Autopilot .... cat's out of the bag now

[Ken_Kopp]

Will HS-34 be able to change heading and course functions for the autopilot?

If not, and an AP-74/76 is not installed I'm assuming we will have to access a menu on the EFIS to make these changes?

If this is true, and the EFIS D10A is the only attitude reference, this may be an interesting issue if flying IFR - virtually requiring the AP-74/76 to be installed for IFR flight (not legally but practically).

 

[6rv]

"FD" is flight director. It lets you toggle it on and off easily.

Did I hear someone say "flight director"? As someone with a D-10A, coupled to an SL-30, and an existing dual axis autopilot, is there any plan on offering the FD functionality without the Dynon autopilot and additional hardware?

 

[dynonsupport]

We don't have such plans at the moment.

Also, since the AP76's design is still being finalized, note that the FD and turbulence buttons on the AP76 are not firm yet.

 

[PilotKris]

Unfortunately, the CDI button on the 430 only changes the output of the analog signals, and doesn't put anything on the ARINC-429 lines that we've been able to figure out, so we can't follow it as you press CDI.

That's incorrect. The CDI button DOES send something over the ARINC 429.

I have a GNS530 hooked up to my Sandel SN3308 EHIS (using only the 429 interface) and I can change the CDI source with with either the "Source" button on the Sandel or the CDI button on the 530.

PilotKris

 

[dynonsupport]

Yes, and you also have your 530 set to "Sandel" format, not ARINC-429 or GAMA-429 format.

When in that format, the 430/530 does send something, but it's all proprietary and Garmin doesn't document it. Since we use the ARINC standard messages, we don't get to see it. Sandel and Garmin work together to make up formats that work best for them.

 

[PilotKris]

Sneaky Bastards...

Any chance of reading the "Sandel" data stream and reverse-engineering it?

 

[Louiscfi]

Very impressive price point.....very glad I was sooooo lazy about ordering that tt. Just 1 question.

ok, I'll ask again.

Is this going to be a system that will annunciate that the autopilot needs a change in trim and then the pilot would need to trim for the autopilot? Or, will the autopilot be able to trim the plane automatically when it is engaged?

If this system does need to be trimmed by the pilot, in your testing so far, how much trimming does it require over the normal flight range, like from cruise through climb and back to cruise again?

Thanks for your time

 

[dynonsupport]

Trim sensing and control won't be a launch feature, but we've designed the system to be able to grow to possibly do both over time.

The initial autopilot features - HDG, TRK, horizontal NAV, and ALT hold/change (either with or without an AP74) haven't required either to be able to fly the core airplanes we've been testing with (RVs).

As we add the more advanced vertical modes such as vertical speed and approach modes with the AP76 later on, we'll be doing some deeper R&D in this area.

And obviously, annunciating out-of-trim conditions is a nice feature for the customer too.

 

[Bill_Denton]

JFTR: Gamin does have VLOC ANNUNCIATE, GPS ANNUNCIATE, and CDI SOURCE SELECT on discrete pins on the 430/530 units. If you have any spare pins on the HS34 you might be able to pick up the info from these pins on the Garmin units.

Yes, and you also have your 530 set to "Sandel" format, not ARINC-429 or GAMA-429 format.

When in that format, the 430/530 does send something, but it's all proprietary and Garmin doesn't document it. Since we use the ARINC standard messages, we don't get to see it. Sandel and Garmin work together to make up formats that work best for them.

 

[Louiscfi]

thanks for the reply...your a/p sounds better every day..

 

[Louiscfi]

any idea of if a warranty will be included on the servos? If so, any idea of what it will cover and for how long? Thanks again.

 

[dynonsupport]

Hmm. There hasn't been any discussion of it being any different from all our other products, so I doubt we'll change it: 3 years from date of shipment, all repairs are covered during that time.

 

[jaba-who]

One of APs I looked at has a riciprocal heading capability. Sort of "if suddenly in the gloop hit a button - AP takes you on a rate 1 turn around and brings you out of the IMC".

Seemed like a good idea to me though never been (or plan to consciously get ) in a spot where I'd need it.

Would it be hard to add that form of feature and if not would you?

 

[dynonsupport]

This is already in our product. There's a 180 turn button in the EFIS menu, and if you have the AP74 or AP76, holding the engage button for two seconds does a 180.

So yeah, we'll add it

 

[dodsond]

Dynon: And obviously, annunciating out-of-trim conditions is a nice feature for the customer too.

This is a safety requirement for any pitch controlling autopilot.  If the autopilot is disconnected while out of trim, unsafe aircraft motion could easily occur.  Automatic control is nice to have, but annunciation is a must have in pitch.

I will probably retrofit from my TT DFC-200 VSG as soon as you have pitch trim feedback.

 

[dynonsupport]

There's thousands (maybe even tens of thousands) of pitch controlling autopilots out there flying around that control pitch yet don't annunciate out of trim conditions. It's nice, but saying that it is a requirement is not something we agree with, based on real world evidence.

One advantage of stepper motor servos like ours is that they have torque limits. If the trim gets too far out, it ends up overriding the motors. Set your torque limit at a sane number, and if the aircraft gets enough out of trim, it will be obvious that it is out, yet it won't be so far out that the aircraft is not flyable. The servo is much weaker than a human, just like it should be.

 

[dodsond]

For the roll axis I would agree. I doubt there are tens of thousands of 2 axis autopilots out there outside of the airlines.

Every aircraft with a pitch axis autopilot I have ever flown has either an up/down trim light or electric trim which the autopilot automatically controls. Even the G1000 C-172 (KAP140 autopilot) has electric trim that the autopilot controls automatically. If the electric trim fails, then the autopilot automatically disconnects and won't engage.

If ALT HOLD was the only mode, and the pitch axis only used when in VMC, then maybe you could get away with it by relying on torque limits. If the aircraft were trimmed for a max rate climb, then the autopilot engaged in VS mode with an ALT capture, there would be a big trim error once in level cruise. A disconnect would give an abrupt pitch up. Under IMC this is way unsafe. In VMC I suppose it wouldn't be too bad as long as it didn't cause an over g.

If torque limits are set low enough to be safe IMO, then the pilot would need to trim accurately all the time. This workload would negate the advantage of an autopilot, again, IMO.

Doug Dodson
Glasair II-S FT
FLight Test Engineer, CFI A-S&ME,I,G

 

[khorton]

If the aircraft were trimmed for a max rate climb, then the autopilot engaged in VS mode with an ALT capture, there would be a big trim error once in level cruise.  A disconnect would give an abrupt pitch up.  Under IMC this is way unsafe.  In VMC I suppose it wouldn't be too bad as long as it didn't cause an over g.

This is an important aspect that should be assessed by the builder, at both forward and aft CG.  If the testing shows potential issues, they may be able to be mitigated by operational procedures that required an autopilot disconnect and retrim if the speed changed by more than XX kt from the speed at autopilot engagement.

The installation instructions should provide good guidance on what flight testing should be done to ensure a safe autopilot installation with acceptable performance.

 

[Brantel]

I have flown in at least 5 different GA aircraft with two axis certified or experimental AP and none of them had pitch trim annunciation or auto pitch trim.

Also, why would someone trim for max climb and then turn on the AP? Why start the AP with an out of trim condition to start with?

Anyway, Dynon obviously will have the abililiy to have torque feedback to the EFIS and this will give them the ability to write the code required for annunciating the need for trim. As far as the automatic trim, I don't want to pay for that as I can do that myself.

For the roll axis I would agree.  I doubt there are tens of thousands of 2 axis autopilots out there outside of the airlines.

Every aircraft with a pitch axis autopilot I have ever flown has either an up/down trim light or electric trim which the autopilot automatically controls.  Even the G1000 C-172 (KAP140 autopilot) has electric trim that the autopilot controls automatically.  If the electric trim fails, then the autopilot automatically disconnects and won't engage.

If ALT HOLD was the only mode, and the pitch axis only used when in VMC, then maybe you could get away with it by relying on torque limits.  If the aircraft were trimmed for a max rate climb, then the autopilot engaged in VS mode with an ALT capture, there would be a big trim error once in level cruise.  A disconnect would give an abrupt pitch up.  Under IMC this is way unsafe.  In VMC I suppose it wouldn't be too bad as long as it didn't cause an over g.

If torque limits are set low enough to be safe IMO, then the pilot would need to trim accurately all the time.  This workload would negate the advantage of an autopilot, again, IMO.

Doug Dodson
Glasair II-S FT
FLight Test Engineer, CFI A-S&ME,I,G

 

[dodsond]

I have flown in at least 5 different GA aircraft with two axis certified or experimental AP and none of them had pitch trim annunciation or auto pitch trim.

Also, why would someone trim for max climb and then turn on the AP?  Why start the AP with an out of trim condition to start with?

Did those 5 aiplanes have VS or any vertical mode besides ALT Hold?

On a missed approach, I would establish a climb, maybe as slow as Vy, trim, then engage the autopilot in vertical speed hold or if I have it, airspeed hold and arm a level off altitude. One that altitude is reached, the autopilot automatically engages altitude hold. If I am not holding, then I would allow the aircraft to accelerate to cruise speed. Without the auto trim, the pilot needs to manually trim. How am I to know when I have the aircraft trimmed without the feed back? What if I fail to trim and disengage the autopilot for some reason?

This is normal operation as I have learned it and as I teach it. Of course, the context is normally IFR. Mishandling an autopilot in VMC is not as big a deal as it would be in IMC.

I can agree that the trim annuciation is 'nice to have" if only altitude hold is available. I would stabilize on speed, trim, and only then engage the autopilot. Altitude changes must be made manually. It would appear that the Dynon basic design would allow autoflight altitude changes with a preset vertical speed. I suppose if the trim remained set for level flight and the preset vertical speed was fairly low, then that too would not be too bad in VMC. Still, throttle changes would cause the trim state to change and out-of-trim disconnects would be an issue.

-Doug