It already had been a long day, flogging my Piper Arrow across what felt like half the U.S. I was aiming for a sizable regional airport to spend some vacation time with an old friend, which was turning into a two-leg, seven-hour slog into headwinds. The weather mostly had been clear, but an undercast crept in below me for the last 200 or so miles. As a result, my destination was advertising 800 feet overcast with good visibility underneath and little wind. It had plenty of approaches, and I had been here before.
Although I received what I considered a late descent clearance, I didn’t anticipate any problems. The airplane was fine and I was instrument-current. The only potential squawk I had was a directional gyro that needed resetting every 15 minutes or so, thanks to precession. The rest of my panel’s steam-gauge instruments and the vacuum system itself were nominal.
Soon I was on a vector to intercept the localizer as ATC vectored a slew of other airplanes for the same approach. They all were getting in fine, and ATC clearly wanted me to keep my speed up in the descent and once I leveled off for the approach. The autopilot was in heading mode, and the rest of the panel was ready for the ILS in use.
Soon I was on a vector to intercept the localizer, still with a good head of steam. As the localizer needle centered, I twisted the DG’s heading bug to align with it and scanned some other instruments. Shortly, ATC remarked that I had flown through the localizer. A glance at the needles confirmed it, although the DG was still indicating the inbound heading. I disengaged the autopilot, thinking I could maintain heading better than the DG. I slowed down a lot, too.
The controller gave me vectors for another try, which didn’t work out much better than before. The DG was precessing madly, and the constant vectoring wasn’t helping. At one point, I glanced outside to see if the terrain was visible. When I looked back, the airplane was in a 30-degree left bank and descending. That won’t do at all.
I was about to request no-gyro vectors for the third attempt when I got a much shallower intercept angle, which allowed for a similarly gentle turn onto the localizer. I was finally established. As I crossed the final approach fix, I dropped the gear and set power for the glideslope. Soon, I had broken out and had the runway. The rest was uneventful.
The directional gyro had been exchanged for an overhauled unit a few years earlier, and didn’t have all that many hours on it. The vacuum system checked out. The flight home was uneventful.
A few months later, I replaced the DG again, with another overhauled unit. The real fix will be to ditch the steam gauges and go all-glass, for which I’m saving my nickels.