Hello,
It looks like some tilt switch.
It has several contacts for the different directions.
Bertus
I think bertus might have it. The barbell-looking component in the center of the glass enclosure
looks like it's free to tilt when the device tilts. Does it? Telling us any details like that will make it easier for us to help.
The "barbell"
looks like it's suspended from some springy support, and can move relative to the center of the (assumed) vertical axis of the device. so that it makes
temporary contact with one or more of the 6 (supposed) contacts surrounding it, when the device is suddenly ("suddenly" meaning fast enough to make the barbell move) moved from its original position. I first assumed that it was designed to be positioned vertically, as shown your first two pics, but then realized it might be designed to be mounted in any orientation, depending on how the "barbell" is weighted
If the weights at either end of the "barbell" are equal, then the barbell will realign with the central axis while the device is in tilted position; and will make
another temporary contact when the device is suddenly returned to its vertical. Strictly speaking, that would make it more of a "rotational accelerometer" than a tilt sensor.
But it might also be weighted for the "barbell" to try to remain vertical (instead of realigning with the axis of the device), in which case the contact barbell will remain in contact while the device is tilted from vertical, until the device returns to its vertical position. That would make the device a true tilt sensor, since it must remain vertical to maintain an open circuit between contacts.
So which is it? Does the barbell, after rotational motion, return to gravitational vertical, or return to device-vertical-axis?
Also, top-down and bottom-up pictures might be helpful, to verify what the overall configuration looks like, and/or show us any details we might be missing.
