What's new
Van's Air Force

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

My RV-4 Learns an F-4 Trick

Vac

Well Known Member
Benefactor
The McDonnell F-4 had a nifty, aural AOA system that allowed you to hear the angle of attack when operating around L/Dmax and slower…it had a “solid tone” that indicated when you were ON SPEED, or flying optimum AOA, let you know when you were slow and really got your attention when you were approaching a stall. ON SPEED was especially helpful during approach and landing as well as maneuvering flight; and the logic allowed the pilot to make fine corrections between slightly slow, ON SPEED and slightly fast when desired. The F-4 didn’t have the greatest handling characteristics, and the system helped the pilot avoid “pulling on the pole” too hard and making bad things happen. I’ve been privileged to have flown lots of airplanes, including fighters and jumbo jets; and I’ve found the F-4 aural AOA system to be the best I’ve used in any airplane.

Recently, I met a couple of smart engineers and fellow RV'ers that have helped to build a prototype system that allows me to program the F-4 AOA tone logic for use in my RV-4 to demonstrate the concept to folks that might not be familiar with it. The “tone generator” is actually a small computer that processes the basic serial output message from my Dynon DY-10A, which includes AOA information from a standard Dynon AOA pitot tube I recently retro-fitted to the airplane. The tone is pumped into the intercom system and the volume is pilot-adjustable (except for the slow tone and stall warning). In the demonstration videos, the camera is receiving raw audio output and it’s much louder than what I’m hearing in the headset; and, to be clear, this is not the Dynon programmable AOA tone.

The AOA tone allows me to hear the “back side of the drag curve” or what you might have learned as the region of reverse command, i.e., where it takes more power to go slower. Optimum AOA lies in this region. At speeds above or AOA’s less than L/Dmax, there is no tone. The way it works, as the airplane slows down (or AOA increases) to L/Dmax, a slow, low frequency beep starts. This reference can be handy, since lots of good things happen performance-wise at L/Dmax. As the airplane slows down more and AOA increases to optimum, the pulse rate of the low-frequency tone increases and changes to a solid tone when you reach “ON SPEED,” or optimum AOA. If you keep pulling and increasing AOA, then the tone transitions to a high frequency and starts to beep again at a slow rate, letting you know that you’re “slow.” And if you insist on stalling, the pulse rate will increase until about 15% or so before the stall a loud, high-frequency quick pulsed tone (that sounds a lot like a stick or pedal shaker) lets you know that the bottom is about to fall out. Here's a picture that illustrates the AOA tone logic:
view
. Here's what that picture sounds like in the cockpit as the airplane slows down from L/Dmax all the way to the edge of the stall: https://youtu.be/S9H6T_tOLe4. The easiest way to conceptualize how this works is that the tone essentially tells you that you are "fast" (low frequency), "on speed" (steady tone), or "slow" (high frequency) when operating at lower airspeeds or higher AOA. The pulse rate allows you to fine tune pitch inputs using this basic logic.

Here’s what an ON SPEED base and final to a low-approach looks like: https://youtu.be/p7PXlOgJUJs. The video starts with the airplane rolling out on downwind and slowing down to begin the base turn. There is about 20 knots or so of over-shooting crosswind at pattern altitude, but only a few knots of right crosswind down near the runway. Prior to starting the base turn, the airplane is configured for landing (Flaps 40) and trimmed for ON SPEED. Throughout the base turn, AOA (tone) is controlled with pitch adjustments. During the turn, you’ll hear some instances of “slightly fast” (a fast-pulsed low frequency tone) and “slightly slow” (a slow-pulsed high frequency tone) as I explore the envelope a bit! Typically, until rolling out on final or approaching the roll-out point, power is at idle for a normal-sized pattern (traffic permitting); but in any case, glide path is controlled with power. Typically, a bit of power is added rolling out on final as the airplane settles “into the groove” for a stabilized approach. ON SPEED is maintained until the flare for a normal landing and you’ll hear a slow tone as I round-out during the low-approach, which is normal. It’s not necessary to look in the cockpit to hear the tone, so the pilot’s eyes are freed up to look at the runway environment and scan for traffic with an occasional cross-check of the airspeed indicator to make sure everything is working as intended. If conditions warrant, then it might be more appropriate to fly a “slightly fast” turn and approach. For example, if conditions are gusty and turbulent or you wish to carry a bit more energy for a wheel landing, etc. Similarly, if stopping distance is critical, establishing a “slightly slow” condition on final might be appropriate.

Currently, I’ve got the system programmed to wake up at 10 knots on takeoff, which lets me hear “what the wing is thinking” as the airplane accelerates; so it’s fairly easy to transition to an ON SPEED takeoff (which approximates maximum performance) for the initial climb segment when it’s necessary or advantageous to do so: https://youtu.be/OtLObb2b_us.

I'll add to this post as we experiment and learn more.

Fly Safe,

Vac
 
Last edited:
Very nice work on your progressive tone indicator. - You should be able to market that to Dynon users or license it to several of the glass manufacturers.

https://youtu.be/p7PXlOgJUJs Yup it looks like an F4 wallowing in off the end of a carrier deck. If the engine quit, I suspect you might end up a touch short of the runway.

Nice, but now lets see you do an "ON SPEED" approach with the mixture knob pulled out to cut off and a windmilling prop. :)
 
Last edited:
I use mine quite a bit since, as you know, there isn't nearly the pre-stall buffet to let you know you're on the gravel/rocks/boulders. It's awesome for getting max turn performance around the top of the egg.
 
Interesting use of available SkyView data. Keep us posted on how it continues to evolve. :cool:
 
I believe that audible tones are an excellent way to indicate AoA. The F4 audible system may be a tad complex for many, but IMHO it's superior to a visual system.

I've done a visual system and a haptic system (stick shaker) for D1x and SkyView systems using the serial data streams but the audible system has the most potential for those who use AoA to get maximum performance. For stall avoidance, however, I think AoA should add haptic feedback. A stick shaker or rudder shaker works well for that.

None of these methods work well, however, without proper practice or training.
 
Maneuvering Flight Application

The aural AOA also assists in maneuvering the airplane, not only in the traffic pattern or after takeoff, but during all phases of flight. The ON SPEED band is actually a key “max performance” point where lots of good stuff happens with turn rate, turn radius and energy management. This stuff is critical in a dog fight, but it’s also important if you are trying not to hit the ground, not stall, and effectively maneuver the airplane within a safe operating envelope. One of the easiest ways to describe how this works is to depict an “ON SPEED” band on a basic VN diagram that all pilots are familiar with. This drawing shows how the ON SPEED band looks for my RV-4 relative to the basic operating envelope for my plane:
view


There are few concepts depicted in this illustration worth pointing out…First, during 1 G, level, unacellerated flight, AOA and airspeed have a nice, linear and predictable relationship; but if you start to change any of those variables, then the relationship between indicated airspeed and AOA changes. In simplest terms, the AOA at which the airplane stalls doesn’t change, but the indicated airspeed at which the airplane stalls varies by load factor (G). This is why the left side of the envelope curves up from zero airspeed at zero G until it reaches structural limit at maneuvering speed (or “corner velocity” which is a term that simply refers to a corner of the envelope). The ON SPEED band mimics the behavior of the aerodynamic limit: as bank angle and/or load factor increase, so does stall speed and, thus the indicated airspeed associated with ON SPEED or optimum AOA. Here’s a good way to see how this works in real life…this is a video of a steep turn. If you look at the indicated airspeed on round dial, you’ll note it’s quite a bit higher for an ON SPEED condition, then when operating in slow flight or flying a gentle traffic pattern: https://youtu.be/BphHzWHbOjo. Throughout the turn, you can hear the tone vary slightly between ON SPEED, “slightly slow” (high frequency) and “slightly fast” (low frequency) as I make pitch input adjustments. The logic is designed to allow this “fine tuning” for relatively precise pitch/AOA control.

Besides the fact that lots of aerodynamic good stuff happens ON SPEED when you are trying to pull the nose around the sky, the best thing is that you will always encounter ON SPEED when you pull before you stall—sort of! Because the only automatic flight controls in an RV are between the pilot’s ears, it’s possible to pull the stick so quickly that bad things can happen without any sort of warning from the tone or even aerodynamic buffet…if I roll inverted at cruise speed and pull the stick as hard and fast as I can, I’ll likely depart controlled flight before I hear anything but the stall tone after the fact; but that’s only if I didn’t pull the tail off the airplane first. Now this is a worst-case scenario, but it illustrates the concept of how you can beat most warning systems if you try hard enough (or channelize your attention enough to not even be aware of the tone). But, if you maneuver at proper speed and use smooth flight control inputs, then the tone can save your bacon if you try to pull too hard. Here’s a video of me beating the system by performing a rapid pull into an accelerated stall: https://youtu.be/DLtamTAh-Is. In the first stall, you might notice that I don’t even hear stall warning before the nose stops tracking and stalls, just slow tone. The system actually has a chance to catch up, and by the time I induce the secondary stall, you hear stall warning.

Another bad scenario is when you overshoot final a bit, throw in some outside aileron to keep your bank from getting too steep close to the ground, but leave inside rudder applied—i.e., a skidding turn. If you were to continue to pull to the stall, an RV will typically snap roll underneath as soon as critical angle of attack is exceed. But in this case, the slow tone should alert you to need to ease your AOA as you look out of windscreen at an angled final, think about going around and are reminded to center the ball up. Here's a link to an old video of a skidding departure (it was part of a test to look at buffet effects on the horizontal stab in different post-stall conditions): https://drive.google.com/file/d/0B8EIT6g2n8o_bjE1VnJRXy1nU0k/view?usp=sharing. There was no aural AOA system installed in the plane at this time; and I'm only posting the video to demonstrate how rapidly the skidding departure (snap roll underneath) occurs after the critical angle of attack is exceeded. As we get deeper into testing the aural AOA tone system, I'll update the video showing the view out the front and how the tone behaves during the departure from controlled flight.

Here are a couple examples of ON SPEED maneuvering other than traffic pattern ops and a simple spin entry…Again, note that it’s normal to hear some variation in tone between ON SPEED, slightly slow and slightly fast when maneuvering. You may also note that there is limited stall warning prior to the spin (just slow tone), but I’ve since adjusted the software threshold to fix that; however it does give you some idea how the system works post-stall and during recovery.

Split S: https://youtu.be/kZJpk_hFIbg

Loop: https://youtu.be/0dkVtURCXiw

Incipient Spin: https://youtu.be/nGaJiHO_st8

I'm sure folks a lot smarter than a bunch of curious RV'ers are working on similar systems; so the key will be to develop maximum ergonomic utility across all phases of flight (not just approach and landing) while accommodating the widest range of pilots. And as Vern wisely pointed out in his post, none of this is worth much unless it's combined with good training resources.

I’ll post more as we continue to experiment with the system, and we're very interested in thoughts and feedback from folks in the community; so please post, drop a PM or e-mail.

Fly Safe,

Vac
 
Last edited:
Nice work, Mike. Thanks for sharing. I've always thought that something like this must exist somewhere.

The closest analog that comes to mind is the near-continuous audio feedback from a variometer in a glider which permits far more precise control of the aircraft than only including the VSI in the visual scan.

The only concern I would have is the same concern I would have with pretty much any other piece of equipment to be used during a critical phase of flight -- what are the failure modes? Thinking pessimistically, the worst-case failure for the aural AOA indicator would be that it sticks in the "on speed" mode and provides the pilot with bad situational awareness. I'm curious (from an engineering perspective here) how your team engineered your gizmo so that this cannot happen?

Again, great work! This is cool.
 
I'm sure folks a lot smarter than a bunch of curious RV'ers are working on similar systems; so the key will be to develop maximum ergonomic utility across all phases of flight (not just approach and landing) while accommodating the widest range of pilots. Vac

I wouldn't be so sure. I stopped with the F-4 so am not familiar with the follow on F-16's & F-15's AOA warning systems. The F-4 aural system was so user friendly but old technology, I wonder if the knowledge is still out there. After viewing your vids, you have mimicked the F-4 system perfectly. Time to market your system!
 
Hi Jamie,

That's a great question!

Short answer: A calibration tone is provided to the pilot when the system is turned on to let him know it's working (and adjust volume to desired level); and there is an LED that provides a "heartbeat" signal for the serial data stream; but the tone generator is just a computer processing the AOA signal from the EFIS, so garbage in, garbage out...IAS (and known pitch/power setting) serves as the back-up to confirm proper operation of the AOA.

Long answer:

Even for a system that had indigenous sensors (say adding pressure sensors to the computer in the tone generator, or replacing the pressure input with a voltage signal from a potentiometer driven by a vane or the signal from the nifty new friction-less magnetic sensors driven by a vane, etc.), the best way to determine AOA validity is to compare it to a known airspeed for the condition.

For approach and landing reference, AOA is either primary or secondary. If it's considered reliable enough to be primary, then the pilot normally computes a back-up IAS for approach and cross-checks throughout the approach to confirm that the tone is "valid." Similarly, if the AOA sensor/display is not considered reliable enough to be primary, then the pilot maintains Vref and uses the AOA signal to validate that. Factors that effect the validity of IAS may also affect AOA (e.g., icing conditions). The tertiary "catcher's mitt" is also comparing AOA/IAS to known pitch and power settings. So in this case, part of the "engineered" solution is training the pilot to compare what he sees with what he hears and also what he knows to be correct for pitch and power throughout the approach.

Part of our testing will be to look at lag time and make a determination whether or not the tone is reliable enough to be a primary indication. If you watch some of my approach testing video very closely, you can detect a bit of null (hysteresis) at the cross-over between ON SPEED, slightly slow and/or slightly fast because I'm actually making a fine pitch correction so quickly that I'm beating the system. On a stabilized approach, this effect is negligible and you have to be looking for it to even notice it. On the other hand, at high pitch-change rates, then lag can be sufficient to mitigate some (or all) of the warning provided by a system like this. That's demonstrated in the accelerated stall videos I posted above.

Interestingly, one of the things we considered was replacing with steady tone with a null tone--i.e., if things are going right, no tone present. The only problem with this logic is that there is already no tone at fast speeds/low AOA; and if a "null" tone fails, how do you know without looking at something? On the other hand, it would be practical from a software logic stand point to add a "hey dummy, the AOA data has stopped coming in" warning tone, so we'll take a look at that...thanks for the idea!

The bottom line is that as nifty as a good AOA tone can be, it's still not perfect and the automated RV flight control system (i.e., pilot) still has to process and evaluate the information.

Cheers,

Vac
 
Departure from Controlled Flight

We are beginning to test aural AOA system performance with side slip present, and as any pilot knows, post-stall yaw will contribute to a tendency for the airplane to depart controlled flight if not checked by the pilot. Let’s look at how the tone performs in an approach and landing environment by taking a crawl, walk, run approach to explore the left side of the envelope a bit…

The first video is a simple, coordinated, power-off approach to landing stall that occurs during a simulated base turn: https://youtu.be/dn3FgdXQUb0. The purpose is to demonstrate how the tone performs as the critical angle of attack is approached and exceeded, and no attempt is made to minimize altitude loss or make a “maximum performance” recovery.

The next video illustrates how the system performs when establishing a full “inside” slip ON SPEED during a simulated base turn. In this example, full aileron is applied in the direction of turn (left) and full rudder (right) is applied to the “outside” (opposite) of the turn. Pitch is adjusted to maintain an ON SPEED condition during the simulated base turn: https://youtu.be/C6NWKRIA8bM. Note the indications on the turn coordinator.

This video shows a normal pattern to a full-stop landing with a full slip applied to assist with glide-path control when starting the base turn. My RV-4 with a light-weight, fixed pitch prop has excellent glide performance, and it doesn’t want to come down too fast in the pattern! https://youtu.be/Td_7QMXtJcE.

Now what happens if you get a bit too aggressive, or, perhaps “tune out” the aural AOA tone? Well, it depends…If you are flying a coordinated airplane, you’ll likely get a simple stall with some wing drop as depicted in the first video in this post. However, if you get too aggressive while trying to slip, this is a possible out-come: https://youtu.be/RtwS_vXQukk. I’m avoiding any commentary so that you can just listen to the AOA throughout the maneuver. You’ll note during the initial stall (the first time critical AOA is exceeded) things really aren’t too much worse than a coordinated stall, however, maintaining the deep stall condition with crossed controls eventually causes a roll in the direction of yaw and if held long enough, a spin entry occurs. Again, the purpose of this is to demonstrate how the tone acts during this type of departure from controlled flight. Note the indications on the turn coordinator.

The next scenario is the most dangerous you’ll face in the pattern: exceeding critical angle of attack while making a skidding turn to final: https://youtu.be/tGItkZzTLUE. In this case, rudder is applied in the direction of turn and aileron is applied to the “outside of the turn” (i.e., left rudder and right aileron during a descending left-hand turn). Note the turn coordinator indications and listen to the AOA tone as the airplane slows below ON SPEED and critical angle of attack is exceeded. The departure is rapid and in the direction of yaw. You can imagine that if this occurred at low altitude, recovery can be problematic. You might note that I’m flying an (almost!) ON SPEED recovery following this departure, which means best “sustained” turn performance, or flying that green band depicted on the VN diagram in the previous post--which is the most energy efficient recovery. If, however, the ground was truly a factor, I’d have to pull right up to the aerodynamic limit while applying maximum power (start of the stall warning tone, or buffet, wing rock, nose rise (stick getting light), etc.—in other words whatever stall cue occurs first)…

If you browse the rest of the You Tube channel, you’ll find a couple other examples of slips or skids combined with a deep stall.

There is a discussion about cross controlled stalls in the transition training document that is maintained as a sticky at the top of this forum page (See the CROSS CONTROLLED STALL discussion beginning on page 327 for more information, if you are interested). From the "note, warning and caution" department, I want to point out that in my RV-4 noticeable buffeting of the horizontal stab occurs when flying a skidding departure in particular. This may, or may not be unique to my airplane. I’m only flying these test points to gather data and validate AOA tone performance through the stall, departure, post-stall and recovery sequence under these conditions. I don’t recommend forcing departures of this type due to possibility of inducing fatigue damage to the tail.

For folks that might be interested in the prototype hardware and software we are using to provide the tone, our engineer has set up an open source page here: https://github.com/dinglewanker/aoa-tone-efis-serial.

Fly Safe!

Vac
 
Last edited:
Stall Warning

The F-4 AOA system essentially provides the pilot with a couple of options as to how he or she would like to “listen” to the tone. If the pilot is interested in “max performing” the airplane, then the entire back side of the drag curve is presented, with the solid tone indicating optimum AOA and the ability to easily differentiate between a “slow”, ON SPEED or “fast” condition. If, on the other hand, all the pilot is concerned with is progressive stall warning (or just the “slow” indication), then it’s possible to adjust the volume to minimize or eliminate all tone for ON SPEED or faster conditions—in other words, the only the 1600 Hz high frequency tone is present and begins when 70% of available lift is used up. This function is similar to progressive stall warning audio currently available in other AOA systems such as those provided by Dynon or Garmin. If adjusted in this manner, then the technique used for flying ON SPEED for approach and landing is to simply slow to ON SPEED and then adjust pitch to occasionally touch the slow tone; so instead of a constant solid tone in the headset, no tone is heard unless the AOA increases into the slow range. Like all AOA systems, the F-4 system provides excellent stall warning, although at high G/AOA onset rates, there is some lag. In other words, if you pull hard enough, fast enough, you can “beat the system” and generate a stall before a stall warning tone is heard (although you will get some slow tone, even at high onset rates). This lag at high onset rates is best illustrated in the accelerated stall demonstrated in a previous post. In that example, the first stall occurs before stall warning is heard, but you’ll note be the second stall in the sequence, the system has “caught up” with pitch rate.

Let’s take a look at two very basic stalls to see how the stall warning portion of the tone works. The first thing to note is at the start of the slow tone, there is still 30% of the total lift available, so if a maximum performance recovery is desired, it is only necessary to break the angle of attack sufficiently to restore flying AOA, and the tone will catch up. In other words, the tone is simply a performance indication under these conditions (as it is in most, but not all, cases).

First we’ll look at a simple, power-off stall demonstration that begins at airspeed approximate to L/Dmax (the bottom of the drag curve) and uses a nominal 1 kt/sec deceleration rate: https://youtu.be/-2nBsFPnpT0. Keep in mind, that if the volume was turned down, only the slow tone would be present. Note that during the recovery, following the initial “unload” (reduction in AOA) to an ON SPEED condition, a momentary slow tone is heard since there is still “lift reserve” available and the objective was to minimize altitude loss following the stall (i.e., the type of maximum performance recovery that would be appropriate if the ground was a factor).

Next we’ll take a look at a basic power-on stall. I’m adding power carefully to control the deceleration and pitch prior to the stall to make things as stable as practical for the demonstration: https://youtu.be/QE0UoBK981k. Like the power-off stall, this demonstration uses a 1KT/sec deceleration rate. I’m cheating a bit end-game with some climb to control the deceleration rate to compensate for the addition of power. In this case, you’ll note the initial recovery is right to the edge of the stall warning (20% lift remaining point). In a power-on condition, it takes a bit longer for the AOA to stabilize during the recovery. So, once again, the AOA is providing performance feedback that must be compared with pitch, power and airspeed.

There are a few key points to understand about any AOA system. First, the system is only going to be as good as its calibration. Second, it is incumbent upon the pilot to understand the system and indications whether they are aural or visual (or a combination of both). Third, AOA indications should, generally, be used like any other “performance” vs a “control” indication. A simple mantra might be “pitch, power, AOA/airspeed” (or swap airspeed/AOA if you wish). However, under stable conditions (e.g., base turn, maximum performance sustained turn or catching ON SPEED during maneuvering, etc.) the AOA can become a de facto “control” indication for fine-tuning pitch inputs. During basic stall recovery, the pilot must use a combination of aircraft feel, airspeed increase and AOA indications to obtain a maximum performance/minimum altitude loss recovery. Some pilots may be familiar with an advanced instrument display called a “pitch limit indication” (PLI) which is a computer generated indication on the attitude display that looks at the pitch/power/airspeed/AOA conditions and does this math for the pilot. The MD-11 I fly at work does not display AOA on the primary flight display, but it does have a PLI, which can be helpful when flying out of a stall. If the ground isn’t a factor, it’s always possible to unload to a low g condition (push to make the tone go away), allow airspeed to increase and AOA to decrease, and then smoothly pull to ON SPEED for recovery (provided that part of the tone volume is turned up sufficiently to be heard).

Some folks have commented that the tone pattern is too complex, too annoying, or distracting in the headset. Unfortunately it’s extremely difficult to convey the utility of this type of system with a simple video demonstration; but for folks that are interested, I’ve compiled some of our test video to put together a brief 5 1/2 minute demonstration of how the system works in various flight conditions: https://youtu.be/5VUbTaeTLAM. With proper training and experience, using a system like the F-4 aural AOA becomes second nature, but it’s possible without that experience it may be difficult to appreciate the utility it offers. And, with all of the options available on the market, it really boils down to a matter of personal preference. The single most valuable contribution of any aural AOA system is that it frees up the pilot’s eyes so they can be out of the cockpit. If it’s properly mechanized, it provides valuable performance cues, especially in airplanes lacking easily interpreted aerodynamic (e.g., buffet) cues.

As we continue to test and develop the tone generating software, we’ll continue to look at ergonomic improvements; and if anyone has any suggestions, please post, drop a PM or e-mail. It would also be insightful for folks with experience with other systems to compare and contrast those systems with some of the examples we’ve posted—in other words, add a post to this thread with thoughts (pro and con) and/or experience with any of the many different options we RV’ers have when selecting an AOA system.

Fly Safe!

Vac
 
Last edited:
I really like what you are doing with this. I have an AFS Sport AOA and while the "Angle Angle Push Push" annunciation could be useful in an inadvertent situation, it does nothing to help me know if I am "on-speed" during approach. I have the light bar on the top left of my panel, and glance at it from time to time during approach, but the tone system would be a huge improvement. I believe there are a lot of the AFS systems out there, would be an fantastic add-on if this could somehow be incorporated into that system. Might be worth a discussion with AFS.
 
Back
Top