Narrative:

In cruise at FL310, I (PNF) went to the aft cabin to use the lavatory. Upon returning to the cockpit, when I looked up, I noticed the command bars had separated from the 'V bar' aircraft on the ADI. The altitude alerter chimed and I looked to see the altitude select at FL310. When I looked back to the ADI we were now in a bank. I knew we had a problem. My initial thoughts were a descent clearance had been issued, but when I saw the bank, I knew the autoplt had fallen off-line. I couldn't communicate with the copilot yet because we use headsets and ear plugs because the noise level in the aircraft is so high in-flight. I quickly put my headset on. All this occurred in a very short period of time, yet long enough to lose 300-400 ft. To make matters worse, the PF (copilot) didn't realize the trim wheel was not functioning. He was trying to correct with it, without placing the passenger in any danger (the selector is extremely sensitive and an over correction the other way is easy to do at 420 KTS). Once my headset was on, I could tell him that he needed to get control of the aircraft manually, which we did, and returned to altitude. This autoplt fell off-line because as I got seated my leg must have brushed against the approach plate, which was in a plastic holder and this bumped the autoplt disconnect button, unbeknownst to either of us. We have 2 selectors to engage the autoplt -- an autoplt select and an autoplt engage lever. To the best of my recollection, the lever on the upper panel remained in a normal position. The PF thought the autoplt was still engaged. This assumption allowed the aircraft to descend further. The copilot with 600 hours on this aircraft was unable during this flight to recognize early the nature of the problem and identify that his initial inputs were ineffective. It should be explained here as well that this autoplt takes a static sense and if outside air pressure falls, the aircraft will descend 150-200 ft very quickly, and if you disconnect the autoplt you lose even more altitude. This is normal for this autoplt. The FAA would not let us change our static source to alleviate the problem as was allowed by raytheon aircraft on the hawker 700, which is the same as the 600 fan. All in all, we managed to regain control of the aircraft in a reasonable amount of time and with no injury to passenger. Callback conversation with reporter revealed the following information: the reporter stated this aircraft is equipped with an autoplt system that is poorly designed and when in use must be continuously monitored. The reporter said the autoplt static and pitot pressure inputs are using raw data as the aircraft is not equipped with an air data computer. The reporter stated with no air data computer input the autoplt is ineffective and subject to abrupt altitude changes in the altitude hold mode.

Google
 

Original NASA ASRS Text

Title: A HAWKER 600 IN CRUISE AT FL310 MAKES A SUDDEN EXCURSION FROM THE ASSIGNED ALT CAUSED BY AN AUTOPLT ALT HOLD DESIGN DEFICIENCY.

Narrative: IN CRUISE AT FL310, I (PNF) WENT TO THE AFT CABIN TO USE THE LAVATORY. UPON RETURNING TO THE COCKPIT, WHEN I LOOKED UP, I NOTICED THE COMMAND BARS HAD SEPARATED FROM THE 'V BAR' ACFT ON THE ADI. THE ALT ALERTER CHIMED AND I LOOKED TO SEE THE ALT SELECT AT FL310. WHEN I LOOKED BACK TO THE ADI WE WERE NOW IN A BANK. I KNEW WE HAD A PROB. MY INITIAL THOUGHTS WERE A DSCNT CLRNC HAD BEEN ISSUED, BUT WHEN I SAW THE BANK, I KNEW THE AUTOPLT HAD FALLEN OFF-LINE. I COULDN'T COMMUNICATE WITH THE COPLT YET BECAUSE WE USE HEADSETS AND EAR PLUGS BECAUSE THE NOISE LEVEL IN THE ACFT IS SO HIGH INFLT. I QUICKLY PUT MY HEADSET ON. ALL THIS OCCURRED IN A VERY SHORT PERIOD OF TIME, YET LONG ENOUGH TO LOSE 300-400 FT. TO MAKE MATTERS WORSE, THE PF (COPLT) DIDN'T REALIZE THE TRIM WHEEL WAS NOT FUNCTIONING. HE WAS TRYING TO CORRECT WITH IT, WITHOUT PLACING THE PAX IN ANY DANGER (THE SELECTOR IS EXTREMELY SENSITIVE AND AN OVER CORRECTION THE OTHER WAY IS EASY TO DO AT 420 KTS). ONCE MY HEADSET WAS ON, I COULD TELL HIM THAT HE NEEDED TO GET CTL OF THE ACFT MANUALLY, WHICH WE DID, AND RETURNED TO ALT. THIS AUTOPLT FELL OFF-LINE BECAUSE AS I GOT SEATED MY LEG MUST HAVE BRUSHED AGAINST THE APCH PLATE, WHICH WAS IN A PLASTIC HOLDER AND THIS BUMPED THE AUTOPLT DISCONNECT BUTTON, UNBEKNOWNST TO EITHER OF US. WE HAVE 2 SELECTORS TO ENGAGE THE AUTOPLT -- AN AUTOPLT SELECT AND AN AUTOPLT ENGAGE LEVER. TO THE BEST OF MY RECOLLECTION, THE LEVER ON THE UPPER PANEL REMAINED IN A NORMAL POS. THE PF THOUGHT THE AUTOPLT WAS STILL ENGAGED. THIS ASSUMPTION ALLOWED THE ACFT TO DSND FURTHER. THE COPLT WITH 600 HRS ON THIS ACFT WAS UNABLE DURING THIS FLT TO RECOGNIZE EARLY THE NATURE OF THE PROB AND IDENT THAT HIS INITIAL INPUTS WERE INEFFECTIVE. IT SHOULD BE EXPLAINED HERE AS WELL THAT THIS AUTOPLT TAKES A STATIC SENSE AND IF OUTSIDE AIR PRESSURE FALLS, THE ACFT WILL DSND 150-200 FT VERY QUICKLY, AND IF YOU DISCONNECT THE AUTOPLT YOU LOSE EVEN MORE ALT. THIS IS NORMAL FOR THIS AUTOPLT. THE FAA WOULD NOT LET US CHANGE OUR STATIC SOURCE TO ALLEVIATE THE PROB AS WAS ALLOWED BY RAYTHEON ACFT ON THE HAWKER 700, WHICH IS THE SAME AS THE 600 FAN. ALL IN ALL, WE MANAGED TO REGAIN CTL OF THE ACFT IN A REASONABLE AMOUNT OF TIME AND WITH NO INJURY TO PAX. CALLBACK CONVERSATION WITH RPTR REVEALED THE FOLLOWING INFO: THE RPTR STATED THIS ACFT IS EQUIPPED WITH AN AUTOPLT SYS THAT IS POORLY DESIGNED AND WHEN IN USE MUST BE CONTINUOUSLY MONITORED. THE RPTR SAID THE AUTOPLT STATIC AND PITOT PRESSURE INPUTS ARE USING RAW DATA AS THE ACFT IS NOT EQUIPPED WITH AN AIR DATA COMPUTER. THE RPTR STATED WITH NO AIR DATA COMPUTER INPUT THE AUTOPLT IS INEFFECTIVE AND SUBJECT TO ABRUPT ALT CHANGES IN THE ALT HOLD MODE.

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.