MGL Autopilot Finally having a Win- Update

[JimJab]

Hi All,

Background,

I have a Voyager EFIS with Trio Servos for my autopilot and I have no other type of GPS connected to my EFIS. Now having just got my autopilot working and not disengaging due to a servo clutch issue, I also have this left of track issue.

I flew a short GPS leg with my autopilot and notice it was flying left of the HIS boxes and maybe the ground track. I did not check if the aircraft was actually flying left of the GPS track over the ground.

I have seen over time where a number of people say their MGL/autopilot flies left of track and the HITS boxes, my autopilot seemed to fly my selected track with reasonable directional accuracy, but it also flew the selected GPS track to the left of my HIS boxes on my EFIS screen, if I had to guess there seems to be some type of track off set maybe in the EFIS programming.

I've added this comment, having now spent some time with my plane and going through the auto-pilot system, my assumption that there could be an offset programmed into the EFIS firmware for left of track, I now believe this comment was totally in correct.

I had read some where that some GPS Units allow you to off set your track in the GPS allowing your aircraft to have a slightly different track, left or right to other another aircraft flying between the same two waypoints, not a bad idea.

So thinking about the offset a bit more, if you are going to offset your GSP track, I would think you would off set your track to the right. This would then have aircraft flying the same track towards you, passing you down your left side, being the side most pilots sit on.

So, I am wonder if the HSI box Track are offset to the right in the EFIS software and the EFIS GPS is actually on track but seems to be left of track when comparing them to the HIS boxes on the EFIS screen.

I am yet to check my GPS/autopilot track over the ground to compare the HIS track to see which track is correct.

Rainer, could you shed some light on this left of Track issue.

Merry Christmas to those who celebrate this Festive Season.

JimJab

[MikeBlackburn]

So I have the same problem…. But I’m to the RIGHT (and no, it’s not a down under thing because I’m also S Hemisphere…… :0) )

Been an issue throughout my ownership of the aircraft. Not sure why but it’s consistently 0.6-0.8nm off and always to the right….

E9B17A4D-53B7-4CA2-BB59-679715C7EE26.jpeg (92 KiB) Viewed 5747 times

Adjusting the GPS OBS doesn’t do anything.

[Diamond]

Hello, Same issue, any ideas?

[Tasmag]

I'm convinced it is a vibration issue, particularly with the lite versions. I had the same thing until I added an SP7, flies straight down the hits now

[JimJab]

Hi All,

Well its been while and we are further down the track with our auto-pilot.

I too have had nearly every problem mention here on this forum in getting my auto- pilot to play nicely in my Jab 430/230.

But we are nearly there, as I have finally got my auto-pilot to fly through the HITS, right on the magenta line on both my EFIS's moving map and my iPad with Ozrunways. Currently in our smooth winter air, it can hold my altitude +/- 10/20 feet for a long Nav with all different headings.

Whilst I am thinking more tweaking is needed with EFIS's auto-pilot setting, we are getting there, but its been a bit of a journey.

It all started with my Trio Gold pitch servo clutch was badly slipping all the time. So the autopilot would disengaged with in seconds of being turned on. After getting that issue fixed, I thought you beauty.

Here is some background info about this and the Trio Clutch.

viewtopic.php?p=4120#p4120


So after getting the pitch servo clutch sorted, Sadly it was not the total solution to all my auto-pilot problems.

Now the pitch servo was not disconnecting auto pilot, if I trimmed my aircraft perfectly for level cruise and I assigned the EFIS cruise height bug to the altitude I was currently at, the pitch servo would stay engaged until we hit a thermal, would tried to maintain my assigned Altitude, but if I did not change my assigned altitude bug to the current height, with in a few seconds the auto-pilot would disengaged due to the pitch servo slipping.

So at this point, I removed the pitch servo from my EFIS's settings and found my auto-pilot could fly the plane in a fashion using just the roll servo without trying to disengage. But it was pretty ugly with one wing down and the joy stick not in the middle. At least the auto-pilot was not disengaging.

Like a dog with a bone, I talked to the Jabiru Australia crew and they said with the Dynon EFIS it can control the heading and altitude with-in pretty good tolerances. So I thought why couldn't the MGL unit do this as well?

Looking at the aircraft, I found there was an issue with the trim block not freely sliding on the trim shaft between the 2 adjustable trim springs. I was told by Jabiru only use graphite power on this block and shaft as it can gum up otherwise. Since we got this aircraft we have been using nothing but graphic power on the shaft every annual. But what we did not realize the trim system was not working as freely as it needed, when using it with the auto-pilot.

I've attached a picture of the trim block and shaft gummed up with some half dried out grease. I am guessing the previous owner was using graphite grease on the trim block.

Edit Added, so how do you know if your trim is gummed up, well I have now looked at a couple of different Jabiru's and they've had varying degrees of trim gum. It's a pretty easy to test your Jabiru Trim System to see if your trim block has free movement on the shaft between the springs.

On the ground, hold the Jabiru trim lever in one hand to stop it from moving and with the other hand rock the joy stick forward and aft a bit, you should be able to move your joy stick without the Trim lever needing to move. If you are feel your trim lever is wanting to move back and forth with the joy stick or it takes some pressure to stop the trim level from moving forward and aft, then the trim block is gummed up on the trim shaft and it needs cleaning.

Before we cleaned up the Trim Block and Shaft, we worked out the pitch servo was not only trying to move the elevator to get to the correct altitude, it was also trying to move the trim system lever and over come the trim levels friction lock. This was placing a lot more strain on the pitch servo so it was disengaging the auto-pilot, as the pitch servo was slipping.

TrimShaft&BlockCleaningV2.jpg (49.12 KiB) Viewed 2577 times

So once I had the trim block sliding between the springs freely, we found the auto-pilot was able to maintain an assigned altitude as long as I had trimmed the aircraft nicely for level flight.

In my experience what I can say, correcting my auto-pilot issues is as much about knowing and leaning how the system works and identifying what could cause the problems and then that was causing the auto-pilot to have it control issues.

Have to go for now, but there is still some stuff to work out, but we can see a light at the end of the tunnel in getting our auto-pilot to play nicely with our aircraft.

Cheers
JimJab

[JimJab]

Hi All,

Well with my auto-pilot, sadly I had been totally focused on my servos, MGL EFIS and settings to solve my auto-pilot issues and nothing I did was improving the way my auto-pilot flew my aircraft, one wing slightly down and ball slightly to the side.

But there's two parts to an auto-pilot, the servo control system and the aircraft/airframe that the auto-pilot controls. I was focusing on the idea that the issues were coming from the MGL auto-pilot control system, right!!!

But I was wrong, as we put the spot light on our aircraft we realized it was not flying that well with hands off in smooth air.

So we went back through the rigging process and this took time, patience and someone who knew aircrafts well. The hardest part was waiting for those smooth air days so we could feel what our aircraft was wanting to do.

After we adjusted a few things and with the way our aircraft was now flying, I actually wondered why we needed an auto-pilot in our aircraft as it flew so nicely, hands off. Well maybe not perfect, but so much better than before.

A comment was made, how can you expect the auto-pilot to fly your aircraft nicely, if your aircraft does not want to fly nicely by itself with hands off.

So how much difference did this make to the auto-pilot, well, it made a world of difference and well worth the time and effort in checking the rigging. The auto-pilot is no longer flying with a slight wing down or left of track.

Why was my aircraft rigging out, well that's another story, which must have come about when the aircraft was de-rigged for trailering and my guess when it was put back together, the owner/builder did not spend the time to fine tune the rigging before we purchased the aircraft.

I was also told you don't have to pulled your aircraft apart to wreck it's rigging, any type of adjustment of the flying surfaces can upset the flying balance of your aircraft.

Cheers
JimJab

[JimJab]

Hi All,

After re-rigging and trimming our aircraft, it was time to start looking again at my MGL auto-pilot settings and servos.

What I had noticed after fixing the pitch servo clutch from slipping, quite often when on the ground and turning on the auto-pilot and selecting "Straight and Level", the roll servo would start chattering for no obvious reason and yet on other days the roll servo would not chatter at all.

So on this particular day, I was sitting in our plane on the ground having turned the auto-pilot on by selecting "Straight and Level" and then started reading the MGL manual about the auto-pilot, whilst I was waiting for the roll servo to start chattering.

Finally I stop reading and by chance shifted my weight in the pilot's seat and then suddenly the roll servo started chattering away. From this I finally worked out, what made the roll servo start chattering, it was the AHRS had moved slightly.

I found that a small area of the adhesive clue had let go of the felt lining where the Velcro strips secured the AHRS unit to roof. Really you would not have believe the small amount of movement it had, maybe a 1mm would make the roll servo start chatter away. So we move the AHRS side ways by about 2cm to where the roof liner was more securely glued to the roof.

Being on the ground, we then turned the auto-pilot on and selected "Straight and Level", I pulled at the AHRS Velcro for a bit and then we tried rocking and rolling the wings. Nothing we did could make the roll servo start and continue to chatter.

At this point, I was wondering how much difference this would make to the way the auto-pilot would fly.

But something else I noticed, when testing the auto-pilot with the MGL pre-flight servo test, I noticed the roll servo centre position did not match the Joy Stick's position in level flight.

So we adjusted the roll servo push rod so the servo centre matched the joystick in level flight. This most likely makes no difference, but hey, if it looks right it must be good.

Cheers for now
JimJab

[JimJab]

Hi All,

Well, I have nothing much to report about my auto-pilot.

We did take the plane for small flight this afternoon and I was meaning to get my Voyager auto-pilot settings, so I could post them here. But I got home without them, maybe next time.

The elevator trim on our Jabiru type aircraft operates very differently to other aircrafts, it uses a frication type mechanism to hold the elevator in the required position to maintain level flight. With other aircraft like Cessna's they have a moving trim tab and using the air flow pressure to hold the elevator in the required position to maintain level flight.

Over time a number of people have said different things like, the Jabiru elevator/trim system does not work with auto-pilots or the MGL auto-pilot system is not good or the Jabiru aircraft can't be flown successfully by an auto-pilot, so basically give up trying.

Well, I am please to say, all that is not quite right. Once you get everything set up correctly, the Jabiru can be flown reasonable well with an auto-pilot. We have flown a few Navs now with our MGL/Trio auto-pilot and our Jabiru does indeed fly reasonable well with our MGL Voyager auto-pilot and MGL flight plan.

So with the auto-pilot, it changes to new way-points and headings, it now is flying on top of the moving map's magenta track line, is flying through the centre of the HITS Boxes, able to fly a holding pattern once over the last way-point and the auto-pilot can maintain our selected altitude +- 20 feet in smooth air and with some moderate turbulence +- 50 feet.

The better I trimmed my Jabiru for level flight, the easier it is for the auto-pilot to maintain the required flight level. I am yet to explore how our auto-pilot will go flying a descent profile to our last way point.

My guess this will be a challenge with the Jabiru type of trim system, but hey do we need this for VFR flight.

Cheers
JimJab