I don't spend enough time working on the website, partly because "Vac" and "website" and "computer" can be mutually opposing forces of nature incapable of achieving any sort of harmony--detente is a better term, but only on a good day. Thus, I've be recalcitrant in my duty to report and archive results; so, in no particular order:
Homebuilding at it's Best
One of our program objectives is to have a design that can be built by someone with basic electronic skills. I certainly don't have those skills, but a surprising number of folks in our EAB community do. The past few months, we've been working with a some talented homebuilders that have built their own systems. These folks have been gracious with their time and patience, but most importantly have helped us develop better assembly instructions. We've also learned about kinks in the supply chain, so we've developed a list of critical components that we are now keeping in bench stock so we have a limited source of parts to help folks that are building their own system. I wanted to call our original group of homebuilders the "Texas Solder Team" only because I thought it was a fun name. Unfortunately, of that group, only one person actually lived in Texas; so there was no real impetus to order tee shirts...I'd like to thank Gary Sobek and Carlos Rodriguez in particular for their efforts to improve our publications. Gary is a talented engineer that donated his time and money to build two systems for the program that will used for operational test; and Carlos is a USAF fellow traveler with an RV-8 that decided to go Full Monty, modify his airplane and build a system from scratch to equip it with. For the Heathkit crowd, we are looking at putting together a kit with some of the one off or more difficult to get components (motherboard, end plate, specialty fittings, etc.)--another thing added to the "to do" list. As always, we are happy to help out anyone that is building--just post in the forum or drop a line. For now, we've consolidated the latest and greatest hardware and software information on our GitHub site.
A Pile of Goodies Showed up the Other Day...
It was pretty exciting for me to get a big box from Phil and a small box from Vern last week. These boxes contained all of our operational test and evaluation (beta test) gear: 18 ship sets of V3 boxes, wiring harnesses, energy display power boards and hardware test gear.
We've decided that in order to begin operational test, we'll need to have sufficient web-based support and auto calibration capability for users in the field. The current calibration technique using GPS horseshoe legs isn't inherently difficult, but it is tedious and time consuming, especially considering the number of runs required to accommodate various flap settings. It also requires post-flight analysis, subsequent system programming and setting tweaking. I fully realize not everyone loves to fly test sorties or spending hours working on a computer to have an operational system, so it will be a bit before our dedicated team of volunteers with the motto "no rush, get the science right" has an effective "calibration wizard." Our current concept for that is a software tool that walks you through a very simple calibration run in flight. Lenny is currently toiling away in the dungeon, and we hope to begin testing soon. Our objective is easy and accurate. We'll see what's in the art of the doable!
Oshkosh CNX'D
Oshkosh is one of the highlights of my year, as I'm sure it is for many folks. And like many folks, I was bummed to see it cancelled; but thought the EAA made a difficult, smart choice and our team is stoked to re-engage next year. The one silver lining is we've got some extra time to continue working on the project, which is good because like anything worth doing the standard time estimate applies: whatever time you THINK something will take, double it and add 30%. In fairness to the FlyONSPEED team, I have to point out that a lot of the friction is the Stick Monkey trying to keep up with the engineers (that's at least the 30% factor).
The RV-4 Gets and Energy Display, Almost
Thanks to Tron, the RV-4 will be getting an energy display soon. Since there isn't a glare shield, and the display makes more sense in a "heads up" location than on the instrument panel; I'm going to mount it using GoPro parts to the canopy skirt. Lenny printed the back plate, all hardware was obtained from the avionics isle at Ace Hardware and one of Vern's nifty power boards is installed. Initial test will be to ops check the display logic, determine sunlight view ability and heat effects. Under the bubble canopy on the RV-4 is about as bright and hot as we can get for testing. There are two versions of the M5 stack available. The one in the photos is grey and has an IMU, a black display sans IMU is also available. The IMU version is over-kill for a simple energy display, but was on hand. This unit is simply cobbled together for test.
How Well Does it Work?
There are lots of AOA systems on the market; but I haven't run across any published qualitative analysis regarding accuracy or performance, with the exception of one FAA report. We think it's reasonable to determine overall system accuracy and define transient response performance. System accuracy can be expressed in degrees, i.e., how accurately is the system measuring angle of attack? Defining transient response is a bit more difficult, but the objective is that that system can keep up with a reasonably quick pilot input, or, perhaps more importantly, an instantaneous gust load as you would experience in turbulence during the base turn or final approach.
Since I'm a tone baby, and have been flying with this logic since 1988, my ear/elbow/butt interface is fairly well calibrated. Thus, I'll happily state: subjectively the system works great. Certainly on a par with similar systems I've flown in fighters. However, what we need are objective facts...my subjective evaluation isn't sufficient and there is no doubt at this point that I'm biased. There are two ways we can perform an objective analysis: post-flight reduction of flight test data; AND, the occasional natural experiment...
...From the latter category, I recently had an airspeed indicator camera installed in the airplane for some flight test work. Normally, I turn the cameras off (or run out of battery) before I RTB; but on this flight, I simply let both cameras run for the duration. My airport is on the shoreline adjacent to Eglin AFB in the Florida Panhandle. It also lies in a "tree canyon." We have some pretty significant shoreline effects on wind (i.e., daily sea and land breezes). This combination results in some occasional wind shear. I'm fairly conservative in my operational risk management, since by this point in my flying career, I figure I've used up most of my luck. During this flight, conditions appeared fine on the ground for takeoff and Eglin ATIS was unremarkable, but as soon as I climbed through about 500 feet I knew that things would be more interesting than required. At that point, I encountered a significant wind aloft (about 30 knots out of the East) with nearly calm winds at the surface. Interestingly, this river of air wasn't very deep. By the time I climbed through about 3500 feet, winds aloft were light and variable. Experience told me RTB was going to be interesting, and it was: https://youtu.be/0RiB6_8bx3c
At the airline I fly for, one of the definitions of severe wind shear is a change in velocity of 15 knots or more. In the video, if you watch the standby airspeed indicator, you'll note that as the bottom drops out, indicated airspeed decreases about 15 MPH almost instantly. You'll also note the tone changes from ONSPEED to slow/stall just as quickly. My reason for sharing this video is to simply demonstrate how well the tone performs under very challenging conditions, including a rough ride down final. It DOES NOT provide trend information in the flare, the bottom simply drops out; but, because my energy is good (ONSPEED) until just above the turf, when it does things get sporty but remain manageable.
Putting on my safety hat, I'd also like to point out that a go-around would have been perfectly acceptable, and the wheels may well have touched during the conduct of that maneuver. I only point that out since most of us are predisposed to land once we get that close to the runway, and an interesting statistic to ponder is that less than 5% of unstable airline approaches actually result in a go-around decision.
Next blog I'll publish some tables and charts showing accuracy and transient response performance of the Gen 2 V3 system.