I do not think it unreasonable to think degradation of the compass/GPS module might be in play. Such a condition has been an issue for many systems, and continues to be one. The issue has been there since aero modeling was first provided GPS coordinated flight. The enemy of electrical components is heat, and the effects of heat are accumulative. Packing all the guidance components into a very confined space with a heat generation source directly beneath them has never made any sense to me. At least elevate the GPS antenna to a location away from a heat source. It would be interesting to place a thermocouple inside the H shell near the GPS/compass unit to establish reference data for how much heat soak the system experiences. If above 40C-50C or so there might be cause for considering heat as a "smoking gun" that could degrade system performance. There is also the possibility that power leads from the battery or main control board are generating noise that interfere with the GPS/compass module. Those that build their own aircraft know to separate the GPS unit as far from batteries and high voltage wires by as much distance as possible, or to shield the GPS from those noise sources.
Two other things are potential factors here though. At 4' or so above the ground the aircraft is still well within the effects of its rotor wash bouncing up from the ground, which is destabilizing. There's also the outside air temperature factor. When cold, batteries do not perform very well, with voltages decreasing rapidly and flight times significantly reduced, at a rate that is not predictable. Halving the flight time obtained under "normal" conditions would not be an out of line initial estimate, and could be even less for the actual conditions. Although a battery produces heat when being used, in cold temperatures they produce a little more heat and generate higher IR, which reduces performance even further while inducing cell damage.
PatR,
Very valid points regarding heat and possible component degradation. We will see what Yuneec says.
I also plan to order another H, maybe just the bird. I need a backup regardless, as well as if I am told to send my original back, I need something to fill the void. It also will give me a chance to test our theories.
At the risk of repeating: my 4 foot test was the height at which I noticed the circling starting. Having repeated the test with a fully charged battery, getting the same results, I do not believe voltage is the issue, at least in this craft.
Also, in my quest to learn everything I can as far as how the H performs in various conditions and attitudes, I am quite confident I had tried the low hover many times in the past without issue. I did testing to observe ground effects and the like, at differing heights.
Another reason I am anxious to get another hull. Test and control units!
One important element I keep forgetting to mention: my first episode with this circling didn't include just the circling.
The first time - the motors did not idle down completely, and tried to tip the H. The motor kill button saved it.
The second time - the motors won and flipped the H before I could get them shut down.
In my subsequent official testing - landings were uneventful once I regained control (stopped the circling) either by turning off the GPS or regaining altitude and landing directly without delay.
Hopefully the motors/flip issue can continue to be mitigated using a direct no hovering landing procedure while we search for a final and complete fix, whether hardware, firmware, procedures, or all, to the overall cause.
Thanks for your comments. This is all good knowledge and theory sharing!
Jeff