Avionics Troubleshooting 101

Common diagnostic tasks to consider when things stop working.

1

When I was a green avionics tech in the early 1990s, the bible in my toolbox was Jerry Gordon’s book Help! My Radio Quit. Published in 1988, the book was a guide written for pilots on how to troubleshoot failed avionics before hollering for help at the local avionics shop. Man were things simple back then in the analog world of Narco Nav 11s, King KN62 DMEs and Terra transponders. Now, not so simple.

Avionics bootcampThese days, our digital avionics suites are loaded with critical data buses, advanced operating software (that may or may not be compatible with third-party brands) and, in some cases, made up of multiple components spread throughout the airframe and connected by a high-speed digital bus. Eventually, something stops working—either a hard failure, consistent failure or one that’s intermittent and very hard to troubleshoot—and you’ll find yourself playing the role of an avionics tech. Maybe you’ve just finished the kit, or maybe it’s an old kit with some miles on it. When it comes to self-troubleshooting, think logically, and don’t dismiss the fact that you may be better off sourcing some hired help from an avionics shop for certain kinds of failures.

Installing used equipment? Team with a trusted avionics shop before going through the effort to install them. Turns out that the mechanical altimeter here had excessive friction and couldn’t be certified.

Troubleshoot Like a Pro

A big part of quickly fixing a problem is having the right mindset, and not everyone has it. You could be an amazingly talented installer but a lousy troubleshooter. Regardless of your skill, the key to any successful avionics diagnosis is having the current installation manual (or the revision that best represents your interface). Hopefully, you or whoever built the machine provided clear and accurate diagrams of the interface. Component installation manuals are only part of the required reference material needed for a practical problem chase—airframe-specific wiring prints (and well-labeled connectors) keep the troubleshooting task on the rails and headed in the right direction. If you built the aircraft and did the avionics, you should have a huge advantage. It’s your work (pro or hack), right?

In addition to a decent multimeter, you’ll need reliable ground power. Whether it’s loading new software or wringing out wiring bundles, troubleshooting is a sure way of flattening the aircraft’s battery. Plus, with a good ground power supply, you can supply the system the precise operating voltage it needs to work best. And there’s a piece of troubleshooting advice right there—keep the charging system healthy. Digital avionics need uninterrupted and stable input voltage. Always keep the battery fresh and alternator strong. Most avionics are happy with anything between 13.4 and 14.2 volts (in a 14-volt system), but you should know the operating voltage range of the equipment and at what threshold it will shut down. You’ll find it in the manual.

Lay harnesses in the panel so you can easily work with them should you have to troubleshoot. Make certain there is enough clearance from the controls. Large bundles can jam them.

A bit on the workspace. It should be obvious, but I’ve seen some ugly work areas. (For tips on how the pros do it right, see the KITPLANES® Metal Magic video series). Troubleshoot in a clean and organized area—away from noise, temperature extremes and other distractions. If you’re working on a heading system, be aware of moving aircraft that can cause interference in the general area. Keep your smartphone or tablet charged, especially if you’re configuring or sending data wirelessly into the system.

Airshow coverage sponsor:
As strong as Garmin’s backplates are, it’s easy for a pin to bend or push back in the connector body.

It’s a big help to have someone available to help. It’s a lot easier wringing out wiring harnesses with two people. If they aren’t skilled in avionics work, tell them to make sure the multimeter’s measuring probe is firmly seated on the connector pin. Got a magnifying glass? You’ll need a couple. If you’ve ever tried to read the pin designation etched into a connector body, you’ll understand and appreciate good work lights. Don’t injure yourself. You may be a contortionist—lying on your twisted back and working upside down in tight areas. It’s a reason many avionics techs call it quits early in their career. The work takes a toll on the body.

Don’t attempt troubleshooting without having solid, stable voltage on the aircraft through an external supply (left). It’s not practical to invest in pricey bench scopes and test equipment (right), so establish a relationship with a trusted shop before you even start the project.

Tips From the Pros

Herewith are some of my top-pick basic troubleshooting tasks that the typical owner pilot should be able to handle. Let’s start at the top of the radio stack with audio systems.

I recently got a phone call from a friend sitting in the run-up area with his airplane, unable to transmit to the tower for a takeoff clearance. The aircraft had a standalone intercom with a King audio panel, and he could key the transmitter—he could hear the click and see the intercom transmit light, but there was no modulation. As with many audio systems, there was an auxiliary mic jack used for a hand microphone. Do not keep the hand microphone plugged in. Better to hang it on the clip unplugged or stuff it in a map pocket—keeping it plugged in increases the chances of problems. You or someone else might climb into the cockpit and catch a foot on it—which is what likely happened in my friend’s airplane because the mic’s cable was stressed (frayed) at the connector, and it was shorting the key line. Resist keeping headsets plugged into audio jacks when they aren’t in use. It’s trouble waiting to happen. (One exception might be where the jacks are well out of the way, and it’s unlikely you or your passenger will kick the wires getting in and out.)

Fabricate quick-disconnect harnesses for switches and annunciators. It will make troubleshooting (and switch replacement) easier.

Additionally, many audio installs have a set of emergency audio jacks that bypass the intercom. Plugging into them connects you directly to the primary radio. Remember, if you are using headsets, you won’t hear sidetone or your own voice as you transmit because the sidetone is generated in the intercom.

Other things to watch out for are failed push-to-talk switches (I always keep a portable switch in my flight bag, and it’s bailed me out on more than one occasion) and also stressed wires within the control stick. When routing the PTT line through the stick, always provide for chafe protection because every time you move the stick, you are putting at least some stress on the wiring. When I wire a PTT switch, I always include a quick disconnect connector that’s easy to access. If the line gets shorted, you could always unplug the connector. When possible, install a second PTT switch but remember to plug it into the set of jacks connected to it.

Consider installing a backup push-to-talk switch on the instrument panel (left) in case the primary switch fails. Yoke/stick switch wiring (right) is prone to chafing when not protected. Always think about the simplest solutions first when troubleshooting.

Transmitting isn’t the only gotcha to troubleshoot. There are also noise issues in the VHF com receivers that you might be chasing. When there’s static in the receiver while flying through precip or wet clouds, it’s time for a look at the antennas. New fiberglass antennas have an invisible anti-P-static coating that wears off over time. Check antenna grounding and, on metal airframes, take the antenna off and inspect the RF connector and look for clean bonding with the skin. Got fabric? Make sure the ground plane is sufficient. Often it is not.

But it’s not always P-static and grounding issues. Consider that in audio control systems, both new and old, there is a potential for both radiated and conducted noise interference, and there’s a good chance you’ll deal with it at some point. Manufacturers can only do so much, and as an installer, you do the rest.

This electrical mess was causing noise in the audio system of a Lancair.

For example, the power supply in PS Engineering’s PDA360EX is designed to reduce electrical noise on the aircraft power bus by at least 50 dB. Although this is a sizable amount of attenuation, it may not eliminate all noise, mainly if noise amplitude is very high. Take out your meter and measure at least 13.8 volts at the power input connector. Back to the healthy charging system requirement, this stable input voltage is what it takes for the power supply to work in its designed range. Otherwise, it cannot adequately attenuate power line noise. There’s no need to rip into the wiring to fix that problem.

In many cases, noise filter only mask the real problem. Drill down to find poor grounds, ground loops and other failures to follow best practices before bolting on more stuff.

We’ve said it before, and we’ll say it again: Shielding can reduce or prevent radiated noise—beacon, electric gyros, switching power supplies—but installation combinations can occur where interference is possible. Modern panels are generally designed in an RFI-hardened chassis and have internal electromagnetic interference (EMI) filters on all inputs and outputs. I’ve seen plenty of amateurs (and pros) try to Band-Aid interference issues by installing noise filters. This may be the last resort and is often an expensive guess.

This is beyond basic troubleshooting but something to keep in mind during initial audio installs, especially if you want noise-free entertainment input. Ground loop interference plagues many audio systems and occurs when there are two or more ground paths for the same signal (i.e., airframe and ground return wire). Large cyclic loads such as analog strobe lights, power inverters and motors can inject noise signals onto the airframe and be picked up in the audio system. Follow the system’s wiring diagram carefully to help ensure a minimum of ground loop potential. Start by using only Mil-Spec shielded wires (MIL-C-275000 or better). And here’s a rookie mistake: Under no circumstances combine microphone and headphone wiring into the same shielded bundle. Always use separate two- or three-conductor bundles for the mic and audio, and properly shield the wire as instructed in the diagrams. It is generally acceptable for the shields to be daisy-chained together and then connected to the ground lugs mounted on the connector backplate. Again, look carefully at schematics.

By far the biggest headache for do-it-yourself avionics installers is getting modern audio panels to work without annoying electrical noise. If there’s any doubt in your ability to properly wire them, farm it out to an avionics shop. If not, follow the manual to the letter.

Troubleshooting Over a CAN Bus

Some systems are interconnected over a CAN (controller area network) bus, and sooner or later, you’ll learn how the various components in the avionics suite are joined. Get a head start and look at the block diagram for the CAN bus in your interface. Garmin’s G3X series is a common CAN interface, and depending on the number of screens and LRUs (line replaceable units), some versions are more complex than others. Most times, a failure that causes the dreaded “red X” on any given piece of data will be a failed LRU. But there is some troubleshooting that you’ll need to do to determine which remote component has failed. It’s  a detailed enough process that we’ll jump into the particulars next month.

Engine/Fuel Display Troubles: Know the Probes

Whether the data is integrated within a big-screen display or on a standalone control head, fuel flow and engine temperature measuring are similar across the board. Eventually, you’ll deal with a faulty temperature probe. Some engine monitors will help you diagnose a faulty probe, while you’ll have to hands-on troubleshoot with others. Know what probes are on the engine, including which ones are gasket probes. For CHT gasket probes (also known as washer-type probes), JP Instruments advises installing them on the top spark plug and to expect as much as a 15° temp difference (on the cold side). Placed on the bottom, you’ll see the same error spread, but on the hot side.

Understand the limitations of different probe types. Otherwise, you’ll be chasing your tail attempting to fix a failure that doesn’t exist. Do keep a spare probe or two on hand. When you suspect a failed or failing probe by spotting erroneous (or missing) temps, it’s an easy task to swap one out. And be sure to provide plenty of protection to the probe’s wiring using proper clamps, standoffs and wiring strain relief.

Install engine probes and harnesses using standoffs and strain relief where appropriate (right). Avoid running sensor wiring alongside spark plug wires.

For jumping temperature readings (usually with all EGT and all CHT values) on a standalone engine display, the first thing to do is check all instrument grounds at the engine. Make sure the probes are secured and fully attached. The chances of all four or six probes failing at once are extremely remote; look for common ground, so to speak.

If you see an error message along the lines of OPEN PRB, you know right where to go. The first thing I might do is swap probe wires with one that is known functional and also connect the suspected faulty probe to the other wire. Don’t swap probes. If the original cylinder is still getting the message, there’s a connection problem, likely a bad crimp. Get your meter and ring it out from the probe wire to the display.

Gross inaccuracies in fuel flow could point to a faulty fuel flow transducer. Minor differences in flow could require a tweak of the K factor. This is the number of electronic pulses per unit of fuel that passes through the transducer. This K-factor setting is accessed and changed in the engine monitor or fuel computer setup/installation menu.

Got a new failure or strange bug? Consider what other systems might be affected when doing software updates. Third-party interfaces may not work to their fullest potential afterward.

Are you not seeing endurance data? Maybe you’ll see something like a BAD DATA error message. That generally means the fuel computer/engine monitor isn’t communicating with one or both of the onboard GPS navigators. This is generally a serial or ARINC connection between the GPS and the computer. Start with the installation manual for your current mod status and software. Remember that software updates can create a change in configuration settings, including format and data speed. Before updating any software, know if it will be compatible with your setup. If the units are configured correctly in the software but still aren’t communicating, you’ll have to take out the meter and wring out the data bus lines wire by wire. Maybe a pin got pushed back, a wire got broken, or if you’re having a really good day, maybe the serial data connector came loose. Also keep in mind you likely won’t see any data if the GPS isn’t locked on, as may be the case when troubleshooting it in the hangar—another potential tail chase.

Modern engine monitors (left) will usually help you diagnose faulty temperature probes. This G3 is showing out-of-tolerance probes (right) in red. Got a fuel totalizer? Troubleshooting might include tweaking the transducer’s K factor value.

Autopilot Woes

Pilots kill themselves troubleshooting autopilot problems. Whether it’s runaway trim, a servo clutch that won’t disconnect or an approach tracking mode that does something wonky, approach autopilot troubleshooting with a healthy dose of respect and attention. It’s nearly impossible to offer solid troubleshooting advice for autopilots given the wide range of models, theory of operation (rate-based versus attitude-based), the servo types and flight computer considerations. But there are some things you can do to help a shop or the equipment manufacturer set the troubleshooting in the right direction. First off, follow all procedures in the autopilot pilot’s guide, the aircraft’s flight manual and any flight manual supplements. There should be a section that covers abnormal and emergency procedures. Know exactly how to pull power off the autopilot’s servos. It’s surprising the number of pilots who don’t know where the circuit breaker(s) is.

Improperly rigged autopilot servos are a source made for sloppy system performance. Check them first (with a calibrated tensiometer) before shotgunning flight computers.

For integrated autopilots, including Dynon and Garmin embedded models, tracking gains and calibrations are accomplished in a dedicated setup menu. Gone, mostly, are the days of turning adjustment pots on a flight computer to get the system to fly the way you want it. But servo motors still fail, and builders still botch the install by using incorrect hardware or mis-adjusting control cable tensions. Maybe the system isn’t holding altitude (or has the annoying porpoising), or it’s blowing through the localizer on an approach. Invest in a good tensiometer so you know the cable tensions are within the manufacturer’s specs.

There isn’t much you can do for problems within the servo, including sticking clutches and weak drive motors. Maybe you’ll break a shear screw, which is fail-safe hardware installed on the servo capstan that’s designed to break when control forces are too great on the clutch, allowing it to freewheel. Dynon servos, as one example, are allowed up to two field shear-screw replacements before they need to go back to Dynon for a look. Keep breaking screws and you may need a heavier-duty servo. Stick with a servo that’s known to work well in your particular application. A few years back, there was a good-sized string of broken screws among Dynon servos in service—I broke two within 3 hours of flying after installing the pitch servo in a fast, heavy single. Still, Dynon makes a good point in advising that you want this shear screw strong enough to not break during normal operations yet yieldable to do its job should the clutch fail.

If the autopilot has any connection with the pitot and static systems, make sure there are no leaks in the system. Keep up with two-year pitot-static inspections. Loose fittings and cracked and brittle lines are a reality, especially in aging aircraft. And like push-to-talk switches, autopilot disconnect buttons and pitch trim command switches can fail. Use the same install considerations,  provide for good chafe protection and use high-quality switches.

Finally, sometimes all an autopilot needs is a tweak of the gain settings to make it fly as you want it to, including more or less aggressive bank angles, roll centering and approach/altitude capturing. Follow the manual closely and always, always make the adjustments in visual conditions with a safety pilot.

WAAS GPS antennas (left) generally require preamp input voltage from the GPS; measure voltage at the antenna connector with the navigator powered on. Also check that the GPS engine is seeing the antenna (center), and read the manual to determine the reason for loss-of-integrity alerts (right).

ADS-B and Transponder Issues

First, understand that ADS-B performance can rely heavily on the transponder system’s health, which includes the altitude encoder. The good news is that ADS-B troubleshooting is easier with the FAA’s Public ADS-B Performance Report or PAPR. You can request them online any time the aircraft has operated within ADS-B coverage areas; plug in your details, and the FAA will send you an e-mailed report. Assuming that the installer got the initial ADS-B configuration programming correct, there generally isn’t a lot to go wrong from an installation standpoint.

On the other hand, it’s up to the installer to properly install the ADS-B L-Band antenna. Shadowing can be a problem, and like the transponder antenna, grease and grime from engine exhaust blow-by can drastically reduce performance. Clean all of the antennas on the belly frequently with a safe aircraft-grade cleaner. In an ADS-B world, it’s more important than ever to keep up with FAR 91.411 and 91.413 transponder/static system testing and certification. Bring it to a shop that can verify the ADS-B system is working, and also one that can bench test the transponder (and altitude encoder) if needed.

Button It Up

While today’s components are generally reliable, failures aren’t uncommon—especially on new systems—where there may be a certain amount of infant mortality. Still, wiring problems are more common, especially when inexperienced installers did the wiring. That’s why I advocate testing as much as you can on the bench or before closing the airframe and installing the interior.

Software updates are generally straightforward, but they aren’t without problems, either. If you are uncomfortable doing your own updates and aren’t sure if the revision will change the existing configuration/setup parameters, have a skilled shop look after the update.

Speaking of shops, some troubleshooting tasks are best left to the pros. We’ll look at when to involve an avionics shop (for troubleshooting and installing) in a separate article.

1 COMMENT

  1. Good article couple of points. There may be multiple data busses, and not all data is on all busses, so sometimes you see partial data, ie csome are missing…indicates maye one specific bus is down.

    Second Canbus lack of terminators can kill the bus, lack of or a broken resistor.Ns sonetimes well meaning people add termi ators to sockets when no tneeded, canbus should be a ” long string, ” you might get away with short tee stubs , but not guaranteed, and i have seen all sorts of intermittencies .when peopke tee too liberally. Its a data bus , not housecwiring, connection with a meter may be ok, but at data frequencies…..

    You cant read activity on a databus with a regularr test meter, and you cant share data by joining databusses .

    And finalky, imho many issues are por or missing grounds, so currrent goes through and other path . I once saw ignition wires melt since with a missing bond airframevto engine, tge starter current return ws through t he screen of the magneto wires.. A one inch wide copper braid just hanging in space…

LEAVE A REPLY

Please enter your comment!
Please enter your name here

This site uses Akismet to reduce spam. Learn how your comment data is processed.