(Image: Header Graphic)

Tuesday, August 20, 2019

Doug's Domain

Doug Vetter, ATP/CFI

dvatp.com has been updated!

June 22, 2016: Donations to dvatp.com are now processed via Stripe. Like this site? It's easier than ever to show your appreciation.

Charging System Failure

(Image: A dark panel following a charging system failure)
In cruise running on battery following a charging system failure

Due Diligence

You may recall from my Annual 2003 series of articles we found a broken lower alternator mount while removing the alternator for an overhaul in accordance with our preventative maintenance program. We didn't know how long we'd been flying with it broken, but we figured it must have broken at some point since the last 50 hour inspection, or about 25 hours.

While examining the engine at that time, I took a close look at the terminations of the wires leading to the alternator and didn't like what I saw. While the terminals appeared secure, several wires exited the terminals at a strong angle and appeared to be under strain, so I asked Marty to replace all of the terminations just to bring them up to speed. Combined with the early alternator overhaul, I figured we'd dramatically increase the reliability of the aircraft.

Problem solved, right? That's what I thought. Fast forward about 50 hours of flight time.

(Image: IMC Departure from Nashua, NH)From 800 and 2...

After a personal trip to Florida and back, my crew partner and I decided to fly to my old haunt -- Nashua, New Hampshire. The forecasts called for generally beautiful VFR weather up through New England so we launched expecting to punch through some scattered to broken decks below 6000 or so, and land under good VFR.

How forecasts change. Just north of Hartford, CT, we heard the Nashua ATIS struggle to reach the noise floor. After listening to it twice, I still thought I was hearing things, but the report eventually came through loud and clear: 800 and 2, landing runway 32, GPS 32 approach in use.

Although we were initially on top during cruise, the clouds came up to meet us, and before long we were solid IMC. The call from Boston approach came to descend and proceed direct to the southern-most fix on the GPS 32 approach. Then the rain came, and it became more intense. It was probably a solid level 2, possibly level 3. We flew the GPS 32 approach on our own, and tried to be smooth about it, in spite of the continuous light turbulence.

A couple miles outside the final approach fix, the Garmin gave us the green "Approach Lock" indication. We kept the needle centered, and broke out at minimums. Visibility in the rain was indeed about two miles, and we had difficulty seeing the runway. A quick call to tower to crank up the runway lights revealed what we needed to descend below the MDA, and we landed uneventfully.

(Image: Landing at Northumberland, PA)To CAVU

The following Saturday dawned in stark contrast to the prior weekend with absolutely perfect weather. Aside from higher than average winds aloft, we were promised a nice day of flying. Our trip would take us to Bradford, PA, by way of Northumberland, for some cheap fuel.

We began the flight like any other, and managed a textbook flight from our base to Northumberland. The 24 hour credit card pump helped us make quick and relatively inexpensive work of our refueling effort, and after talking with a friendly local pilot, we launched on our way to Bradford.

As a matter of procedure, we checked the ammeter several times; once immediately following start (when charging current is higher and therefore more obvious), once prior to takeoff as part of the Before Takeoff procedure, once during the takeoff roll, and again during the After Takeoff procedure at 1000 feet AGL. All checks revealed normal conditions this fair weather day, so I made our usual call to Harrisburg Departure as we tracked toward Selinsgrove VOR courtesy of the GPS.

NORDO

And then it happened. As I completed my most recent scan for traffic in the vicinity of the VOR I pulled my eyes back inside the cockpit, and began a quick survey of the panel. I glanced at the ammeter, and did a double-take. What the 8(*! is this? I found the needle far to the left of its normal, near vertical position. I figure it indicated a discharge of approximately 3 needle-widths, or about 12-15 Amps. I instantly knew our alternator was offline.

(Image: C172 Schematic Diagram)At this point I fell back on my training, and focused on solving the problem, while my crew partner flew the airplane on the clearance. I began the troubleshooting procedure by looking at the field and main bus breakers. The field breaker was the non-pullable 5A original, while the main bus was a brand new 55 Amp unit replaced at the annual due to a Cessna service bulletin. I found both in the closed position, so I redirected my attention to the Master Switch only to find both sides the correct up / on position.

In a newer 172, now would have been the time to check the overvoltage protection relay by turning the Master Switch off for a few seconds and then turning it back on. Unfortunately, this vintage 172 lacks an OVP relay, and the POH-specified procedure to combat an excessive charge rate is simply to turn off the alternator side of the master switch to disable the alternator field current. For this reason, I decided to turn off the alternator side of the master switch at this point, but I also figured I'd kill the battery master as well, just for good measure.

It was at this point that I glanced outside and thanked the weather Gods for providing such a nice day. What could have been a major deal on our unexpected IMC flight to Nashua, was today merely a major annoyance. I quickly relayed our condition to Harrisburg and asked to go NORDO for 30 seconds. They approved. Unfortunately, we soon reported to them our unsuccessful attempt to fix the problem, and requested cancellation of our IFR clearance.

The helpful controller offered any services we needed, but I told him we were going to simply put our tail between our legs and head home, transponderless. He wished us luck, and we said our goodbyes. I then did something I haven't done in an airplane other than for training purposes...I turned off all lights and avionics other than the audio panel, and watched the ammeter spring back to the general vicinity of the vertical, and the voltage creep up to 12.1. As expected, however, the low voltage LED on the electrical system profiler remained illuminated as a reminder of our predicament.

As I glanced over at the vacuum gauge, and watched our gyros continue to spin merrily in spite of the crippled electrical system, I thanked my stars that I didn't have an all-electric airplane at that moment. Cirrus and Lancair pilots, eat your hearts out!

Threading the Needle

On the recent Florida trip, I found myself using our high-tech GPS to "thread the needle" through thunderstorms on the way up the coast from Florida, and today I found myself using old-fashioned pilotage to thread the needle through the airspace of eastern Pennsylvania and New Jersey that didn't require a transponder. We started a turn to the southeast, traded the instrument chart for a sectional, and began to hop our way from one prominent landmark to the next.

(Image: Using a battery charger to top off the battery)Thanks to a strong tailwind, a mere 45 minutes later, we arrived a few miles outside of the New York Mode C ring, which signalled our need to turn the transponder on. Fortunately, due to our power conservation efforts, there was more than enough juice remaining to power the transponder.

While I normally begin to broadcast my position on Unicom 10 miles out as recommended by the AIM, in the interest of conserving power, I turned the slimline secondary comm (a Narco Com810) on about 3 miles out over a well- known pattern reporting point. I made all the usual calls, and set ourselves up on the downwind.

Abeam the touchdown point, we followed our SOP of 80 knots and 10 degrees of flaps, again emphasizing the need to normalize operations during this abnormal operation. However, I decided to restrict further flap operation to conserve battery power in the event we needed to go around. No sense having 30 or 40 degrees out on a go around.

That turned out to be a good decision, too, because just as I started to turn base, someone reported on frequency that the geese that normally congregate near the runway decided to congregate smack dab in the MIDDLE of it. To make a long story short, as the airplane neared, they got the message and figured it would be better to move on, rather than be speared by our prop. The airplane soon came to a stop, and we taxiied back to the hangar, determined to figure out what went wrong.

Impromptu Maintenance

With the airplane safely back in the hangar, I decided to put a charger on the battery and call my partner to let him know what happened. He suggested we get together the next day and try to troubleshoot the problem, and that's exactly what we did.

We began by looking at the obvious stuff. We performed a cursory inspection for frayed or disconnected wires, and found none. I began to think that it may have been a simple alternator failure, but my partner pointed out (correctly, I may add) the most common failure mode -- a diode failure. This failure diminishes the alternator's maximum output current, but does not prevent it from serving its role as battery charger at reduced bus load. So, his position was that the alternator might still be good, and we should look more closely at the regulator. I gladly accepted his viewpoint, if for no other reason than we had just spent $350 on the unit and I really didn't like the thought of replacing it. :-)

We happened to have a spare regulator handy, so we decided to swap it to determine if the regulator was at fault. Unfortunately, a quick run-up revealed no difference. The ammeter continued to show a discharge.

(Image: Closeup of broken terminal) Dumbfounded, we then decided to take a closer look at all the terminations. We peeled back the boot covering the output (A+) terminal and gave an extra strong tug on the output wire that is normally routed to the main bus breaker. The problem quickly became obvious.

The 8 gauge multi-strand wire had completely broken through at the intersection of the terminal crimp. A fairly common failure mode, but here's the strange part. Only a few of the breaks looked "fresh". It was pretty obvious this had been fracturing over time, and only a few strands broke suddenly.

The wire didn't budge that much on prior inspection because it was routed along with the regulator feedback wire, which was still well crimped in its own terminal and affixed to the same binding post on the alternator. And, it was just our luck that the plastic insulation of the terminal obscured the break. This was a good lesson in looking but not seeing.

We then reinstalled the approved regulator and reterminated the alternator output wire. A&P friend gave it a thumbs up and we considered that a job well done.

Lessons Learned

So, you may have been wondering, as I was, how a new termination could break so easily. Well, the fact is that while most of the terminals were replaced per my request, several, including the main feeder, were NOT for some reason.

To Marty's credit, this wasn't his doing. He told his (former) employee to replace the connections as I'd directed, and both Marty and I assumed he did the job properly. I can imagine him saying to himself, "aw, that looks okay, and those big connections *never* fail". Riiiiiight. It's true what they say -- it's hard to find good help nowadays.

(Image: Closeup of broken main bus feeder wire) Recall the broken alternator bracket. It's pretty safe to say that we were flying for a few hours with the alternator somewhat less than "fixed" to the engine. We figured that the extra vibration caused additional stress on all the terminations, but the heavy gauge wire in particular, simply due to its mass. It couldn't flex like the smaller wires to absorb the vibration.

As each individual strand of the wire broke, it increased the resistance of the connection. Combined with the recent night flights and 25+ amp loads, the remaining strands likely heated to the point that the copper became brittle. In spite of the fact that we replaced the alternator bracket and reduced the vibration level, the damage was already done. It was only a matter of time before the wire failed.

For this reason, I think it's important to redo critical connections like this periodically, say every few years or 500 hours. In fact, that will be our plan from this point forward. Also, after getting a very close look at the 30+ year old firewall forward wiring, we're also planning to replace it all with a modern shielded, teflon coated equivalent at the upcoming annual.

If there is a good side to this story it's our incredible luck. Not only did the failure occur during good weather, we're quite fortunate that the main feeder wire was the one that failed, because when it broke, it produced the effect of turning off the alternator field. Had the main feeder remained attached and the A+ regulator feedback wire failed, it's likely that our expensive electronics would have received a potentially lethal voltage spike.

So, when was the last time you took a VERY close look at your charging system?