by Michael “Mick” Kaufman
In my previous column, I again addressed the issue of “icing,” as did Greg Gorak in his excellent article. It was unfortunate that Wisconsin got to see another fatal GA icing accident since that publication.
In this issue, I will cover “autopilot glitches, failures and anomalies”– what to do in each case and why some of those things happen. Also in this issue, I will touch on the new Avidyne IFD-540 GPS navigator based on some recent flights, and do a full review in a future issue of Midwest Flyer Magazine.
Without repeating myself like a defective vinyl record, I need to say that if you should get into icing conditions of any magnitude, report it to air traffic control (ATC) without delay, and have a plan to get out of it ASAP. That plan may be a climb to VFR conditions on top or a descent to a lower altitude, if the air is warmer below as per Greg’s article. I have been in similar conditions myself, and my lack of being assertive with ATC almost ended in disaster. I would like to mention that Greg Gorak was one of my instructors and mentors in the late 1960s, and I consider him one of the top aviation educators of our time.
Over the years, which I have been giving instrument and proficiency training, about one out of four pilots complain or have a comment on their autopilot. The manufacturers of autopilots have the responsibility to provide as many fail-safe features as they can in order to avoid crashing the airplane. One way to do that would be to make a heading and altitude hold autopilot only, but pilots and aircraft owners ask for much more. The autopilot in my Bonanza is an S-Tec 50, and it does just that and does it well. It will not descend or climb with altitude preselect or intercept and fly a radial or glideslope or create a top of descent profile (TOD). Once I have established my aircraft on a VOR radial or GPS course, the autopilot will track it but nothing more.
I hand fly all of my transitions and approaches. When I made the decision to install a new autopilot in my aircraft more than 10 years ago, I had the option of going with a more sophisticated autopilot, but declined because it would cause me to become lazy and lose my proficiency in hand flying approaches.
Ask yourself this question: If I lost my autopilot in hard instrument meteorological conditions (IMC), could I safely fly an approach to minimums? If you answer “No” to that question, go and see your local certified instrument flight instructor (CFII), or come and see me and get proficient again.
Why am I so insistent that we be able to hand-fly approaches as these sophisticated autopilots rarely fail? The answer is, there are safety features built into autopilots to intentionally keep them from flying an improperly established approach or capturing a false glideslope. I will cover this later in this article.
If I were to try to list the issues in order of occurrence of mis-coupled approaches, they would be as follows:
Number one would be “pilot error. “
Number two, “a bad vector. “
Number three, a true “autopilot problem” or “bad installation” or “set up” by the avionics shop.
The first issue of pilot error, and the most common problem, is not switching from GPSS mode to approach mode soon enough (call it buttonology). It is my recommendation when flying an approach, especially a full approach with a course reversal, to let the GPSS feature get you established inbound and then manually make the switch to NAV or APP mode, depending on the autopilot.
There is a feature on the world’s most popular navigator, the Garmin 430/530, to switch automatically from GPS to VOR/Localizer (VLOC) mode once established inbound on a non-GPS approach. The pilot can go to “preferences” and change this feature, but most units I see have this feature enabled.
Many pilots are not aware that when there is a change from GPS course guidance to VLOC, the autopilot will continue to fly the digital output signal coming from the navigator. If the pilot fails to switch to the NAV or APP mode on the autopilot, the system appears to work well until the final approach fix.
In the early days of GPSS, many pilots elected to fly the GPSS course, especially on Localizer-only approaches, which was not legal or the intent of the system. Garmin made a firmware update to stop the digital output on their navigators at the final approach fix on non-GPS approaches. The digital output will resume at the missed approach point to enable GPSS again while flying the missed approach procedures. When we wait too long to switch our autopilots to the NAV or APP mode or we get a bad vector, we have other issues to deal with.
If my memory is on track from previous research, the Air Traffic Controllers’ Handbook states that a pilot must be vectored to intercept the final approach course a minimum of 2 miles outside the Final Approach Fix (FAF), or the pilot optionally may decline the approach. This is due to the fact that certain autopilots may not have adequate time to respond properly and couple.
To explain some of these conditions, we need to remember that most of the autopilots were designed before we were flying in the precise GPS environment that we fly in today. If you refer to the Aeronautical Information Manual (AIM), there is a mention of false localizer and glideslopes, which few people are aware of today because GPS navigation routes you to get established on the approach.
Fig 1
One of my favorite demonstrations of yesteryear was flying the VOR transition from the Stevens Point, Wisconsin VOR (STE) to the ILS 08 approach to Central Wisconsin Airport in Mosinee, Wis. (KCWA) (Fig 1). Once established on the 321-degree radial of the STE VOR (NAV-2), we begin to look for the localizer needle on (NAV-1) to come alive. When we see the navigation needle come off the peg, we know we need to turn to the outbound localizer heading of approximately 251 degrees to do the procedure turn. By not confirming our position with the outer marker or NDB, chances are we have intercepted a false localizer, which will take us outbound for a short distance and then finally disappear into infinity.
Each autopilot and navigator has different safety circuits to prevent autopilot missed coupling on approaches, and I will highlight these later. There are so many different autopilots in use in General Aviation that it would be difficult to cover every one in the scope of this article. If you have any questions on a specific model of autopilot, I would be happy to answer these via email or telephone.
To generalize, King autopilots must see a centered localizer or GPS indicator before the glideslope centers for capture. Century and S-Tec autopilots must see an active needle, and it takes time (usually between 15 and 40 seconds depending on model) before the glideslope will couple. A pilot should check the autopilot manual for specific requirements for doing an approach. I have seen some very unusual requirements on some autopilots for proper approach coupling. I found a very unusual autopilot while trying to do coupled approaches in an Epic. The autopilot would not capture the glideslope if the button pushing to do the approach and altitude hold were pushed in the wrong sequence.
When you are on an approach and the weather is bad and you have some ice on the airplane, it is no time to learn how to use your autopilot or experiment. Rather than try to fix or adjust the autopilot, it is my recommendation to disconnect it and hand fly the approach, but only if the approach is salvageable.
Remember that the required procedure is to make a missed approach any time that either the localizer/GPS indicator or glideslope indicator should peg. On some autopilots, pressing the ALT or GS button multiple times will cause the autopilot to capture a glideslope that is already below the center. I cannot recommend this procedure as I have seen some rather steep descents to catch the glideslope, which result in an unstabilized approach.
The third cause of autopilot glitches deals with failure or improper installation, set up, or repair.
The most common autopilot problem usually involves brushes in the servomotors. The symptom for this is overcorrection while making minor corrections in heading or altitude. The cure is to have the servo cleaned and new brushes installed. This is a minor fix if your avionics technician is willing to do it, but if he insists on sending it into the factory, the bill will frighten you.
Avionics technicians have a tough job trying to get all of the devices in an aircraft to talk to each other. This combination may have never been done before in your aircraft. There are so many different navigators and autopilots, and most of the technicians are not pilots who can test fly the aircraft upon completion of a complicated install and may not have any idea of what to expect in performance.
Many technicians will not get in an aircraft with a pilot aircraft owner, and I can understand that as well. Because of this, I find many avionics installs that are not done right, and the aircraft owner does not know the difference. It may be a year or more before some knowledgeable pilot detects the problem.
One case I remember was having two GPSS units installed in the aircraft that were fighting each other to see which one would win. After disabling one of the units, the autopilot flew fine.
My recommended solution would be to include an in-flight checkout of the pilot/owner and equipment with the price of every major avionics installation by a flight instructor who is familiar with the expected performance.
Avidyne IFD-540 Navigator
Last month, I had the privilege of using an Avidyne IFD-540 navigator in an F33 Bonanza on a flight from Tucson, Ariz. to Edwards AFB in Calif. I plan on doing a complete review of this unit in my column in a future issue of Midwest Flyer Magazine, but I can tell you I have a very favorable first impression of the equipment.
The Avidyne IFD540 (and IFD440) FMS/GPS/NAV/COMs combine the best flight management system available, with a multi-channel digital VHF radio, in an easy-to-use plug and play design. Each provides VHF communication and SBAS/LPV precision navigation and are designed to meet the accuracy and integrity requirements for ADS-B as part of the NextGen airspace initiative.
The installation of the IFD540 in an aircraft that had a Garmin 530 previously takes about 2 hours. This saves thousands of dollars in labor. It is a slide-in replacement, and all of the devices that were connected to the 530 will work once the installer programs the box to recognize that equipment.
I find that the IFD540 is easy to learn to operate, especially for someone with Garmin GNS 480 experience. Victor Airways are included making flights on airways easy to program. You do not need to enter every waypoint along the airway. Every programming function can be accomplished by two methods, touch screen or a knob and button combination. This way the pilot can program a route on the ground using the touch screen, and use the knob and buttons while in flight when turbulence makes the touch screen method too difficult to use. Rumor has it that you will be able to communicate with the navigator and upload flight plans from Foreflight on your tablet via Bluetooth or WiFi. Look for a full review coming soon!!
I am looking forward to spring and better flying weather; I will gladly trade in-flight-icing for dodging thunderstorms. At least I know where the thunderstorms are and how to avoid them.
EDITOR’S NOTE: Michael J. “Mick” Kaufman is a Certified Instrument Flight Instructor (CFII) and the program manager of flight operations with “Bonanza/Baron Pilot Training,” operating out of Lone Rock (LNR) and Eagle River (EGV), Wisconsin. Kaufman was named “FAA’s Safety Team Representative of the Year for Wisconsin” in 2008. Email questions to captmick@me.com or call 817-988-0174.
