airguy
Well-Known Member
Dynon - any plans on interfacing with electric pitch trim devices to keep them in trim automatically instead of having to manually adjust?
Autotrim?
- Thread starter airguy
- Start date
- Joined
- Mar 23, 2005
- Messages
- 13,226
It's definitely interesting, and something we've though about, but there are some real challenges that make it not as simple as people may think:
So first of all, trim runaways commanded by autopilots kill, that much we all know from reading NTSB/news reports. Our autopilot is designed to mitigate that possibility in a couple of ways. First, the autopilot only pushes on the control surface actuators directly, just like your fingertips do on the stick. To limit the autopilot's ability to get you into trouble in a hurry, the servo has limited control authority, and so once you get somewhat out of trim, the servo starts to "slip" when it can't do what it wants to against the forces that are required to overcome an out of trim condition. Additionally, there's the trim annunciation, which tells you when you ARE out of trim but DO still have enough autopilot control authority such that the servo isn't slipping.
In today's methodology, if you're being asked for some trim by the AP, you'll tap the trim button or spin the wheel some. You shouldn't have to do this often, by the way. Every now and then in cruise, perhaps at the start and and of climbs/descents. You, the pilot, are part of the AP control system, and this is important: It's very unlikely that, if, somehow, the autopilot trim detection system failed, that you'd just keep applying trim mindlessly. You KNOW (because you engaged the AP relatively in trim) that it should be small adjustments.
An system that auto-drove the trim, however, doesn't have the same ability to detect when it's being asked for excessive trim when its sensors fail. Yes, the technically oriented among you can think of ways to characterize to mitigate this by characterizing the trim required, or limiting the amount of trim that can ever be put in by the autopilot to some subset of the full trim range, but, in the end, if you miss a failure case, and the system runs away with the trim though some catastrophic failure, it can be a bad day. A very out of trim airplane can create forces that are between hard and impossible to physically overcome without first reversing the out of trim condition.
Anyway, our view is - once you automate the trim control, there's a whole other level of redundancy and failsafeing that needs to be considered, and it gets complicated quickly.
So first of all, trim runaways commanded by autopilots kill, that much we all know from reading NTSB/news reports. Our autopilot is designed to mitigate that possibility in a couple of ways. First, the autopilot only pushes on the control surface actuators directly, just like your fingertips do on the stick. To limit the autopilot's ability to get you into trouble in a hurry, the servo has limited control authority, and so once you get somewhat out of trim, the servo starts to "slip" when it can't do what it wants to against the forces that are required to overcome an out of trim condition. Additionally, there's the trim annunciation, which tells you when you ARE out of trim but DO still have enough autopilot control authority such that the servo isn't slipping.
In today's methodology, if you're being asked for some trim by the AP, you'll tap the trim button or spin the wheel some. You shouldn't have to do this often, by the way. Every now and then in cruise, perhaps at the start and and of climbs/descents. You, the pilot, are part of the AP control system, and this is important: It's very unlikely that, if, somehow, the autopilot trim detection system failed, that you'd just keep applying trim mindlessly. You KNOW (because you engaged the AP relatively in trim) that it should be small adjustments.
An system that auto-drove the trim, however, doesn't have the same ability to detect when it's being asked for excessive trim when its sensors fail. Yes, the technically oriented among you can think of ways to characterize to mitigate this by characterizing the trim required, or limiting the amount of trim that can ever be put in by the autopilot to some subset of the full trim range, but, in the end, if you miss a failure case, and the system runs away with the trim though some catastrophic failure, it can be a bad day. A very out of trim airplane can create forces that are between hard and impossible to physically overcome without first reversing the out of trim condition.
Anyway, our view is - once you automate the trim control, there's a whole other level of redundancy and failsafeing that needs to be considered, and it gets complicated quickly.
airguy
Well-Known Member
Those are definitely valid points - but I would think it would be relatively simple (and granted, I'm not designing an AP for experimental aviation) to interface with one of the electric trim providers like Ray Allen that include a potentiometer in the trim device for position feedback. With that done then you can have a range of absolute position that is "open season" for the autopilot to use with full authority (say 35% to 65%) with another 5% on either side of that which the autopilot can use while giving a trim caution warning. If the AP comes up against the hard limit, give a hard alert for 2 seconds and disconnect. That leaves the trim inside a range that can be handled manually even if it's out-of-trim, and ideally would disconnect at the hard edge with a near-in-trim condition that the pilot could easily handle.
The calibration procedure for this could even require the pilot to set min/max trim positions in flight to determine their comfort level with the final disconnect at the hard edges.
Just thinkin' out loud, it would be very nice to have...
The calibration procedure for this could even require the pilot to set min/max trim positions in flight to determine their comfort level with the final disconnect at the hard edges.
Just thinkin' out loud, it would be very nice to have...
Yaw trim?
What about Yaw trim? I read on another board about the success someone had with TruTrak's yaw trim. Basically, he said the auto yaw did much better than he was capable of doing with his own feet.
I didn't even know auto yaw existed, let alone could work that well. To me, that would be even more of an amazing feature.
The safety issues of runaway yaw would be less slightly less of an issue i would imagine.
Jae
What about Yaw trim? I read on another board about the success someone had with TruTrak's yaw trim. Basically, he said the auto yaw did much better than he was capable of doing with his own feet.
I didn't even know auto yaw existed, let alone could work that well. To me, that would be even more of an amazing feature.
The safety issues of runaway yaw would be less slightly less of an issue i would imagine.
Jae
I am finishing an F1 Rocket with Dynon Skyveiw 10" with all of the extras except auto trim. I have flown Lear Jets, Citations, Falcons, Challengers, Boeings, and several turboprops and they all had auto trim. It is a very nice feature the aircraft is always in perfect trim when the auto pilot is disconnected. There has to be safe guards built into the auto trim system but do not see that as being a complicated thing. In my case I have installed a Vertical Power VX-P and Skyveiw is capable of trimming it already. Write a little software some safeguards say 1/2 second burst no more than four per minute or better yet let the operator set thoughts options, and you could charge me a few for the soft ware but Garmin includes there auto trim. It is a much safer with auto trim. 13,000 + hours never had a problem except in the sim that thing always brakes.
Jerry
Jerry
vlittle
Active Member
- Joined
- May 7, 2006
- Messages
- 544
This is a tad premature, but this fall there will an open-source trim controller design released from MakerPlane that can be programmed to talk to a SkyView and sniff the data streams.
It's feasible, with some firmware development, to make this into an auto-trim device.
It would be a lot easier if the SkyView did the heavy lifting, however, and issue command messages on a serial port. These could be runtime:direction or just emulate button pushes or whatever. I've developed a protocol that would work, but it's easy to adapt to whatever Dynon would like.
So if Dynon does implement autotrim through a VP system, my trim controller could also be supported with a little bit of effort in providing serial message handling... making the retrofit applications much easier to address.
Vern
It's feasible, with some firmware development, to make this into an auto-trim device.
It would be a lot easier if the SkyView did the heavy lifting, however, and issue command messages on a serial port. These could be runtime:direction or just emulate button pushes or whatever. I've developed a protocol that would work, but it's easy to adapt to whatever Dynon would like.
So if Dynon does implement autotrim through a VP system, my trim controller could also be supported with a little bit of effort in providing serial message handling... making the retrofit applications much easier to address.
Vern
TRCsmith
Member
I'm waiting for Dynon to come out auto everything. I'm so tired of take-offs, flying, and landing the plane. Plus when leveling off sometimes the auto pilot gives me a cool yellow arrow (up or dn) to trim. Cool, I just tap the trim sw and done. O, butt his should not be necessary. Dynon do you hear us.. You must come out with an auto everything! And how about the auto radio frequencies. The plane must know when it's near an airport and auto select the Atis and then auto select the tower or unicom, about 10 seconds after lnd select gnd. By the way do the same for nav.
Auto everything for everybody, cuz we hate hand flying the plane! ;D
Auto everything for everybody, cuz we hate hand flying the plane! ;D