Hi Princess,
great, that narrows it down quite a bit. I would say that the MFD is probably fine, but it is receiving varying headings from the sensor. You mentioned two different sensors, a Garmin Steadycast (which to my knowledge has only a NMEA 2000 connection) and some other sensor that you seem to bring in via an Actisense gateway. Is that right?
I assume that you are not using both sensors at the same time, as that could leads to alternating different headings.
So try a sensor at a time, then, as you have it on your tale at home, turn the sensor slowly through 360 degrees and watch if the heading on the MFD follows. If it does and you find that depending on the direction it points the jitter gets less or more, then it is likely an interference. Note that the typical heading sensor sends out the heading 10 times or more per second, but in the stationary case, where the sensor isn't moving, all sends should show the same heading. For my sensors this is definitely true.
The sensors do no smoothing on purpose as they should report any any changes in heading as instantly as possible to stabilise the display of radar targets on the chart and on North up displays. The chart orientation in head up mode however could bee smoothed quite long as it just needs to display approximately with the general direction up. For that purpose, some MFDs have a course up option that can be active while going towards a waypoint. If your boat then meanders around a bit, the chart stays aligned to the active waypoint ahead.
That said, on a docked boat or the sensor on your kitchen table should result in a steady heading, that doesn't need any smoothing.
It is hard to imagine, that your sensor is sending out a stable heading and the MFD 'invents' the variations. And since the data transmission is checksum or CRC protected it is also not likely that the data gets corrupted on the way.
So it seems that your sensor is indeed seeing some changes in the magnetic field. Normally interfering fields that come from a stationary source would need to change in strength to cause a varying effect on your sensor. But that also means that the amount of variation you see would change with the orientation of your sensor. That's why I suggested to try turning the sensor.
Harald