AUto Trim Function with SKyview Autopilot

[JohnH]

I have been giving the Dynon Autopilot trim annunciations a lot of thought and have wondered whether Dynon could consider the idea I have.

If the Skyview (or other Dynon EFIS is able to annunciate a triom requirement on the screen, how difficult could it be to provide a pin out for trim up and trim down. This pin could simply provide a signal, eg. ground signal to a relay system connected to the Ray Allen trim servos for automatic trim functions with Ray Allen trim servos or any other trim servo system.

The idea is that I would like to build a relay that is triggered by such a signal and have the system switchable and pulsed with very brief pulses to the Ray Allen servos on the RV7 I am building.

All it requires is 2 signal grounds (or 12V) that provide the signal when the EFIS signals a trim requirement. The builder would need then to provide whatever interface to the trim system. I would not be difficult to build a relay system that could provide a small pulse signal to the servos. It could also be wired with a manual override switch to switch off auto trim functions. It may take a few iterations to get it right, but if I could have the signal provided by the EFIS I am willing to give it a try.

 

[dynonsupport]

The issue is that all systems like this are control systems and delays, speed of trim movement, and other factors all play in. It's pretty easy to end up with a system that just chases itself. Thus, the hard part here is not the hardware, it's the software. Likely software that needs some tweaking for each airplane it's installed in, just like the general autopilot does.

The SkyView system already outputs trim indications on the serial port, so it's possible for someone to read this, and write an algorithm to control the trim. We have considered doing the same thing. But it's not as easy as just moving the trim up when the screen says "Trim up" and vice versa, so it isn't just a day's work, and it hasn't been more important than other projects so far.

So, someday...

 

[JohnH]

Thanks for the reply. I realised from the beginning that this will be more involved than a simple input to the trim servo. It will require testing in an installation to get it working right with lots of tweaking.

I am about a year away from my final installation and would really like to have some method to get the autopilot to adjust pitch trim when engaged.

Reading the serial data requires a unit that can read and interpret a string of data. I also understand that using the annunciation lights only may result in too much or too many trim changes, but this can be controlled by adjustable hardware with control of the trim servo with very small bursts of trim along with control of the servo speed and the frequency of the bursts.

I understand that due to nature of Dynon's customers, there is an enormous range of requirements, but if you make something available, to us customers to start with, then we can perhaps solve our own individual issues.

Maybe someday????

 

[NASA515]

I'm not a fan of auto-trim.  All the newer transports have the feature - and a number of accidents have ensued over the years.  I've discussed it at length with my engineering and test pilot pals at Boeing.

Here's a webpage I created that discusses it in some detail:

http://www.rbogash.com/Safety/autopilots.html

Bob Bogash
N737G

 

[Skee]

And now garmin has done it to their G3X system Autopilot, auto trim.

 

[airguy]

It's certainly do-able, but not easy. You'll want a deadband of servo torque centered on null where no trim inputs are made (just let the servo handle it), and then your active bands on either side would need to move the servo (maybe even 2 or 3 different speeds, using PWM, depending on how far outside the deadband you are) in order to get the servo torque values back inside the deadband. Adjusting the width of the deadband and the PWM trim speed would let each user be able to set it up pretty much how they want it.

 

[ljbuller]

I programmed a pic microprocessor, had a circuit board made, all tested and ready to go. However, I came to the conclusion that it is more of a solution waiting for a problem. I basically need to make one trim adjustment leveling off and maybe one more as fuel burns off on my RV. I had thoughts of making it available commercially, but there HAVE been accidents where people have input control inputs contrary to the autopilot and then the trim running away. I decided not to mess with it, even on my own plane.

 

[airguy]

I programmed a pic microprocessor, had a circuit board made, all tested and ready to go.  However, I came to the conclusion that it is more of a solution waiting for a problem.  I basically need to make one trim adjustment leveling off and maybe one more as fuel burns off on my RV.  I had thoughts of making it available commercially, but there HAVE been accidents where people have input control inputs contrary to the autopilot and then the trim running away.  I decided not to mess with it, even on my own plane.

Can't say I blame you. It's not impossible to do, you can bet there are many individuals that have followed that train of thought (and action) at least as far as you have. I'm all for that in the spirit of experimentation, but to sell it commercially brings a layer of liability upon the manufacturer that is a risky proposition for the little guy. You could post plans for it, and force them to build it themselves - that would absolve you of any realistic liability that would hold up in court.

No, I don't want it in my plane either - just sayin'!

 

[purlee]

Hi Guys,

I am new here, but with a similar time in aviation as Bob Bogash, I am with him all the way. It is one of those things which sounds like a good idea, but really isn't.

Anything which can create a badly out of trim airplane, without the pilot being aware of it, has the potential to be truly dangerous and using features such as VNAV or ALT HOLD whilst handing control of trim to the autopilot can do that.

On a light aircraft the need for major trim changes is minimal and so long as there is a clear indicatin of a trim conflict between the trim control and the AP manual trimming is always going to be better and safer.

Bad surprises when disconnecting an autopilot in low visibility and close to the ground produces more "pucker factor" than any of us really need