New auto pilots becoming available, but........

[dgapilot]

While an AP installation may sound easy there is a lot to it. Stability, flutter, control forces, effort to slip the clutches, and the list goes on. Just the flight test requirements, operate through the entire speed range and at all the extremes of the W&B envelope.

Remember with amateur built, there is a reason for those 2” letters that say EXPERIMENTAL. Every flight is an experiment.


Sent from my iPad using Tapatalk

[Keith Turner]

And every EAB aircraft is a one off.
The conventional approach to autopilot optimization is to measure the airplane response and develop the complex variable equations to describe those functions in terms of weight, CG, Q (dynamic pressure), control force and deflection and a bunch of second order terms. Those equations are then inverted, scaled and used in the control laws applied to move the surfaces. With modern solid state transducers used in the ADAHRS modules in Dynon and Garmin autopilots, the sensor performance is far superior to the gyro systems used in the first generation systems. Coupled with high bandwidth digital processing the Dynon and Garmin systems allow optimization through inflight tuning that results in very high levels of stability, performance and capability. I have a Dynon autopilot in my RV-7 that holds altitude and heading well inside IFR limits at 150 knots even in moderate turbulence. Considerably better than the gyro based analog system used in Mooney and Bonanza aircraft I flew in a flying club I once belonged to.
The bigger issue may be the physical installation of the servos, clutches and interfaces with the existing control circuits. That the Pacer and Tripacer don’t have mass balanced control surfaces is another area to consider in terms of control source stability and control law sensitivity. Making sure that all failure modes are accounted for and don’t result in any unacceptable handling conditions, assurance of being able to disconnect the autopilot, no jammed controls or loss of control.
I did some of that work in a past life, part of a cast of thousands working for a big kite factory here in the Pacific Northwest. Trying to get all that work done on a limited budget with just a couple of people and limited test equipment is really tough.
Designing a high quality autopilot servo installation for pitch and roll to interface with the flight control cables acceptable to the FAA DER would be the first step.

[Subsonic]

Designing a high quality autopilot servo installation for pitch and roll to interface with the flight control cables acceptable to the FAA DER would be the first step.

I'm wondering if the servo array design from the Brittain Industries AP could be used as a jumping off point?
IM 2-906-01 Rev. A, Piper PA22, Model B2C(1).pdf

[dgapilot]

While an AP installation may sound easy there is a lot to it. Stability, flutter, control forces, effort to slip the clutches, and the list goes on. Just the flight test requirements, operate through the entire speed range and at all the extremes of the W&B envelope.

Remember with amateur built, there is a reason for those 2” letters that say EXPERIMENTAL. Every flight is an experiment.


Sent from my iPad using Tapatalk

Something else to think about. The You kin Mystery Pacer had a removable Tru Track autopilot installed. Jim designed not only the Tru Track series of auto pilots, but all the Edo/Mitchell/Century autopilots up through the Century IV. He knew Auto Pilots and the certification process well. When he started TruTrack he said he never wanted to have to deal with FAA certification BS again.

It was much easier to certify an analog auto pilot than a digital one. The level of design assurance required for the software on a digital autopilot costs 3-4X for a certified digital autopilot than one to be used in the uncertified world. As soon as you have software running a flight critical system the development costs go way up. At work, we needed to replace an analog air data computer on one of the jets (fleet of less than 50). Someone screwed up with contract requirements and basically said make a box that has the same inputs and gives the same outputs as this analog box. The vendor did that, but when it came time for a flight clearance, no joy! It did not meet the software safety requirements for design assurance for safety critical systems. It cost the government almost $2 million to have the vendor redesign the software with the appropriate level of design assurance.


Sent from my iPad using Tapatalk

[Keith Turner]

I believe that is an analog pneumatic servo. The problem is what would you interface it with for electronics? I was thinking along the lines of replacing steam gages with a Dynon HDX, ADAHRS and S32 servos. All those parts can be procured as “certified” and the STC for PA 22-150 covers all but the autopilot. I have a lot of familiarity with the “experimental” HDX, ADAHRS and autopilot and S32 servos. The RV’s all have push/pull control tubes so a bell crank interface works well. The control laws and the capstan drive is available for other Piper aircraft so the big effort is in adapting the servo interface to the PA-22 tube structure in a way acceptable to the FAA. I don’t know - but I guess - the difference in control laws could be accommodated in the inflight tuning capability inherent in the “experimental” version. Originally I was just kicking ideas around to see what could be done to replace older avionics and displays with modern avionics. The sticker shock came in looking at the cost of “certified” compared with “experimental”. There is a real risk of ending up with a $ 40K airplane with a $60K panel.
In the end it comes down to the mission for the airplane. As I totter towards antiquity an easy to fly grass field capable VFR high wing is probably all I need (and be capable of getting insurance on). An autopilot would be nice but not essential.
I looked at going experimental exhibition but that brings some serious limitations. A secondary consideration for the airplane mission was to make an airplane available to the local flying club that would be at a heavily discounted rate for young ( high school age) students so it has to stay in the certified category.

[Keith Turner]

Something else to think about. The You kin Mystery Pacer had a removable Tru Track autopilot installed. Jim designed not only the Tru Track series of auto pilots, but all the Edo/Mitchell/Century autopilots up through the Century IV. He knew Auto Pilots and the certification process well. When he started TruTrack he said he never wanted to have to deal with FAA certification BS again.

It was much easier to certify an analog auto pilot than a digital one. The level of design assurance required for the software on a digital autopilot costs 3-4X for a certified digital autopilot than one to be used in the uncertified world. As soon as you have software running a flight critical system the development costs go way up. At work, we needed to replace an analog air data computer on one of the jets (fleet of less than 50). Someone screwed up with contract requirements and basically said make a box that has the same inputs and gives the same outputs as this analog box. The vendor did that, but when it came time for a flight clearance, no joy! It did not meet the software safety requirements for design assurance for safety critical systems. It cost the government almost $2 million to have the vendor redesign the software with the appropriate level of design assurance.


Sent from my iPad using Tapatalk

Yes - the effort required to verify and validate the software and to meet the design and documentation requirements in the first place is quite onerous and considerable. DO -178B level 3 software and hardware designed to DO-254 standards certainly adds a lot to the developmental costs especially for small companies that don’t have the infrastructure in place. For those companies that have already got glass, digital flight instruments and digital autopilots certified across a range of general aviation aircraft the work to add new general aviation models to the certification STC shouldn’t be a major redo. I suspect the major difference between the experimental versions and certified versions is in the level of documentation, configuration and revision control. Since going through the experimental market was on the way to the larger, more lucrative, certified market for GA and business jets I would guess that Dynon and Garmin both followed DO-178B and DO-254 standards right out of the gate. I doubt they would have made the connection to the certified world without following that trail.
Of course if we want to talk about what is required for part 121 and military aircraft systems ………………

[moto657]

The Retropilot system that Uavionix is researching could be promising. They have a history of bringing functional products at reasonable prices to GA.

https://youtu.be/5RTgV4VUl9I?si=K0LAYgYmHkcvZzfM

https://uavionix.com/retropilot-interest/

[Keith Turner]

The Retropilot system that Uavionix is researching could be promising. They have a history of bringing functional products at reasonable prices to GA.

https://youtu.be/5RTgV4VUl9I?si=K0LAYgYmHkcvZzfM

https://uavionix.com/retropilot-interest/

Yep been very impressed with Uavionix. They have been fueled by the high end (and military) drone market. Taking a serious look at the VA30 and the tail beacon ADSB out. Not sure I like the idea of a 2lb weight swinging about on the yoke. Looks like the worst aspects of experimental hiding inside certified and getting away with it because it is a “temporary” installation installed like an IPad on the yoke by the pilot. Think I’ll pass on that one. The idea of a couple of bungee cords hung off the right side of the left yoke and the left side of the right yoke going down to a servo Velcro’d onto the floor in front of the flap handle driven by either a HDX or VA30 is something to think about though. What could possibly go wrong?

[Subsonic]

I'm thinking it will be more refined in the final iteration. I like the idea.

[Keith Turner]

Still trying to get my head around all the restrictions and limitations of keeping this 1956 Tripacer in the standard certified category. I was just joking about the bungees hanging off the control yokes but it does beg the question of the extreme limits of sloppy engineering that one would have to go to to stay within the “certified” category without a full blown STC.
I thing my next adventure will be to go talk to a couple of DAR’s and DER’s to get a better feel for what I can expect in the experimental exhibit category and the limitation likely in the program letter. I owned a German factory built sailplane for many years registered in the EE category so I have some experience of the program letter limitations and how the process works. May be different for a 56 Tripacer but may be worth asking the question.