Let me try again. I have a D10A and a D100, each have their own remote compasses which are mounted one on top of the other. After upgrading to 3.0 I swung the compasses using my King stone age whirling wheels gyro coupled to my Sandel as a reference. Previous measurements have shown the Sandel to be within one degree of the cardinal headings. After settnig up the Dynon units they where within a degree or two of the Sandel, I was pleased as Punch (an English outburst of positive emotion). I completed the condition inspection on my plane last week, a part of which was checking to see that the remote compass units weren't rattling about in the wingtip, they were there as snug as a bug. I flew about 2.5 hours this last weekend and noticed thet the two Dynon units were still within a degree of each other , but were 10 degrees out compared to the Sandel. Of course it must be the fault of the whirring wheels gyro! But NO!!! Rolling out on the runway the Sandel was about 3 dgrees off the runway heading and the Dynons were around 6 or 7. This error has been creeping up over time as I've noticed a slightly elevated error when compared to the Sandel, up to 5 or 6 degrees, but 10 degrees caught my attention. So I am now suspect that the Dynons my be incorrect. I know this is not an exhuastive test but on the surface things don't look too good. The puzzling thing is that the two Dynons are so close to each other. So if anyone has sage words regarding this I would appreciate reading them. Meantime when the weather clears I'll run off to a compass rose and check things more closely. The error was observed at a northerly and southerly heading (oubound and inbound from my trip)