Feb 5 Update

Lot's to write about! We've been busy and in no particular order:


Beta Test Hardware


Our production team (read Lenny and Phil!) have done some outstanding work getting the beta test hardware ready to go. Without getting too "MilSpecy," we are endeavoring to apply best engineering practices to the hardware to make it as bullet proof as practical and meet our design price point of $250 in components for the through-hole design using off the shelf components. We use an off-the-shelf hardware box, but custom make the end plates. One is metal and accommodates pressed fittings into which the 1/4" OD quick release fittings mount and the other is plastic, to allow wifi connectivity. Flight and ground testing have shown the wifi capability to work well, so the plastic plate accomplishes it's intended function. Lenny and Phil are currently working on the hardware test protocol, and we hope to have everything ready by the end of the month--about six months behind where we wanted to be. Oh well!


25 boards ready for boxing and final bench test!
CNC'd endplates ready for pressed fittings and DSUB.

Hardware lesson learned: the "long lead time" component. We did learn that we've got one component in the box that isn't as easy as a search on Amazon...that's the small brass fitting we use to distribute static pressure to the various sensors on the motherboard. It's made by Beswick Engineering and is a miniature fluid power fitting. Beswick is a small, specialized manufacturer that doesn't sell in small quantities; so we are going to procure some "bench stock" (parts on hand) so that we can accommodate folks in the field that need help acquiring the part as well as the occasional box that we hand-build. In the military, slang for anything that has the potential to impede progress is "limiting factor," abbreviated LIMFAC. This small fitting definitely falls in to that category!


Three-way "LIMFAC" pressure fitting

Software Updates


We are working on numerous software updates. Most of these are either data-recording or "usability" related...our core AOA software remains unchanged. Each one of these updates requires ground, flight test or both; so it's a bit of a slog for yours truly when you are working with talented engineers that are several light years ahead of you! Fortunately, I've got a new hangar at an airport much closer to home, so time available for test has increased dramatically. One of our biggest hurdles has been to integrate our two onboard Gen 2 systems with the calibrated test boom so that all of the data is recorded properly. The boom manufacturer (Spin Garage out in Mohave) has update the EPROM software in the wireless system at Lenny's direction; and we've finally cracked that nut--we have excellent capability now and all of the data is recorded on the SD card inside the primary Gen 2 box in my airplane. Since the boom angles are accurate up to 40 degrees with less than 0.1 degree error and it's equipped with a sensitive Keil probe for pitot and static; we now have an effective "ground truth" source to compare Gen 2 performance against.


Wireless boom data receiver mounted to RV-4 roll bar.

Perhaps our biggest software update is a new wifi interface that will allow the pilot to adjust any system settings using an iPhone. As described in some previous posts, the way we currently calibrate the Gen 2 system is to fly GPS speed runs to derive one or more "aircraft curves," and then we establish "set points" for the various tones (L/Dmax, ONSPEED, etc.). Until the development of this interface, I had to fly each desired set point, record AOA data, land, adjust software and fly again--a painful, iterative process that consumed 100LL and time. With the new software, it's now practical to simply adjust settings real-time, while flying. We don't expect folks to be entering a third order polynomial whilst pulling G's, so much of the interface is just designed to simplify entering settings (no need to use the Arduino software that we use to run the system code). In flight, only the "USE LIVE AOA" buttons are applicable along with adjusting damping settings...and only during calibration. Once a system is dialed in, settings don't change. This is similar to other interface protocols used by commercial AOA manufacturers, but on steroids (i.e., lots more capability). And, one of our primary beta test objectives remains automating this entire process.


iPhone WiFi setting interface.

3D Audio


Evolution has bestowed a great capability in us primates to very precisely determine the direction a sound comes from. In the military, we tested systems that provided warning cues from the appropriate azimuth. It's the same reaction you have when you turn your head to face the direction that a sound comes from. For short-hand, we called this "3D audio cuing." We decided to add this capability to Gen 2 to "slew" the tone to follow the slip/skid ball. In other words, if the ball slides out of the race to the left, you hear the tone "slide" to the left in the sound field. Obviously, this requires stereo so Gen 2 either uses a stereo ICS or bluetooths to a stereo headset to provide this functionality. Pretty cool to be in the base turn, and "listen" to your rudder coordination without having to turn your head to glance at the ball!


High Alpha Performance


Recently, we started high AOA maneuver testing (stalls, spins and sustained stalls). We learned fairly quickly that our initial technique of capturing the aircraft curve with a 2nd order polynomial works well up to the stall, but falls short in a post-stall environment. Turns out, what the doctor ordered was a 3rd order fit. Fortunately, it's just math to the computer, so it's easy enough to implement an improvement that allows the system to accurately capture alpha up to about 40-45 degrees or so. This algorithm change will allow us to better capture AOA spikes as well as assisting post-stall. This is depicted in the two figures below:


2nd order fit. Note that curve eventually goes negative...
3rd order fit does a much better job of capturing alpha above stall, and doesn't trend negative.

There is another LIMFAC with a coefficient of pressure system as well: side slip angle. Engineers typically refer to sideslip as "beta" angle. Wind tunnel data has shown a representative coefficient of pressure probe (Alpha Systems) to work up to 6 degrees of beta angle. In flight test, we are observing the Dynon probe to work up to about 9-10 degrees. Pressure falls off with side slip, but does so proportionally; so the difference between pressures remains constant (which allows for accurate AOA computation). Currently, when maximum beta angle is exceeded, the tone simply goes away. Our plan is to incorporate the gyros on the IMU chip to seamlessly integrate a combined pressure/IMU derived alpha/beta solution. At lower angles, pressure will be the predominant input, and at higher angles, the algorithm will transition to inertial inputs. This capability is important if we want to use the AOA and sideslip signals to drive a post-stall recovery display.


Visual Display


I've finally said "uncle" as regards a visual display! While I think audio is the bomb in an operational environment where your head is on a swivel; why not simply provide both and allow folks to experiment with what they like best? Gut tells me that after a couple of trips around the pattern, the aural cue will be a favorite; but different strokes for different folks. The visual display may be helpful in event the pilot becomes task saturated and begins to tune out audio cues.


Originally, we developed HUD software that would allow any typical Dynon/Garmin/AFS EFIS drive a HUD with AOA cuing. Due to the cost and complexity of a HUD installation, we've come up with Plan B courtesy of Vern Little. The result is something we are calling an "energy cuing display." It does just that: provide energy cues: airspeed, G and AOA in a simple, graphic format. AOA performance cues and trend information are provided in a standard, military format. The AOA display has a "fast" yellow chevron, a green ONSPEED "doughnut" and a red "slow" chevron along with the trend bar, L/Dmax "barbells" and % lift display. It uses an inexpensive, sunlight readable display (about $40) that can be mounted in the pilot's line of sight. We'll get some improved graphics and video posted soon; but here's a first look mounted on the glare shield of Cecil's RV-8:


Energy cuing display mounted on the glare shield

The "Texas Solder Team"


OK, they aren't all in Texas, but I like the way the name rolls off the tongue...We've got a small group of volunteers that are going to build Gen 2 systems from scratch to test and validate the assembly instructions. Our plan is to use the forum on this site for technical questions and lessons learned to not only document the process; but share lesson's learned.


0 views

© 2019 FlyONSPEED.org

Contact Us

team@flyonspeed.org

 

  • Mailing Address
  • 1202 Windward Circle
  • Niceville Florida  32578
  • YouTube Clean