What's new
Van's Air Force

Don't miss anything! Register now for full access to the definitive RV support community.

Angle of Attack and Energy Maneuverability

Vac

Well Known Member
Benefactor
"If you want to go up, pull back on the stick. If you want to down, pull back a little more. If you want to go down real fast, pull back harder..." -an old aviation proverb.​

Nothing makes a pilot's eye's glaze over faster than an energy management discussion; but ultimately it's failing to manage energy that brings many pilots to grief. There are pilots with extensive experience and a well calibrated “seat of the pants” feel for their airplanes. They can safely and masterfully fly the wing throughout the flight envelope, extracting maximum performance when required, and maintaining positive aircraft control. There are also pilots that lose control of their airplanes, kill themselves and their passengers during normal pattern operations. Surprisingly, more during the takeoff and initial climb phase than during approach and landing. What’s the difference? The first pilot has an intuitive feel for “flying the wing,” while the second does not. Sometimes, they are the same pilot. Pilots that experience loss of control aren’t restricted to any age, experience, or proficiency level. No pilot intentionally spins an airplane into the ground. They get into trouble due to misunderstanding or lack of information, or become distracted and miss a cue. This is understandable, because the two fundamental skills required to fly an airplane are controlling angle of attack to make lift and power to counter drag and make speed and altitude. There is no way to see angle of attack, and we don’t have a power required gauge in the cockpit if the airplane, so there is no direct feedback for the two most critical tasks a pilot must perform. Learning to fly is a complicated process of using surrogate information provided by flight and power instruments and visual cues to control the flight path of the airplane.

What if we could simply see and hear how hard the wing is working and how that work is trending? What if we automate “seat of the pants” feedback using modern technology and provide it to the pilot in a simple, intuitive manner? What if an ergonomic cue allowed the pilot to adjust both angle of attack and power to achieve key performance parameters, maintain positive aircraft control and avoid stalling when maneuvering? Not nearly as glamorous as wearing leather flying kit and listening to the wind in the wires, or as effective as fully automatic flight controls, but certainly better than what we have now.

To reduce loss of control mishaps and improve flying precision, the US military was an early adaptor of angle of attack (AOA) technology. These systems reached their maximum effectiveness with 3rd generation fighters when combined with energy maneuverability training as dog fighting re-emerged as a requirement during the Vietnam war. As fighters have advanced, AOA has been incorporated in automatic flight controls, in essence making the airplane smarter than the hairless ape flying it. Until GA flight controls are automated in a similar manner, we are much like the old 3rd generation fighters with manual flight controls, and could benefit from adapting some of the technology and teaching techniques that lowered military attrition rates and improved combat effectiveness over a half century ago.


Fly safe,

Vac
 
Last edited:
"If you want to go up, pull back on the stick. If you want to down, pull back a little more. If you want to go down real fast, pull back harder..." -an old aviation proverb.​
The point about dying at low Gs is really interesting. For some reason my brain wrongly assumed the “base to final stall/spin/crash/burn” scenario involved noticeable G loading.
 
we are much like the old 3rd generation fighters with manual flight controls, and could benefit from adapting some of the technology and teaching techniques that lowered military attrition rates and improved combat effectiveness over a half century ago.​



Fly safe,

Vac

I would say this is a huge understatement. AOA should be taught to all pilots as part of the PPL, and all aircraft should be fitted with an AOA indicator.

Just my opinion.
 
Watch the altitude and attitude during the banking demonstrations. If the attitude is keep level during bank you quickly hit stall with a G load of only 1.2.
 
Protest noted, what were loss of control mishap rates in 1946? :unsure:

Respectfully, I get it; which is why I stated "There are pilots with extensive experience and a well calibrated 'seat of the pants' feel for their airplanes. They can safely and masterfully fly the wing throughout the flight envelope, extracting maximum performance when required, and maintaining positive aircraft control." Unfortunately, even well trained experienced pilots end up losing control unintentionally, myself included. As old guys, we have to beware hubris and remember we are all one moment of inattention or misperception away from being a statistic in any airplane any time. AOA/power required cuing isn't a perfect answer, but it's step in the direction Wolfgang Langewiesche started marching when he wrote "Stick and Rudder" in 1944, two years before your Cub was built. You might notice in the video in the original post how fast things go from "OK" to "smoking hole" in the base turn departure scenario. I will stall at low altitude for flight test and demonstration purposes under controlled conditions to demonstrate the relationship between AOA and power required, but only in coordinated flight, since an uncoordinated stall below a couple hundred feed AGL would likely be non-recoverable. And then who would edit the video and type up lesson's learned? ;)


Respectfully,

Vac
 
Last edited:
Vac, you and I have emailed each other several times as I have attempted, and failed, to create an On Speed device for my RV-8 and SeaRey. I have installed a Huvver display for the AOA but I would really prefer to have the aural tone than the visual display as my eyes seem to be busy elsewhere. Still no option for completed units?
 
Great video Vac. Thank you. How do we get the On Speed hardware and software integrated into our current EFIS, in my case a GRT Sport SX?

One suggestion. I was initially confused by the references to slow tone vs fast tone. I understand now you mean the tones corresponding to airspeeds less than and greater than On Speed, but the actual tones also pulse at slow vs fast rates, the fastest pulsing rate being just before stall. So a fast pulsing tone is also the higher frequency “slow tone“ that corresponds to an airspeed slower than On Speed. Hope that makes sense. Anyway, I think it would be clearer and more intuitive to refer to the tones for slow and fast airspeed as the high and low frequency tones respectively so they don’t get confused with slow and fast pulsing tones. Just my two cents. Thanks again. I’m hoping to be able to incorporate On Speed into my RV in the near future.
 
Vac, you and I have emailed each other several times as I have attempted, and failed, to create an On Speed device for my RV-8 and SeaRey. I have installed a Huvver display for the AOA but I would really prefer to have the aural tone than the visual display as my eyes seem to be busy elsewhere. Still no option for completed units?
When you say "completed units", is it that you haven't been able to build the unit but have all the parts, or is it that you want an assembled unit? You live between the 2 most knowledgeable OnSpeed people there in Orlando, but if it's about putting the unit together, I can still help you over here in Texas.
 
Leighton Collins started Air Facts in 1939. Initially it was just a compilation of accidents. The accident rate in that era was awful. Collins talked a lot about the prewar light aircraft ailerons being much more prone to causing inadvertent spins than the postwar ailerons.
Wolfgang wrote articles for Air Facts starting around 1940. Stick and Rudder evolved from those articles. Wolfgang went from being a Cub pilot to a production test pilot on the F4 Corsair in just a few years.
 
Hi David,

Thanks for the question. The team has been working on a new, refined hardware design that isn't based on the current Teensy 3.6 platform, but we still have to work the software. As for the current V3 configuration, we are willing to share what parts we have available, but unfortunately the long pole in the tent is the Teensy 3.6 CPU we are using--it's officially extinct and we are "bingo" (i.e., out of them).

Erich,

Thank you for the insight. Lemme' think on that for a bit--unfortunately I'm saddled with 30 years of nomenclature and experience (meaning I'm part of the problem, not the solution) but it's observations like yours that make effective communication happen!

As for your EFIS, we'd be happy to work with any manufacturer that is interested in working with us. Our group is a 501(c)3 non-profit, open-source organization. We have reached out to the major avionics manufacturers, including GRT and offered to help with implementation of the capabilities we've developed and tested, but no one has shown interest. I suspect part of this may be because of an understanding in general aviation that AOA is an adjunct stall warning "instrument" vs the key to performance cuing feedback that assists the pilot with maneuvering and "max performing" the airplane when required (e.g., takeoff and landing). This is essentially the same feedback loop employed by automatic flight controls. We are simply substituting our meat computers for the flight control computer. And because I'm a washed up Air Force guy that hauls cargo out of Hong Kong in the middle of the night, I'm sure another part of it is that I don't have the first clue about what it takes to integrate new tech into existing avionics hardware!

Observationally, I will point out that there are a lack of performance requirements and standardization in commercial AOA systems. Our work is only one way to skin the cat, but to provide this level of functionality, the system has to accurately measure AOA across the entire speed band of the airplane, have good "transient" response (i.e., the ability to handle G whether induced by the pilot or a gust of wind), and still provide filtered information to the pilot in a timely manner to avoid a PIO amongst other things. The system has to be as caveman simple to use as reading an airspeed indicator, and, calibration should to be automatic (or so dirt simple it would be impossible to have a bad calibration) so that the indications in my RV-4 are the same as the indictions in your RV-7.

I'm old enough to know what a significant paradigm shift it was in the military to change the way we equipped airplanes and trained pilots, so there is a lot of education between where we are and where modern tech may allow us to get. Honestly, the USAF and the Navy invert the "standard AOA indexer" cues to this day, so there is still room to do better after 60 years of trying. The first fly-by-wire Sling is being engineered now. Eventually, we'll get there, in steps. A first step is a "hack" like our system: provide the pilot with the two key bits of info (how hard the wing is working and how much power is required during maneuvering) required to "fly the wing." The second step is flight envelope protection provided by an automatic pilot. We already see examples of that in high end GA airplanes and the nifty avionics we have the privilege of experimenting with--things like the blue TruTrack/Honeywell "Auto Recover" button come to mind. And, ultimately if you think this is all unnecessary, you can still build a Sopwith Camel in your garage, just don't try to break right when the Red Baron rolls in on you if it's got a rotary engine and you forget to hit the "blip switch" :)

v/r,

Vac
 
Last edited:
Hi David,

Thanks for the question; unfortunately, no significant progress to report over the past 18 months or so on a follow-on hardware design. Our small, all-volunteer team is exactly one man deep in our design engineering and coding department and he has to rightly prioritize putting food on the table over putting new AOA hardware in our test birds. Like the paving on the off ramp to the Underworld, we still have every intention of doing that and creating a new hardware. We are willing to share what parts we have available, but unfortunately the long pole in the tent is the Teensy 3.6 CPU we are using--it's officially extinct and we are "bingo" (i.e., out of them). No excuse for the lack of progress.

Erich,

Thank you for the insight. Lemme' think on that for a bit--unfortunately I'm saddled with 30 years of nomenclature and experience (meaning I'm part of the problem, not the solution) but it's observations like yours that make effective communication happen!

As for your EFIS, we'd be happy to work with any manufacturer that interested in working with us. Our group is a 501(c)3 non-profit, open-source organization. We have reached out to the major avionics manufacturers, including GRT and offered to help with implementation of the capabilities we've developed and tested, but no one has shown interest. I suspect part of this may be because of an understanding in general aviation that AOA is an adjunct stall warning "instrument" vs the key to performance cuing feedback that assists the pilot with maneuvering and "max performing" the airplane when required (e.g., takeoff and landing). This is essentially the same feedback loop employed by automatic flight controls. We are simply substituting our meat computers for the flight control computer. And because I'm a washed up Air Force guy that hauls cargo out of Hong Kong in the middle of the night, I'm sure another part of it is that I don't have the first clue about what it takes to integrate new tech into existing avionics hardware!

Observationally, I will point out that there are a lack of performance requirements and standardization in commercial AOA systems. Our work is only one way to skin the cat, but to provide this level of functionality, the system has to accurately measure AOA across the entire speed band of the airplane, have good "transient" response (i.e., the ability to handle G whether induced by the pilot or a gust of wind), and still provide filtered information to the pilot in a timely manner to avoid a PIO amongst other things. The system has to be as caveman simple to use as reading an airspeed indicator, and, calibration should to be automatic (or so dirt simple it would be impossible to have a bad calibration) so that the indications in my RV-4 are the same as the indictions in your RV-7.

I'm old enough to know what a significant paradigm shift it was in the military to change the way we equipped airplanes and trained pilots, so there is a lot of education between where we are and where modern tech may allow us to get. Honestly, the USAF and the Navy invert the "standard AOA indexer" cues to this day, so there is still room to do better after 60 years of trying. The first fly-by-wire Sling is being engineered now. Eventually, we'll get there, in steps. A first step is a "hack" like our system: provide the pilot with the two key bits of info (how hard the wing is working and how much power is required during maneuvering) required to "fly the wing." The second step is flight envelope protection provided by an automatic pilot. We already see examples of that in high end GA airplanes and the nifty avionics we have the privilege of experimenting with--things like the blue TruTrack/Honeywell "Auto Recover" button come to mind. And, ultimately if you think this is all unnecessary, you can still build a Sopwith Camel in your garage, just don't try to break right when the Red Barron rolls in on you if it's got a rotary engine and you forget to hit the "blip switch" :)

v/r,

Vac
Sounds like what we need is for us customers to request that On Speed functions be incorporated into the various EFIS products. Seems like it would be a great selling point and a distinguishing feature to whoever does it first. Anybody out there listening?
 

Pilot135pd and Vac, I have 2 Arduino boards from when I foolishly attempted to make them myself. I'd have to get the exact models when I'm at the airport tomorrow. Niceville, FL is a short 6+ hr drive from Orlando. Is there someone closer who could help me?

 

Pilot135pd and Vac, I have 2 Arduino boards from when I foolishly attempted to make them myself. I'd have to get the exact models when I'm at the airport tomorrow. Niceville, FL is a short 6+ hr drive from Orlando. Is there someone closer who could help me?

I used to live in Orlando (Hunter's Creek) and still own some homes there, but I'm 90% happy in Texas now. :ROFLMAO:

Vac might know someone who has time and is closer to you to help you, but you can't get nicer than Niceville, it's baked into the name !!
 

Pilot135pd and Vac, I have 2 Arduino boards from when I foolishly attempted to make them myself. I'd have to get the exact models when I'm at the airport tomorrow. Niceville, FL is a short 6+ hr drive from Orlando. Is there someone closer who could help me?

PM me. If you would be willing to mail the kits that you have I might be able to assist. I know the designer of the ori v3 circuit boards.

- larosta
 

Pilot135pd and Vac, I have 2 Arduino boards from when I foolishly attempted to make them myself. I'd have to get the exact models when I'm at the airport tomorrow. Niceville, FL is a short 6+ hr drive from Orlando. Is there someone closer who could help me?

I'd be happy to assemble and ops check them if you like. Final install is pretty straightforward. I work on this project with Vac up here in Niceville.

Bob
 
Hi David,

Thanks for the question; unfortunately, no significant progress to report over the past 18 months or so on a follow-on hardware design. Our small, all-volunteer team is exactly one man deep in our design engineering and coding department and he has to rightly prioritize putting food on the table over putting new AOA hardware in our test birds. Like the paving on the off ramp to the Underworld, we still have every intention of doing that and creating a new hardware. We are willing to share what parts we have available, but unfortunately the long pole in the tent is the Teensy 3.6 CPU we are using--it's officially extinct and we are "bingo" (i.e., out of them). No excuse for the lack of progress.

Erich,

Thank you for the insight. Lemme' think on that for a bit--unfortunately I'm saddled with 30 years of nomenclature and experience (meaning I'm part of the problem, not the solution) but it's observations like yours that make effective communication happen!

As for your EFIS, we'd be happy to work with any manufacturer that interested in working with us. Our group is a 501(c)3 non-profit, open-source organization. We have reached out to the major avionics manufacturers, including GRT and offered to help with implementation of the capabilities we've developed and tested, but no one has shown interest. I suspect part of this may be because of an understanding in general aviation that AOA is an adjunct stall warning "instrument" vs the key to performance cuing feedback that assists the pilot with maneuvering and "max performing" the airplane when required (e.g., takeoff and landing). This is essentially the same feedback loop employed by automatic flight controls. We are simply substituting our meat computers for the flight control computer. And because I'm a washed up Air Force guy that hauls cargo out of Hong Kong in the middle of the night, I'm sure another part of it is that I don't have the first clue about what it takes to integrate new tech into existing avionics hardware!

Observationally, I will point out that there are a lack of performance requirements and standardization in commercial AOA systems. Our work is only one way to skin the cat, but to provide this level of functionality, the system has to accurately measure AOA across the entire speed band of the airplane, have good "transient" response (i.e., the ability to handle G whether induced by the pilot or a gust of wind), and still provide filtered information to the pilot in a timely manner to avoid a PIO amongst other things. The system has to be as caveman simple to use as reading an airspeed indicator, and, calibration should to be automatic (or so dirt simple it would be impossible to have a bad calibration) so that the indications in my RV-4 are the same as the indictions in your RV-7.

I'm old enough to know what a significant paradigm shift it was in the military to change the way we equipped airplanes and trained pilots, so there is a lot of education between where we are and where modern tech may allow us to get. Honestly, the USAF and the Navy invert the "standard AOA indexer" cues to this day, so there is still room to do better after 60 years of trying. The first fly-by-wire Sling is being engineered now. Eventually, we'll get there, in steps. A first step is a "hack" like our system: provide the pilot with the two key bits of info (how hard the wing is working and how much power is required during maneuvering) required to "fly the wing." The second step is flight envelope protection provided by an automatic pilot. We already see examples of that in high end GA airplanes and the nifty avionics we have the privilege of experimenting with--things like the blue TruTrack/Honeywell "Auto Recover" button come to mind. And, ultimately if you think this is all unnecessary, you can still build a Sopwith Camel in your garage, just don't try to break right when the Red Barron rolls in on you if it's got a rotary engine and you forget to hit the "blip switch" :)

v/r,

Vac
I have to say that while I like the concept and routinely use aoa at work and in my -10, I have ZERO interest in a GA aircraft controlled by an automatic pilot using control laws. Dont get me wrong, when I fly my -10 cross country, I absolutely use the auto pilot. When I am just putting around, it is OFF. I want an aircraft that I can FLY not manage through an automatic pilot. It is kind of like buying a sports car with an automatic transmission; sure it can and does shift faster and more efficiently than the guy in the seat but that’s not the point. I want the visceral experience…balancing the clutch for a launch; shifting through the gears; listening to the huge IC engine rev and load. The same applies to aircraft, at least to me. I want the visceral experience. I want to feel the airplane respond. When you have an intermediary, you lose some of that experience. Can it potentially make flying safer? Maybe…but it can also cause problems. Hypothetical situation: the autopilot intermediary is programmed such that you cannot exceed any airframe limitations. Now you are put into a situation where you NEED to exceed those limits to survive, say a CFIT situation. Without the intermediary with its restrictive control laws, youUrvive but bend the aircraft. With the intermediary, you can pull as hard as you want but the autopilot will NEVER allow the aircraft to exceed its design limitation. It will, however, allow you to fly into cumulogranite, as long as you do not exceed any limitations.

Yes, hypothetical but it illustrates my point.

Again, just a different perspective in this world of flight. It won’t be the same when you do not have the final say as to what the aircraft does, and let an autopilot dictate what you can and cannot do.

Sorry for the long post but I keep seeing auto thus and auto that…kind of takes the fun and challenge out of the hobby.
 
I have to say that while I like the concept and routinely use aoa at work and in my -10, I have ZERO interest in a GA aircraft controlled by an automatic pilot using control laws. Dont get me wrong, when I fly my -10 cross country, I absolutely use the auto pilot. When I am just putting around, it is OFF. I want an aircraft that I can FLY not manage through an automatic pilot. It is kind of like buying a sports car with an automatic transmission; sure it can and does shift faster and more efficiently than the guy in the seat but that’s not the point. I want the visceral experience…balancing the clutch for a launch; shifting through the gears; listening to the huge IC engine rev and load. The same applies to aircraft, at least to me. I want the visceral experience. I want to feel the airplane respond. When you have an intermediary, you lose some of that experience. Can it potentially make flying safer? Maybe…but it can also cause problems. Hypothetical situation: the autopilot intermediary is programmed such that you cannot exceed any airframe limitations. Now you are put into a situation where you NEED to exceed those limits to survive, say a CFIT situation. Without the intermediary with its restrictive control laws, youUrvive but bend the aircraft. With the intermediary, you can pull as hard as you want but the autopilot will NEVER allow the aircraft to exceed its design limitation. It will, however, allow you to fly into cumulogranite, as long as you do not exceed any limitations.

Yes, hypothetical but it illustrates my point.

Again, just a different perspective in this world of flight. It won’t be the same when you do not have the final say as to what the aircraft does, and let an autopilot dictate what you can and cannot do.

Sorry for the long post but I keep seeing auto thus and auto that…kind of takes the fun and challenge out of the hobby.
Just push the little AUTOPILOT DISCONNECT button and you'll be fine.
 
I commend those that contributed to this project. I am surprised it hasn’t gained any traction with any of the avionics companies. I guess they don’t see a profit in it or see a marketing opportunity against their competition.
I think it would be an excellent addition to aircraft piloted by the average Joe who doesn’t have the same “zest” for flying and operating a machine as many of us do.
I agree with Bob. FADEC, SDS, AOA systems…., all do a better job than I can and probably add efficiency and improve safety. However, the day they take my knobs away, or a computer tells me what to do, this hobby is over for me.
There is a place for it, just not in my airplanes.
There’s another thread going on about slide rules and whiz wheels. Pilotage is a lost skill. Someday, manually operating an aircraft may go the same way….
I am a bit of a hypocrite as I do have a starter, electrical system, EMS, GPS, Fuel Computer etc…. But when it comes to stick and rudder skills, I’m in Bobs camp.
 
I got your point but I can hand fly all day in all positions then have an autopilot and only use it when I'm flying home straight and level tired. They're not mutually exclusive.
Again, you are missing the point.

You are describing what we have now. What has been suggested is a system where you “fly” THROUGH an autopilot all of the time, letting it control the aircraft based on programmed laws.

If that is what you like, go for it but dont call it flying; you are then nothing more than a manager who makes suggestions.
 
Well, the F-22 has fully automatic flight controls and is an absolute joy to fly; and so is the 777 I fly at work. The older I get, the more I appreciate the assist, but I was dead serious about my "you can still build a Sopwith Camel in the garage" comment although the Red Baron part was tongue firmly implanted in cheek. That's the glory of experimental aviation, which now constitutes about 30% of the GA fleet and continues to grow. Legacy aircraft are now unaffordable for the average pilot, and don't offer more than better avionics than they did 30 years ago. So, having said that, respectfully request we veer back to AOA/EM :cool:

Here's a short demo of what happens to the power required vs power available curve when we maneuver. The takeaway, is that at 1 G, we will run out of AOA on the backside of the power curve before we run out of power, but as G load increases, we eventually run out of power before we run out of G. This shift in the power required curve is essentially the heart of energy maneuverability math. I'm flying the demo without any airspeed indications in the cockpit, and the AOA indexer only provides AOA info, so there is no G displayed. The accelerated stall occurs at about 2.5 G's in this example.


Technically, what we are talking about is "excess specific power" which is abbreviated "Ps" and pronounced "P sub S." No matter what airplane we fly, eventually we run out of power when power available equals power required. The heart of this is the basic relationship between thrust and drag--two of the four forces we learned about in flight school. The math is actually pretty simple: (Thrust x Velocity) - (Drag x Velocity) / Gross Weight x G. From a pilot perspective, it's Thrust-Drag/Effective Weight. "Effective weight" is how much lift I need right now . At an on speed condition, you'll note that thrust and drag are balanced for the current effective weight. It's easy to visualize this relationship in level flight like the demo in this post, but if you go to 5:40 in the original post, you'll see a flight path effect--we can get thrust (or drag) from Mother Earth as well. In the demo in the OP, I don't adjust the throttle, but allow the velocity vector (where the airplane is GOING, not necessarily where its POINTING) to drop below the horizon as I increase G. That provides the additional airspeed I need to maintain an on speed AOA condition. While this sounds complicated, in practice the purpose of the tone/display is to do the cockpit math: you adjust power or angle of attack based on what you hear and flight path by looking out the window.

v/r,

Vac
 
Last edited:
Well, the F-22 has fully automatic flight controls and is an absolute joy to fly; and so is the 777 I fly at work. The older I get, the more I appreciate the assist, but I was dead serious about my "you can still build a Sopwith Camel in the garage" comment although the Red Baron part was tongue firmly implanted in cheek. That's the glory of experimental aviation, which now constitutes about 30% of the GA fleet and continues to grow. Legacy aircraft are now unaffordable for the average pilot, and don't offer more than better avionics than they did 30 years ago. So, having said that, respectfully request we veer back to AOA/EM :cool:

Here's a short demo of what happens to the power required vs power available curve when we maneuver. The takeaway, is that at 1 G, we will run out of AOA on the backside of the power curve before we run out of power, but as G load increases, we eventually run out of power before we run out of G. This shift in the power required curve is essentially the heart of energy maneuverability math. I'm flying the demo without any airspeed indications in the cockpit, and the AOA indexer only provides AOA info, so there is no G displayed. The accelerated stall occurs at about 2.5 G's in this example.


Technically, what we are talking about is "excess specific power" which is abbreviated "Ps" and pronounced "P sub S." No matter what airplane we fly, eventually we run out of power when power available equals power required. The heart of this is the basic relationship between thrust and drag--two of the four forces we learned about in flight school. The math is actually pretty simple: (Thrust x Velocity) - (Drag x Velocity) / Gross Weight x G. From a pilot perspective, it's Thrust-Drag/Effective Weight. "Effective weight" is how much lift I need right now . At an on speed condition, you'll note that thrust and drag are balanced for the current effective weight. It's easy to visualize this relationship in level flight like the demo in this post, but if you go to 5:40 in the original post, you'll see a flight path effect--we can get thrust (or drag) from Mother Earth as well. In the demo in the OP, I don't adjust the throttle, but allow the velocity vector (where the airplane is GOING, not necessarily where its POINTING) to drop below the horizon as I increase G. That provides the additional airspeed I need to maintain an on speed AOA condition. While this sounds complicated, in practice the purpose of the tone/display is to do the cockpit math: you adjust power or angle of attack based on what you hear and flight path by looking out the window.

v/r,

Vac
I appreciate what you are doing with aoa and the information is valuable and pertinent to the RV community.

We are NOT flying the F22 ( which I wish I could), or the 777. Yes, we can get back on topic…
 
Again, you are missing the point.

You are describing what we have now. What has been suggested is a system where you “fly” THROUGH an autopilot all of the time, letting it control the aircraft based on programmed laws.

If that is what you like, go for it but dont call it flying; you are then nothing more than a manager who makes suggestions.
I don't know who suggested that, looks like I missed it. Regarding your last comment, I guess since way back when the B-777 came out we don't have pilots anymore, just managers.

EDIT: I'll have to check between comments because I got the email about your comment and came here and answered, then saw the others who replied after you but before I got chance to. I guess we do agree on the 777 too.
 
Envelope 'Limiting' versus Envelope 'Protection' in Fly-By-Wire airplanes.

The Airbus FBW airplanes have envelope limiting. For example, if the pilot tries to exceed Vmo or Mmo, at Vmo +10 kts (IIRC) and at Mmo +0.05M (IIRC) the control laws command the airplane to automatically roll towards level and pitch up to slow to Vmo or Mmo, even if the pilot was pushing the sideskick full nose down. The pilot's control input is ignored once past defined limits.

The Boeing 777 (at least the originals, not sure about the latest 777X versions) have envelope protection. The pilot can exceed Vmo or Mmo if needed, but the control laws cause the stick forces to dramatically increase requiring the pilot to push harder to go faster when the airplane exceeds Vmo or Mmo, giving the pilots that feedback. Likewise, if the pilot rolls more than 35° of bank, the roll forces increase dramatically and the pilot has to apply more force to continue rolling past 35°. The FBW control laws do not limit the bank angle, but provide definite feedback to the pilot. So the pilots can do what they feel is necessary to fly the airplane.

Note that Envelope 'Limiting' or Envelope 'Protection' is available in the 'Normal Law' (Airbus) or 'Normal Mode' (Boeing). In the degraded modes of each company's FBW control systems, not all the envelope limiting or envelope protections are available.
 
Last edited:
Hi David,

Thanks for the question; unfortunately, no significant progress to report over the past 18 months or so on a follow-on hardware design. Our small, all-volunteer team is exactly one man deep in our design engineering and coding department and he has to rightly prioritize putting food on the table over putting new AOA hardware in our test birds. Like the paving on the off ramp to the Underworld, we still have every intention of doing that and creating a new hardware. We are willing to share what parts we have available, but unfortunately the long pole in the tent is the Teensy 3.6 CPU we are using--it's officially extinct and we are "bingo" (i.e., out of them). No excuse for the lack of progress.

Erich,

Thank you for the insight. Lemme' think on that for a bit--unfortunately I'm saddled with 30 years of nomenclature and experience (meaning I'm part of the problem, not the solution) but it's observations like yours that make effective communication happen!

As for your EFIS, we'd be happy to work with any manufacturer that interested in working with us. Our group is a 501(c)3 non-profit, open-source organization. We have reached out to the major avionics manufacturers, including GRT and offered to help with implementation of the capabilities we've developed and tested, but no one has shown interest. I suspect part of this may be because of an understanding in general aviation that AOA is an adjunct stall warning "instrument" vs the key to performance cuing feedback that assists the pilot with maneuvering and "max performing" the airplane when required (e.g., takeoff and landing). This is essentially the same feedback loop employed by automatic flight controls. We are simply substituting our meat computers for the flight control computer. And because I'm a washed up Air Force guy that hauls cargo out of Hong Kong in the middle of the night, I'm sure another part of it is that I don't have the first clue about what it takes to integrate new tech into existing avionics hardware!

Observationally, I will point out that there are a lack of performance requirements and standardization in commercial AOA systems. Our work is only one way to skin the cat, but to provide this level of functionality, the system has to accurately measure AOA across the entire speed band of the airplane, have good "transient" response (i.e., the ability to handle G whether induced by the pilot or a gust of wind), and still provide filtered information to the pilot in a timely manner to avoid a PIO amongst other things. The system has to be as caveman simple to use as reading an airspeed indicator, and, calibration should to be automatic (or so dirt simple it would be impossible to have a bad calibration) so that the indications in my RV-4 are the same as the indictions in your RV-7.

I'm old enough to know what a significant paradigm shift it was in the military to change the way we equipped airplanes and trained pilots, so there is a lot of education between where we are and where modern tech may allow us to get. Honestly, the USAF and the Navy invert the "standard AOA indexer" cues to this day, so there is still room to do better after 60 years of trying. The first fly-by-wire Sling is being engineered now. Eventually, we'll get there, in steps. A first step is a "hack" like our system: provide the pilot with the two key bits of info (how hard the wing is working and how much power is required during maneuvering) required to "fly the wing." The second step is flight envelope protection provided by an automatic pilot. We already see examples of that in high end GA airplanes and the nifty avionics we have the privilege of experimenting with--things like the blue TruTrack/Honeywell "Auto Recover" button come to mind. And, ultimately if you think this is all unnecessary, you can still build a Sopwith Camel in your garage, just don't try to break right when the Red Baron rolls in on you if it's got a rotary engine and you forget to hit the "blip switch" :)

v/r,

Vac
I emailed GRT support today suggesting incorporation of OnSpeed functions into their EFIS models and got this response:
“I will make a note in our to-do list. We just haven't had a lot of time to try these things.”
Perhaps we can get some movement on this with a few more similar requests! Just takes a minute people..
 
Thank you Carl!

Now let's explore another facet of AOA... performance cues. Obviously, there is no better form of "stall warning" than being able to watch (or hear) your angle of attack approaching critical; but now that we understand that we can get energy feedback as well from AOA, what else can we do with this information? Quite a bit, it turns out. A sufficient amount to "max perform" the airplane without a need for an airspeed reference in the cockpit. That's because airspeed is simply a surrogate for AOA in most cases.


I think this is a succinct demonstration of takeoff, landing and efficient maneuvering that demonstrates most of the goodness an accurate AOA system provides.

In the demonstration, I'm using the AOA system to determine rotation, best angle of climb initially, best rate of climb for climb to simulated engine failure, on speed for maximum sustained turn rate, maximum endurance glide, and approach and landing. Note there are no airspeed indications in the cockpit. I'm flying the airplane solely by reference to the AOA system. In this case, the visual display in the cockpit is AOA indexer-only. The optional visual display has 5 pages available: the "energy" display in the previous videos that combines airspeed, G and AOA information, an AOA indexer only page, an attitude indicator page used to verify proper operation of the internal IMU, and deceleration and historic G displays used for flight test.

This diagram plot of 1 G power available vs power required shows important airspeeds (AOAs):

Screenshot 2024-03-18 at 5.20.05 AM.png


During takeoff, rotation and lift off occur on speed. You'll see in the video that as the airplane accelerates during takeoff, you hear the tone pattern in reverse. This makes sense since we start the takeoff roll in a stalled condition. One of the pressures we are using to measure AOA is dynamic (pitot) pressure--the same pressure that drives the airspeed indicator. Thus, as pressures build up during the takeoff run, the system wakes up and begins computing AOA. This wouldn't be practical if AOA was measured with a vane. Note at 0:21 in the video the momentary "slow" indication as the airplane transitions to flight. This is the feedback I use to adjust pitch to control AOA in the transition to initial climb. After lift off, I maintain an on speed condition for best angle of climb until I retract the flaps about the time I'm just above the trees visible left and right of the runway. If you look at the diagram above, you'll see maximum excess thrust is available on speed. Best angle of climb occurs when you have maximum thrust available from your propeller. After flap retraction, I adjust pitch/AOA to fly L/Dmax. Notice the proximity of L/Dmax (Flaps 0) to maximum excess power. Best rate of climb occurs at maximum excess power. I don't do a super job of precisely holding L/Dmax because I've go much excess power in the RV-4 I can afford to be sloppy. Which is a good tactical problem to have.

I'm using an iPhone app that is updating the physics required to determine minimum turn-back altitude at 1Hz. My 1/2 Hz brain, Mark I eyeball isn't sharp enough to that with the same precision. Once the smart phone has done the heavy lifting and I turn back for landing, all I do is fly an on speed condition and maneuver the airplane. All thrust is provided by gravity when the engine fails. Astute readers will note in the diagram that is not an L/Dmax condition, where best range and best range glide occur. They will also note that is not Vmin where maximum instantaneous turn performance occurs. You would think that if you wanted to extract maximum performance, the right answer would be a "bat turn," followed by maximum range glide prior to transitioning to landing. You wouldn't be wrong, but I'm not that good a pilot to accomplish that a couple hundred feet above the ground during a maneuver that will be over in about 30-45 seconds. The reason I compromise and fly on speed is:

1. It's s single cue I can use throughout maneuvering, approach and landing.
2. It is a maximum endurance glide condition.
3. It's maximum sustained turn rate.
4. It's the AOA I use every day for approach and landing.
5. It provides sufficient "aerodynamic margin," i.e., distance from the aerodynamic limit of the envelope--just another way of saying "stall margin." I can't stall in an on speed condition.

Not perfect, but the 90% solution.

Fly safe,

Vac
 
Vac, can you explain fractional lift? Sounds like the higher the number the less lift available. Also the 1st vid shows the departure on a base to final turn. What was the actual loss in this departure? Speed, angle or g’s?
 
Vac, can you explain fractional lift? Sounds like the higher the number the less lift available. Also the 1st vid shows the departure on a base to final turn. What was the actual loss in this departure? Speed, angle or g’s?
Hi Blain,

Great question! You are correct, but more like “the higher the number the harder the wing is working.” Another way to think about it is “percent lift.” Just prior to the stall, the wing is working at 100%. There are three key performance parameters that can also be expressed in fractional lift: L/Dmax (50%), on speed (60%) and maneuvering speed (100 divided by the G limit of the airplane). Each of these conditions is also an AOA so fractional lift is just another way to display AOA.

The departure (from controlled flight) is caused by exceeding critical AOA (we can stall at any airspeed in any attitude). The snap roll is caused by adverse yaw (in other words an uncoordinated turn). I like this demonstration because the low G’s associated with the accelerated stall are imperceptible. Here are some properly coordinated accelerated stalls at low G. You can see the airspeed, AOA and G on the display. After stalls, there are some more examples of “skidding departures,” rudder in the direction of turn and opposite aileron. The last couple demos are “slipping departures,” aileron in the direction of turn and opposite rudder. The last one is a gee whiz oscillatory condition. It demonstrates how the system works under dynamic conditions with sideslip present. The key takeaway is that if you watch that demo, you’ll see the system always detects the stall. The bottom line is that that it’s very easy to loose control in a skid, and hard to loose control in a slip—you have to ignore serious feedback from the airplane in a slip (but it WILL depart :unsure:). In a skid, the airplane snaps as soon as it stalls.


Fly safe,

Vac
 
Last edited:
Let's take a closer look at fractional lift...There is an AOA at which the wing produces no lift. Engineers call this the “zero lift line.” A cambered airfoil achieves zero lift at negative alpha, about -1.3 deg for my RV-4 at flaps 0, whereas a symmetrical airfoil is at zero lift when AOA equals zero. If we "unload the wing" and go to zero G, we aren't asking the wing to produce any lift. On the other hand, if we "load the wing" and pull as hard as we can while still maintaining control (just prior to the stall and loss of longitudinal stability), the wing is generating 100% lift. This simple “0-1” scale is another way to think about how hard the wing is working. It’s also another way to display AOA to the pilot.

Here's an example of an old-fashioned AOA gauge that displays fractional lift on a 0-1 scale. Note that an on speed condition (.6 fractional lift) is positioned at "3 o'clock," i.e., the needle is horizontal for approach and landing. The military used different styles of round dials when jets still had steam gauges, and even different non-dimensional units to display AOA, but on speed was always at "right 3" to keep things standard and easy to interpret. Even on this simple gauge, we can see some key performance parameters, including best angle of climb, maximum endurance glide, best range AOA, best range glide, approximate best rate of climb; and most importantly, approach/landing AOA/zero excess specific power ("p sub s").

Screenshot 2024-03-24 at 2.22.13 PM.png

Some manufacturer's use a variation of the round AOA display on the EFIS, Cirrus being a good example. You can see the "0 to 1" round dial on the left and the "green doughnut" on the airspeed tape (red arrow):

Screenshot 2024-03-24 at 2.27.00 PM.png


We've chosen to combine a fractional lift display with an old-fashioned military style "doughnut/chevron" display...just another way to skin the cat, not necessarily better, just different. Although the "green doughnut" is standard, the USAF and the USN approach it from opposite directions, so even after 60 years, we still have some standardization work to do! We use standard USAF convention and light the lower chevron first as the airplane slows down. We've also added a trend indicator (the horizontal white bar that moves up as AOA increases) as well as a visual cue for an L/Dmax condition. These are the white balls adjacent to the lower chevron--when the white bar aligns with the balls, the airplane is at L/Dmax.

Screenshot 2024-03-24 at 3.18.05 PM.png


Let's revisit the 1 G power required curve and see how the various AOA performance cues correlate with the airspeeds we are used to flying:

Screenshot 2024-03-24 at 3.41.21 PM.png


So all fractional lift is just another way to think of AOA, and certain fractions of "how hard the wing is working" are associated with some key performance conditions:

Stall Warning (90% lift)
1. Maximum instantaneous turn rate
2. Stop pulling or something bad is going to happen

On speed (60% lift)
1. Vapp (approach and landing AOA)
2. Zero excess specific power
3. Maximum sustained turn rate
4. Maximum endurance Glide
5. Optimum maneuvering condition at low altitude (best blend of turn performance, glide performance, aerodynamic (stall) margin, and optimum energy for landing transition)

L/Dmax (50% lift)
1. Maximum range
2. Maximum range glide
3. Approximate best rate of climb

Carson's Cruise (22% lift)
1. Optimum cruise speed (maximum "MPG" or the least wasteful way to waste gas)

Maneuvering Speed/Corner Velocity (100/G limit)
1. Minimum % lift at which one control surface can be deflected one time using maximum rate input without incurring structural damage to the airframe

Fly safe,

Vac
 
Let's talk...airspeed.

I was discussing this NTSB Final Report after a colleague reached out to me to discuss RV-4 stall characteristics. Data contained in the report are derived from ADS-B, and the number that caught my eye was a 50 KCAS turn from base to final. The report did not indicate if the airplane was equipped with stall warning, and there is no requirement to do so for EAB aircraft. If the plotted ADS-B data in the report are correct, it doesn't take much maneuvering/G at this condition to exceed the aerodynamic limit. If we expand the corner of the RV-4 flight envelope for this area of operation, it looks like this:

Screenshot 2024-03-30 at 5.40.20 PM.png


Sorry to mix apples and oranges, but in the plot the X-axis is airspeed in miles per hour, whereas the report data are knots. Notice that at 50 KCAS, the airplane was already in a distinctly "slow" (high AOA) condition, and the G load required to stall the airplane is quite low. As demonstrated in some of the videos above, this would be effectively imperceptible. With rear cockpit occupied as it was in this case, the aft CG would also impact handling characteristics. If the pilot used any right aileron to counter a left wing drop at stall, an almost instant snap roll would be the result as it is in this demonstration:


Whatever actually occurred in the cockpit, both occupants were killed when the airplane impacted terrain. The pilot was ATP-rated.

One of the other ONSPEED team members developed an interesting presentation using a familiar, certified training airplane, a Cessna 172. What I like about this presentation, is that it uses all of the data contained in the POH, which actually isn't too much:

Screenshot 2024-03-30 at 5.52.25 PM.png

Screenshot 2024-03-30 at 5.52.46 PM.png
Screenshot 2024-03-30 at 5.53.09 PM.png

Screenshot 2024-03-30 at 5.53.23 PM.png
Screenshot 2024-03-30 at 5.53.37 PM.png
Screenshot 2024-03-30 at 5.53.55 PM.png
Screenshot 2024-03-30 at 5.54.09 PM.png
Screenshot 2024-03-30 at 5.54.26 PM.png
Screenshot 2024-03-30 at 5.54.41 PM.png


Continued in next post...
 
Last edited:
I emailed GRT support today suggesting incorporation of OnSpeed functions into their EFIS models and got this response:
“I will make a note in our to-do list. We just haven't had a lot of time to try these things.”
Perhaps we can get some movement on this with a few more similar requests! Just takes a minute people..
Just curious as I am not fully up to speed on the "OnSpeed" implementation ...

Would this be software that GRT would incorporate into their EFIS and REPLACE the "AOA" that they have now (calculated or measured) or would it be to "ENHANCE" the software that they have for that already.

The reason I ask is so that if I approach key people there, I am speaking with a reasonable understanding of how much "work" I am asking them to do (adding to a LOOONNG list of things I have already suggested. :) )

Thanks!
 
Screenshot 2024-03-30 at 5.54.54 PM.png
Screenshot 2024-03-30 at 5.55.13 PM.png


Fly safe,

Vac
Just curious as I am not fully up to speed on the "OnSpeed" implementation ...

Would this be software that GRT would incorporate into their EFIS and REPLACE the "AOA" that they have now (calculated or measured) or would it be to "ENHANCE" the software that they have for that already.

The reason I ask is so that if I approach key people there, I am speaking with a reasonable understanding of how much "work" I am asking them to do (adding to a LOOONNG list of things I have already suggested. :) )

Thanks!
Hi James,

We are an independent group and have developed our own hardware and software. We are not affiliated with any avionics manufacturer, but are trying to raise the bar for the performance and standardization of GA AOA systems as well as demonstrating the utility of an accurate system with good transient response and automated calibration. We are a non-profit, all volunteer, open source group. You can checkout our website at FlyONSPEED.org.

v/r,

Vac
 
I'm currently 180 NTSB reports into a planned study of 480 NTSB reports on takeoffs and landings, covering a two year period. Preliminary results are that AOA-related events, either stall/spin or events where a visual AOA might have helped, are in fact quite rare. Also, glides to a precise forced landing spot don't seem to have been addressed in AOA-for-guidance videos. An earlier finding, substantially reinforced by results so far, is that many "takeoff" events are better described as failure to achieve sustained flight.

One flight was a successful 180° turn after engine failure after takeoff, but after expending all the altitude on the turn, it was not possible to find a suitable forced landing spot. Sadly, it was fatal. I'm not sure that this kind of scenario has been discussed much.

Forced landings straight ahead after engine failure on takeoff handily outnumber impossible turn events.

300 more reports to go. Hopefully I'll make sense of all this for an Oshkosh presentation.
 
Hi Ed,

Hopefully you will, and I look forward to reading it. I hope that knee is feeling better!

As far as I know, we don't have any data about successful turn-backs :), they don't generally get reported; but we do have the science we did in our EAA 2022 simulator experiment in addition to the NTSB data. One of the conclusions of that drill and data analysis similar to the one you are working on now is that any sort of power loss on takeoff is likely the most "it depends" scenarios the average pilot may encounter. Scarily, a large percentage of turn back attempt fatalities occur from attempts started below 200-300 feet. A turn back after power loss is a "no good alternatives straight-ahead or laterally, only considered in advance, no better options maneuver." A personal takeaway from lots of instrumented flight test is that no matter how good you think you are, how your airplane is instrumented, or how much you practice, it's still a "gutsy move, Mav." One interesting variable we don't always consider is residual thrust during training/practice, which has to be accounted for, but the single biggest takeaway (assuming the airplane has sufficient performance, which most RV's do and many common certified trainers do not) is how absolutely critical the wind is. We can build a small amount of "slop" into the assumptions, but not much, and just a couple of knots or a few degrees of azimuth change can mean the difference between success or failure. The bottom line is that I'm not a good enough pilot to do it without technical help from a planning app that updates at 1Hz after I start moving and an accurate AOA system that provides constant feedback when I maneuver and doesn't require me to look inside the cockpit. Sufficient that I don't need an airspeed reference and can "max perform" the airplane, if required, and "average out" overall performance by flying the airplane at an optimum compromise condition, using only AOA performance cues. Obviously, a good pilot on their A-game could do that with conventional instrumentation, but it wouldn't take much of a gust or momentary distraction to ruin the day. Much nicer to know how hard the wing is working right now (and how that work is trending) vs math in public. Respectfully, we've got another educational thread called the "Impossible Turn" dedicated to that conversation: https://www.vansairforce.net/threads/the-impossible-turn.221622/, so that's a great place to add any lessons learned specifically regarding that scenario and join that conversation.

AOA "Human Factors" Points to Ponder: What's the most effective way to get required information to the human brain when the pilot is sucking up the seat cushion or just flying to get a 100 dollar burger?

“Directive” vs “Descriptive” Information. Conventional flight instruments present descriptive information to the pilot. That information must be interpreted, usually visually, and processed. Then the pilot must react. Assuming the pilot sees the cue, it takes about a half second to interpret and another half second to react; so, in a perfect world about a second to make a required adjustment. Once the pilot becomes task saturated, channelizes attention or becomes distracted, all bets are off. An AOA tone provides directive and descriptive information. Assuming the pilot hears the cue, they can respond in half of the time. And there is no requirement to look at anything or be perceptive enough to note an airspeed decay, for example. Like a flight director, the tone directs the pilot what to do with the pitch and power controls and is a more effective means to communicate critical information. A visual AOA display provides similar information when properly designed but requires scan time and brain bytes, some of which can be ameliorated with placement of the visual display in the pilot's predominate line of sight. A combination of the two is probably optimal (especially if the system is designed for performance cuing), but if only one was available, I'd want tone (with a big old volume knob :cool:).

Another way to think of this is to think of the impact of moving map displays. Would you rather have a fixed card ADF and a paper map or a GPS with a moving map display? It’s much simpler and intuitive to look at a moving map display on a smart phone or a panel mounted navigation display. This means of conveying information is so much more effective than old-fashioned navigation techniques that controlled flight into terrain accidents have been considerably reduced since widespread adaptation by GA.

Bottom Line. While speed considerations have an important place in properly operating your aircraft, there are times when a speed reference is inappropriate. Years of training to speed approximations for wing efficiency have created a culture that is inherently susceptible to dangerous flight excursions. Adaption of AOA (including requisite education and a degree of standardization of both displays and performance requirements) has the same potential to reduce LOC mishaps as moving maps have reduced CFIT mishaps. It worked in the military, a community that suffered a disproportionate number of LOC events until it changed the way it did business. I realize that GA flying isn't military flying, but an airplane is an airplane. I've had the privilege to fly and teach in both communities, and the difference in LOC mishap statistics speak for themselves. Control of AOA when maneuvering is a subject misunderstood by many pilots, regardless of background. AOA control is basically the same in any airplane regardless of type of power plant, wing form, size, or color. Pilots that experience loss of control aren’t restricted to any age, experience, or proficiency level. We either adopt some 21st Century technology to the "angle of incidence string" on the Wright Flyer, provide it to pilots and teach them how to use it, apply some lessons learned in other flying communities, or we keep making high AOA LOC smoking holes at the current rate.

v/r,

Vac
 
Last edited:
Hey Vac,

I'm looking at this and of course the issue with building one is the lacking teensy 3.6. I looked at the 4.1 but aside from the pins being in different locations it seems that it also lacks the DAC (what is needed to make the tones). Adding a DAC isn't that hard and a new board is already needed, but are those the only problems or are there software issues with this running on a teensy 4.1?

Also, the ESP32 that is needed by wifi isn't really much slower than teensy 3.6 and it already has a DAC on board... could this project be entirely implemented on an ESP32? I see those being around for a long time. What would be really great is to design a board that has the RS-232 chip, the power protection/regulator circuitry, some GPIO pin protection, the pressure sensors, and sockets for the ESP32 and the gyro as the gyro chips are a huge pain to solder....

I suppose I'm mostly curious as to what direction the project is heading hardware wise....
 
One of the issues that has been widely discussed and mostly ignored: how many successful turn backs have there been compared to fatal attempts. There is almost no documentation of the successful events.
The Pitts S1S has somewhat of a reputation as a streamlined brick. I totally disagree with that but it is certainly not a great glider.
In the era when a Pitts with a starter was very rare I watched two successful engine stopped landings on two consecutive days at a contest. Tailslide at a much too low altitude, prop stopped. Pilot made no attempt to land on the end of the runway but instead flew a base leg directly to approximately 1000' from approach end. He was maybe 10' AGL as he rolled wings level from a 45 degree bank. This guy was a airline pilot, very experienced aerobatic pilot but I would give him F on judgement for flying with a known problem that caused the two engine stoppages.
Most who have watched Bob Hoover in the Shrike or David Martin in the Beech Baron would agree that there is no need to die because of engine failure. AOA is a fantastic idea but none of the above had an AOA except maybe the one between their ears.
 
Akschu,

That’s exactly the tack we are taking for the new hardware design: ESP-32 based. We just ordered components to fabricate 5 systems for software development :). Tones are the heart of the system, so sound quality is paramount.

Jrs14855,

Excellent point regarding successful dead stick landings!

Cheers,

Vac
 
Vac,

Is the Gen 3 schematic on GitHub? I just designed a trim controller that is all surface mount and good for -40c to 85c (or something like that) with input power power protection that will survive a pretty harsh airplane environment. I wouldn’t mind helping out.

Schu
 
Vac,

Is the Gen 3 schematic on GitHub? I just designed a trim controller that is all surface mount and good for -40c to 85c (or something like that) with input power power protection that will survive a pretty harsh airplane environment. I wouldn’t mind helping out.

Schu
The current stuff is on github here....


Current hardware is Gen 2 Version 3. We are currently working on Gen 2 Version 4.
 
Thanks for the reply Bob,

I was looking for the experimental version where the teensy was eliminated and the ESP32 used instead.

Looking at this a bit more, I'm not sure that the esp32's DAC is good enough to produce decent audio. I saw some tests here:

Ultimately the person making the video started using the MAX98357 chip and passing audio over i2s.

Another option is this interesting library: https://github.com/tierneytim/Pico-USB-audio That seems to make decent enough audio with passive components.

Have you gotten reasonable tone output from the DAC on the ESP32?

Using the raspberry Pi pico with the library above might be a good option. I don't expect those to go away for a very long time and it has the wifi. Another option is the Pi Zero W. You would need to run a linux kernel, but it's not like you would need a full running system. The filesystem could be RO and init can call the binary directly. That gives you the ability to implement in python which might be easier and you would probably never again be dependent on a board being made.

Is there a better place to discuss development of this project? I'm not sure VAF is the right place for this....

schu
 
Thanks for the ESP32 tip, schu. I'll pass that along to the guys developing the next version of hardware.

[Update] Just checked with the hardware guys. A separate external DAC is part of the plan.

Bob
 
Last edited:
Hey Vac,

I'm looking at this and of course the issue with building one is the lacking teensy 3.6. I looked at the 4.1 but aside from the pins being in different locations it seems that it also lacks the DAC (what is needed to make the tones). Adding a DAC isn't that hard and a new board is already needed, but are those the only problems or are there software issues with this running on a teensy 4.1?

Also, the ESP32 that is needed by wifi isn't really much slower than teensy 3.6 and it already has a DAC on board... could this project be entirely implemented on an ESP32? I see those being around for a long time. What would be really great is to design a board that has the RS-232 chip, the power protection/regulator circuitry, some GPIO pin protection, the pressure sensors, and sockets for the ESP32 and the gyro as the gyro chips are a huge pain to solder....

I suppose I'm mostly curious as to what direction the project is heading hardware wise....
Teensy 3.6 can be bought on eBay but they are asking 180 to 400 dollars. You can see if they will take an offer.
Perhaps there is Vans AF member that is sitting on unused project. Might do a Classified.
I got mine from this forum.

Max
 
Akschu,

That’s exactly the tack we are taking for the new hardware design: ESP-32 based. We just ordered components to fabricate 5 systems for software development :). Tones are the heart of the system, so sound quality is paramount.

Jrs14855,

Excellent point regarding successful dead stick landings!

Cheers,

Vac
Great news Vac. Making hardware is major step forward.
Hopefully you will have a new prototype for test flights before years end.
 
I'm currently 180 NTSB reports into a planned study of 480 NTSB reports on takeoffs and landings, covering a two year period. Preliminary results are that AOA-related events, either stall/spin or events where a visual AOA might have helped, are in fact quite rare. Also, glides to a precise forced landing spot don't seem to have been addressed in AOA-for-guidance videos. An earlier finding, substantially reinforced by results so far, is that many "takeoff" events are better described as failure to achieve sustained flight.

One flight was a successful 180° turn after engine failure after takeoff, but after expending all the altitude on the turn, it was not possible to find a suitable forced landing spot. Sadly, it was fatal. I'm not sure that this kind of scenario has been discussed much.

Forced landings straight ahead after engine failure on takeoff handily outnumber impossible turn events.

300 more reports to go. Hopefully I'll make sense of all this for an Oshkosh presentation.
Keep the wing flying rather than the aircraft dropping like a rock is generally the preferred conventual wisdom.
Look forward to see you presentation at OSH too.
 
Back
Top