detailed how ESP works in the manual.

Plus it?s easy to install a switch to a discrete input on the G3X system to disable ESP. We did so on our RV-7A and labeled it upside down ?FUN? to be used for aerobatics.
 
After recreating the event on a later flight...

A few more thoughts on the matter:
* The reason why we didn't hear the autopilot dis/connect tones on was likely because the stall warning was going on the whole time. At least, that's what was observed on a later flight when I tried to recreate the incident and confirmed the aural warnings by video. I'm planning another flight to excite ESP with bank angle, not low speed, so I can experience the autopilot aurals with the stall warning silent;
* When I recreated the incident on a later flight and the autopilot (as designed) put the elevator trim to full nose down, it was natural to assume that the additional stick pressure required was the normal pressure required for that flight maneuver -- an interesting human factors phenomenon and not a systems problem. Not sure I've ever read about that one;
* Regarding power to the elevator trim motor: in my installation, the trim circuit breaker provides power to the position indicator and to the pilot-controlled trim switch. As mentioned above, the pitch trim servo, if it has power, can vary the trim speed based on airspeed. But if the autopilot is flying the plane, the pitch servo can run the elevator trim, even if there is no power to the pilot's trim switch. (I pulled the trim circuit breaker to verify this in flight.) In designing cockpit switchology, it's important to know exactly what power sources affect each failure mode. Maybe folks with Dynon and TruTrak can comment on whether those systems work the same way;
* Since we were maneuvering entirely by outside references (I was demonstrating maneuvers to another pilot and critiquing his efforts), visual indications inside the cockpit were of little value. That's undoubtedly why the aural alerts were designed in (Check FAR 23.207(b) for a similar point of view);
* Not that we thought about it at the time, but slowing down might have increased the back pressure required. Might try that on a later flight;

There's always more to learn about flying and about airplanes, more so if you're interested in human factors. And if your focus is safety, there's even more to learn.
 
Last edited:
AP Disconnect button function

In the G3X and G5 autopilots with the GSA 28 servos, the AP disconnect will keep the servo motor from running and will also keep the servo from driving trim. If the servo is to ever do something that you do not expect, you should press and hold the AP disconnect button until you have resolved the situation. The AP disconnect is a hardware disconnect and does not rely on software to disconnect the autopilot or to stop running trim.

If you do hold down the AP disconnect, be aware that it could engage your autopilot upon release of this button. The G3X can be configured to engage the AP from Control Wheel Steering (CWS). If you have this configured and the intent of holding the button was to stop the servo from trimming or other AP function, you would need to disconnect the autopilot if it engages upon releasing the AP disc/CWS button.

If you have auto trim, you should have a checklist for how to handle runaway trim which would include pressing and holding AP disconnect until you can pull power from the servo/trim circuits.

As always, our email address is below. We are happy to answer any questions you have.

Thanks,
Levi Self

Team X Project Manager
 
Back
Top