Narrative:

We were level at our cruise altitude of FL280. ATC issued a descent to FL240. The PNF selected the new altitude in the altitude preselect and we each confirmed the new altitude as per company sops. I (PF) reduced thrust slightly; selected vs mode; and rolled in 300 fpm down. My technique is to initiate a shallow descent for the first 200-300 feet; then increase the rate of descent to a normal target of 2;000 fpm down once the aircraft has started down. I find that this creates a comfortable transition for the passengers and allows me time to smoothly set a target power setting for descent. After selecting vs and 300 fpm down; the aircraft began a slow climb; which I thought would soon reverse itself. Going through 28;200 ft I selected altitude to capture the altitude and stop the climb; but the aircraft kept climbing. Passing through 28;300 ft I disconnected the autopilot and pushed the nose over. In order to keep from plastering the passengers on the ceiling I smoothly transitioned from a climb to a descent; which resulted in us reaching a maximum altitude of 28;500. There was no traffic in our vicinity so I wasn't concerned about an immediate conflict. I quickly returned through FL280 and continued the descent to FL240. I reselected the autopilot on and the rest of the flight was uneventful. I have been flying this airplane for 16 years and it has a tendency to abruptly pitch up during descent if it senses the airspeed is too close to redline. In this case I was cruising at .76 with the redline at .80; which is usually enough of a buffer for a normal descent. I suspect the autopilot pitched up to reduce airspeed and to avoid a potential overspeed condition; but I'm not sure why selecting altitude didn't capture our current altitude. This is about the worst case of altitude control I have seen with this airplane; but it is something that all of our pilots are aware is a potential issue. The autopilot has been checked numerous times by different maintenance shops over the years and no faults are ever found. I think it is a built-in protection that just doesn't work very well; poor design.

Google
 

Original NASA ASRS Text

Title: Hawker 800 flight crew reported the autopilot started to climb after the crew initiated a descent. The problem was caused by a built-in speed protection feature of the autopilot.

Narrative: We were level at our cruise altitude of FL280. ATC issued a descent to FL240. The PNF selected the new altitude in the altitude preselect and we each confirmed the new altitude as per company SOPs. I (PF) reduced thrust slightly; selected VS mode; and rolled in 300 fpm down. My technique is to initiate a shallow descent for the first 200-300 feet; then increase the rate of descent to a normal target of 2;000 fpm down once the aircraft has started down. I find that this creates a comfortable transition for the passengers and allows me time to smoothly set a target power setting for descent. After selecting VS and 300 fpm down; the aircraft began a slow climb; which I thought would soon reverse itself. Going through 28;200 ft I selected ALT to capture the altitude and stop the climb; but the aircraft kept climbing. Passing through 28;300 ft I disconnected the autopilot and pushed the nose over. In order to keep from plastering the passengers on the ceiling I smoothly transitioned from a climb to a descent; which resulted in us reaching a maximum altitude of 28;500. There was no traffic in our vicinity so I wasn't concerned about an immediate conflict. I quickly returned through FL280 and continued the descent to FL240. I reselected the autopilot on and the rest of the flight was uneventful. I have been flying this airplane for 16 years and it has a tendency to abruptly pitch up during descent if it senses the airspeed is too close to redline. In this case I was cruising at .76 with the redline at .80; which is usually enough of a buffer for a normal descent. I suspect the autopilot pitched up to reduce airspeed and to avoid a potential overspeed condition; but I'm not sure why selecting ALT didn't capture our current altitude. This is about the worst case of altitude control I have seen with this airplane; but it is something that all of our pilots are aware is a potential issue. The autopilot has been checked numerous times by different maintenance shops over the years and no faults are ever found. I think it is a built-in protection that just doesn't work very well; poor design.

Data retrieved from NASA's ASRS site 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.