Narrative:

We were on a short IFR flight (about 30 mins in duration). We had planned an instrument approach into fairhope based on the nearest reported WX at bfm (mobile downtown). We were cleared by mobile approach control to descend to 4000 ft after having been held above 10000 ft longer than usual. This caused me to set up a steeper than usual descent in order to get down to 4000 ft early enough to be cleared for the approach without having to take vectors and/or circle to descend. I was using the autoplt throughout this portion of our flight. After I completed setting up the descent, the copilot and I then completed the descent checklist and began briefing for the upcoming instrument approach we had been told to expect. I began to shallow out the angle of the descent (using the autoplt trim wheel) when coming through about 6000 ft, but still kept it about 2000 FPM. My attention then returned to completing the approach briefing and setting up the correct frequencys and headings for this approach. It was during this time that something went wrong with the autoplt/flight director. The altitude hold did not capture at 4000 ft and I inadvertently descended through this altitude. I did not catch this malfunction until ATC called and reminded us that we had been clear to descend only to 4000 ft. I immediately disengaged the autoplt and took control of the aircraft to stop the descent. I had inadvertently descended to 2600 ft and then immediately climbed back up to a newly issued altitude clearance of 3000 ft. I continued to fly the aircraft by hand for the remainder of this flight. During the ensuing approach to fairhope, the flight director once again failed to give the command to level off at the preset altitude dialed in by the copilot. I ignored its erroneous command to continue descending below the altitude published for that portion of the approach, of course, but was now more aware of what kind of problem we were having. I then canceled the flight director and flew the rest of the approach using raw data to the missed approach point. I then did indeed have to execute a missed approach due to the ceiling being below the published MDA. During the missed approach procedure, I decided to try the flight director again, but continued to hand-fly the aircraft. ATC issued us vectors and a clearance to fly an ILS into bfm. During this phase, I saw no more of the problem we were previously having. The flight director issued correct commands during this portion of the flight and we landed uneventfully at bfm. After waiting for our 4TH passenger, we departed mobile and made our planned flight to 4r4. Watching the operation of the flight director and autoplt with much vigilance, I used it during much of this flight with no problems noted. But, as we were rolling out after landing at 4r4, the altimeter and airspeed indicators on the pilot's side began to malfunction. This was accompanied by their respective 'off' flags being displayed and the sounding of the overspd warning horn. I recognized this as a probable failure of the air data computer and perhaps the cause of our earlier problems. I had maintenance personnel come over and confirm my suspicions. They ordered a new air data computer and replaced the failed one early the next morning. We have had no more problems with either the autoplt and/or the flight director since the replacement of the failed air data computer. The malfunction of the autoplt/flight director contributed to this incident, but the momentary lapse in our total situational awareness was the primary contributing factor in this incident. I feel that the somewhat steep descent and the relatively short duration of the flight were also contributing factors. This incident could have been prevented by my more judicious use of CRM.

Google
 

Original NASA ASRS Text

Title: TWO ALTDEVS DURING AN APCH WHEN AUTOFLT SYS IS AFFECTED BY A MALFUNCTIONING AIR DATA COMPUTER.

Narrative: WE WERE ON A SHORT IFR FLT (ABOUT 30 MINS IN DURATION). WE HAD PLANNED AN INST APCH INTO FAIRHOPE BASED ON THE NEAREST RPTED WX AT BFM (MOBILE DOWNTOWN). WE WERE CLRED BY MOBILE APCH CTL TO DSND TO 4000 FT AFTER HAVING BEEN HELD ABOVE 10000 FT LONGER THAN USUAL. THIS CAUSED ME TO SET UP A STEEPER THAN USUAL DSCNT IN ORDER TO GET DOWN TO 4000 FT EARLY ENOUGH TO BE CLRED FOR THE APCH WITHOUT HAVING TO TAKE VECTORS AND/OR CIRCLE TO DSND. I WAS USING THE AUTOPLT THROUGHOUT THIS PORTION OF OUR FLT. AFTER I COMPLETED SETTING UP THE DSCNT, THE COPLT AND I THEN COMPLETED THE DSCNT CHKLIST AND BEGAN BRIEFING FOR THE UPCOMING INST APCH WE HAD BEEN TOLD TO EXPECT. I BEGAN TO SHALLOW OUT THE ANGLE OF THE DSCNT (USING THE AUTOPLT TRIM WHEEL) WHEN COMING THROUGH ABOUT 6000 FT, BUT STILL KEPT IT ABOUT 2000 FPM. MY ATTN THEN RETURNED TO COMPLETING THE APCH BRIEFING AND SETTING UP THE CORRECT FREQS AND HEADINGS FOR THIS APCH. IT WAS DURING THIS TIME THAT SOMETHING WENT WRONG WITH THE AUTOPLT/FLT DIRECTOR. THE ALT HOLD DID NOT CAPTURE AT 4000 FT AND I INADVERTENTLY DSNDED THROUGH THIS ALT. I DID NOT CATCH THIS MALFUNCTION UNTIL ATC CALLED AND REMINDED US THAT WE HAD BEEN CLR TO DSND ONLY TO 4000 FT. I IMMEDIATELY DISENGAGED THE AUTOPLT AND TOOK CTL OF THE ACFT TO STOP THE DSCNT. I HAD INADVERTENTLY DSNDED TO 2600 FT AND THEN IMMEDIATELY CLBED BACK UP TO A NEWLY ISSUED ALT CLRNC OF 3000 FT. I CONTINUED TO FLY THE ACFT BY HAND FOR THE REMAINDER OF THIS FLT. DURING THE ENSUING APCH TO FAIRHOPE, THE FLT DIRECTOR ONCE AGAIN FAILED TO GIVE THE COMMAND TO LEVEL OFF AT THE PRESET ALT DIALED IN BY THE COPLT. I IGNORED ITS ERRONEOUS COMMAND TO CONTINUE DSNDING BELOW THE ALT PUBLISHED FOR THAT PORTION OF THE APCH, OF COURSE, BUT WAS NOW MORE AWARE OF WHAT KIND OF PROB WE WERE HAVING. I THEN CANCELED THE FLT DIRECTOR AND FLEW THE REST OF THE APCH USING RAW DATA TO THE MISSED APCH POINT. I THEN DID INDEED HAVE TO EXECUTE A MISSED APCH DUE TO THE CEILING BEING BELOW THE PUBLISHED MDA. DURING THE MISSED APCH PROC, I DECIDED TO TRY THE FLT DIRECTOR AGAIN, BUT CONTINUED TO HAND-FLY THE ACFT. ATC ISSUED US VECTORS AND A CLRNC TO FLY AN ILS INTO BFM. DURING THIS PHASE, I SAW NO MORE OF THE PROB WE WERE PREVIOUSLY HAVING. THE FLT DIRECTOR ISSUED CORRECT COMMANDS DURING THIS PORTION OF THE FLT AND WE LANDED UNEVENTFULLY AT BFM. AFTER WAITING FOR OUR 4TH PAX, WE DEPARTED MOBILE AND MADE OUR PLANNED FLT TO 4R4. WATCHING THE OP OF THE FLT DIRECTOR AND AUTOPLT WITH MUCH VIGILANCE, I USED IT DURING MUCH OF THIS FLT WITH NO PROBS NOTED. BUT, AS WE WERE ROLLING OUT AFTER LNDG AT 4R4, THE ALTIMETER AND AIRSPD INDICATORS ON THE PLT'S SIDE BEGAN TO MALFUNCTION. THIS WAS ACCOMPANIED BY THEIR RESPECTIVE 'OFF' FLAGS BEING DISPLAYED AND THE SOUNDING OF THE OVERSPD WARNING HORN. I RECOGNIZED THIS AS A PROBABLE FAILURE OF THE AIR DATA COMPUTER AND PERHAPS THE CAUSE OF OUR EARLIER PROBS. I HAD MAINT PERSONNEL COME OVER AND CONFIRM MY SUSPICIONS. THEY ORDERED A NEW AIR DATA COMPUTER AND REPLACED THE FAILED ONE EARLY THE NEXT MORNING. WE HAVE HAD NO MORE PROBS WITH EITHER THE AUTOPLT AND/OR THE FLT DIRECTOR SINCE THE REPLACEMENT OF THE FAILED AIR DATA COMPUTER. THE MALFUNCTION OF THE AUTOPLT/FLT DIRECTOR CONTRIBUTED TO THIS INCIDENT, BUT THE MOMENTARY LAPSE IN OUR TOTAL SITUATIONAL AWARENESS WAS THE PRIMARY CONTRIBUTING FACTOR IN THIS INCIDENT. I FEEL THAT THE SOMEWHAT STEEP DSCNT AND THE RELATIVELY SHORT DURATION OF THE FLT WERE ALSO CONTRIBUTING FACTORS. THIS INCIDENT COULD HAVE BEEN PREVENTED BY MY MORE JUDICIOUS USE OF CRM.

Data retrieved from NASA's ASRS site as of July 2007 and automatically converted to unabbreviated mixed upper/lowercase text. This report is for informational purposes with no guarantee of accuracy. See NASA's ASRS site for official report.