Step 1: History
In an experiment, I connected a loose HD to a computer power supply, and felt the resistance to its change of position. I figured a HD in the horizontal and one in the vertical plane (X- and Y- axis) would dampen almost all unwanted shaking.
The 2 Gyro stabilizer model is very successful, but practical use of the device is limited, due to bulk and weight.
A single Gyro has its theoretical limits. A strong Gyro in the front / back plane (Z - axis) could dampen shaking, but would not correct rotation. But if rotation around the optical axis is only a minor component of motion blur... , the advantage is less weight and power requirement.
Step 2: Design
The camera is mounted on a hinge with fastening screw. The other part of the hinge mounts a HD.
The HD can be tilted in the vertical ( Y-) plane from way under the camera to a hugging position for storage.
(my understanding of...) the theory would indicate the best position of the axis of rotation is parallel to its optical axis.
But in this design, any position of the HD Gyro can be tested. It also makes storage and transport much easier than the 2 Gyro model.