|
|
Below are some shots of the BMA "brainbox" that was on display at OSH. At the core is a Single Board Computer (SBC) available from places like Jumpteck/Adastra or Kontron (www.kontron.com). It is franked by a Systron Donner AQRS automotive rate sensor (www.systron.com), tilt sensor, data acquisition board and LVDS display interface. All you need to do is add a power supply, inertial reference, a high brightness monitor from a company like Computer Dynamics (http://www.cdynamics.com/) and some "game" software and VIOLA, instant EFIS!
Building experimental avionics using COTS hardware that is glued together has a long tradition. Early Avidyne systems used an Adastra 486 SBC. Darkangel used the motherboard pulled out of a Dell laptop and there have been several articles in the literature on the subject. There is a strong economic driver to use COTS hardware one of which is that you don't have to come up with the six figures (or more) it take to design a purpose built system. But lets face it, at some level ALL hardware is COTS. If you design a system you still use COTS components (resistors, capacitors, ICs) so what is the difference if you build it or they build it? Well, as they say, the Devil is in the details.
Think of it this way, would you use sheet metal screws to attach the engine cowling on your plane? No? Why not? They are, after all "screws" and aren't you suppose to use screws? Here on the ground they seem to hold the cowling on perfectly fine. Heck I can even fly around the pattern. Why should I spend $25 for one livelock when I can get a box of 100 at the 98-cent store? So what if a few fall out? There are a lot more. Besides, what is the worst that could happen?
The guys who designed the COTS hardware KNEW that they would never have their lives depend on the quality of their design. "How can I do the same thing for less dollars." was their first consideration. "How can this be manufactured with less effort" was probably second. Never, in the design process, did the COTS engineer say "If this thing breaks I will be putting my person at extreme risk". Speaking for CFS, ALL of the engineers involved fly behind the hardware in their own airplanes. At the core of the CFS system is a COTS Pentium processor and the support devices that with around it. Surrounding that are multiple layers of mechanical and electrical protection. Think of the CFD system as an M60 tank. The processor is the driver. Now think of a guy in street clothes sitting in a convertible Yugo. That is the COTS SBC. There is a tornado coming. Where would you rather be.
The "Axis of Evil" of EFIS failures are Static, Vibration and Software. Static is deadly. It can get in anywhere and in a flash (NPI) and kick the grapes out of the most expensive silicone on earth. The really wicked part is that Static, and his big brother Lightning are the petulant siblings residing on IMC street. A street you sometimes have to take to get home.
Vibration is like cancer. It lurks in the system looking for weakness. Those pretty ribbon cables with the neat IDCs (Insulation Displacement Connectors)..Gotcha! That soldered wire without strain relief..Gotcha! Tin plated connector contacts instead of gold...Gotcha! Those lovely, tall, unsupported legs..Gotcha big time! What, no locknuts...Gotcha again! This is too easy. After viewing the system below I asked my 11 year old son Matt what he thought would happen if we put this system on the shaker rig and conducted a standard vibration test. His answer was: "It would turn into an Electric Pińata." Out of the mouths of babes.
Software, what can I say about the potential to screw up in software. I don't think I need to say anything. It is that bad, but there are three things you can to to get reliable software, test, test again and test some more. Even good code usually cannot account for all the permutations that happen when the forces centripetal, inertia and gravity go dancing. Works on the ground, easy. Works in high speed flight, very very scary.
These aren't the only pitfalls. There is electromagnetic interference, sand and dust, moisture, power spikes, altitude, temperature variations and mechanical shock. All of which live for the opportunity to bite you in the spar bolts on short final to minimums. But wait, you also get the opportunity to have your system screaming broad spectrum digital obscenities at all the other panel dwellers. While designing their electronic "Mosh Pit" did the COTS engineers ever even vaguely wonder, or care, if your aircraft COM or NAV radio would peacefully coexist with it's noisy neighbor?
But hey, they call it "Experimental" for a reason, right? You bet. If you want to build or buy and install an EFIS system put together "Mr. Potato Head" style from COTS hardware then go for it! It is fascinating stuff and you will learn a lot in the process. There is one rule you will need to obey if you want your cause of death to not be aircraft related. Never use your experimental EFIS as a primary display in actual IMC. All Lancairs should have at least one, good old fashion, Attitude Indicator. Even if you have two EFIS systems, whatever kills one can kill the other. The AI is different technology with it's own, but separate, demons.
At OSH the boys at BM actually advised me that their system was ALL you needed and that the best backup was a second system. "For IFR flight?" I asked. "Well no, legally, it's for VFR flight only <wink>." was the reply. I wonder if this guy knows how many people die from unintentional flight into IMC or how quickly his system will go Tango Uniform at the first brush with P-Static? You can't placard against IMC flight any more than you can placard against stalls or fuel exhaustion. If it were only that easy.
As we were walking out of the tent of Blue Mountain my son asked "Hey Dad, if you can only use their system outside the clouds then why do you need it?". The warm flush of a father's pride for his son filled me. "You don't, Matt, you don't need it at all. You just need to look out the window".
Regards
Brent Regan
BlueM1.JPG
BlueM2.JPG
|
|