I’ve spent quite a bit of time thinking about whether or not to incorporate backup instruments in the panel. My primary reasons against adding them are the cost, weight, complexity, and the already redundant Dynon installation. My primary reason for wanting to include some kind of backup is to be able to have enough information to keep the airplane upright and under control no matter what.
First, I’ll address the “no backup” reasons. What is the cost (including purchase, installation effort, weight, etc) of adding conventional instruments? Well, that depends on the instrument. It’s pretty easy to add in an airspeed indicator or altimeter. Those instruments don’t require any electrical power, so they are fairly reliable. I’ve had an airspeed indicator fail in flight before, but not an altimeter. The problem with those two instruments is that they really don’t tell me much. If I’m in an emergency situation where both Dynon’s have quit working (see below), the altimeter is of little value (except in a non-precision approach in very low IFR- otherwise a glideslope will really be what I need), and the airspeed is something that I can gather from several other cues such as wind noise and GPS ground speed. What I would really like to have as a bare bones minimum emergency standby is an attitude indicator. The problem with attitude indicators is that they are either electric or pneumatic, and those two sources are troublesome. If the electrical system is operating normally, then there is really only one case where at least one Dynon unit will not work (see below). Pneumatic systems are either engine driven or venturi driven. The engine driven vaccuum pump is really not even on the table as an option. First, they are expensive and unreliable. Second, the vacuum pad on the engine is going to host the second alternator instead. The venturi is intriguing in some ways, though the drawbacks are of course continuous drag (whether I need the instrument or not) and spotty performance, especially in any kind of unexpected icing conditions.
Now, about the inherent redunancy in the system. Since the D100 and D180 each have their own internal air data computers (airspeed, altitude, vsi) and AHRS (DG heading, Attitude), that provides one layer or protection. If one has a failure, the other one will still be there. I asked the Dynon techs about how the units have failed in the past, and they said that the units do occasionally fail during their initialization, but that in-flight failures are exceptionally rare. In fact, I think I remember him saying that he didn’t know of any. But what about a loss of electrical power to the units? My electrical system includes a single battery with two alternators. Each Dynon is powered by a different bus, and there are only a few exceedingly rare situations that would cause both of those busses to lose power on the same flight. Even if that does happen, the Dynon unit has an internal battery backup so that it can run in a completely isolated mode for long enough to land. Perhaps the only serious vulnerability that I can see is that the units are built by the same company, and both run with similar parts and software. If there was some sort of software breakdown that was not unique to a single unit, it could potentially impact both of my dynon units at once. This is the heart of the “yes backup” argument.
So what about that goal of always wanting to have enough information to keep the airplane upright and under control? How does the equipment in my panel serve that goal, especially when compared to a traditional instrument panel? I would suggest that the goal is unattainable. There are no mechanical devices that will operate in all conditions no matter what. Even in modern airliners, the last resort is still a backup instrument that almost never fails. The best that system designers can do is to design a system that will tolerate failures, with the least hardship to be associated with the most likely failures. My installation is in many ways superior to traditional panels of light aircraft, which are usually not equipped with any significant backup. So far, I’m not planning to add anything other than the two Dynons. I’ll have plenty of VMC hours to gain confidence in them and learn more about their vulnerabilities before I depend on them in IMC.