Auto Pilot

[rhtrudder]

Having trouble with my trio avionics AP not holding a straight course , wants to hunt , the linkages are all tight, wondering if it could be the 296 garmin not good enough. , tried all the settings

 

 

[Robbo]

Does GPS go off track even if AP disconnected?

 

 

[jetjr]

Had a long running issue with Dynon AP and a 296, eventually fixed with software upgrade from Dynon

 

Maybe the output is not clean or consistant enough

 

Usually theres sensitivity adjustments that can be made in AP

 

 

[Guest Andys@coffs]

I'm using a Garmin 295 with mine and its works fine......Hunting is an interesting term..... If you are expecting the wing leveller to roll to a course and fly perfectly wings level without any adjustments at all then that's not really likely...mine holds a course so that reviewing track on a GPS using breadcrumbs shows that the course was exact, but the wings easily are +/- 5 degrees of level as the system hunts around the required heading.....

 

If you look at the manual (I've assumed an EZ model) http://www.trioavionics.com/Manual rev 2.3.pdf specifically section 7.9.1 then the settings in this section relate to how diligently the system chases the required heading, the larger the numbers the greater the system gain and the more likely the system is to oscillate around a heading...try reducing the gain numbers....

 

I remember being in the RAAF on a C130H model flying to Malaysia and noting that the heading hold AP had the aircraft constantly rocking Left and Right with a cycle time of about 15 seconds....it was noticeable visually, but not at all by closing your eyes and using our inbuilt balance detection etc....

 

The following from 7.9.1 deals with the issue:- (Bold and underline are mine)

 

For instance, assume that you are flying in the

 

CRS

 

mode (in smooth air) and you find

 

that the aircraft “wanders” a few degrees to the left and right. You may wish to enter the

 

calibration screens to adjust the

 

FINE CRS QLT

 

gain to lessen the deviations.

 

Additional

 

gain may improve the performance, but be aware that too much gain may also cause

 

oscillations. You should adjust it to find the “sweet spot” for your aircraft where it holds

 

the course best.

 

Use the following procedure.

 

You can discount the GPS by simply turning it off. the AP will then use the internal Gyro to hold a heading...see what happens oscillation wise when you do that. (knowing that the internal Gyro isn't corrected and over time the heading needs to be corrected to fly a constant compass heading.....

 

 

[rhtrudder]

I did not know that turning off the GPS the AP would still work , that will be first on the list , I can get it sometimes to barely move other times have to turn the the thing off , had a tru track before this one, done the same thing , so maybe it is the gps

 

 

[Guest Andys@coffs]

Well "still work" is qualified......with GPS providing valid NMEA Data it will follow a route, with it off all it does is basically hold a single heading that degrades over time and has to be fine tuned every few minutes or so with the L-R switch.......but the if the GPS data is providing valid NMEA data then your route info from the GPS will be seen on the AP instrument. If you are getting a GPS waypoint identifier and stable ETE and range etc to the waypoint then I suspect the NMEA data is working fine. If however the ETE is jumping all round the place then the GPS NMEA data is probably shagged but that shagged DATA is being wrapped in valid NMEA protocol data (pretty unlikely I would guess)

 

For clarity NMEA Data is serial Data if your GPS is sending it and the AP is receiving it no issues, then a steady time to run and distance countdown as well as a Track that doesn't change dramatically (last true only if you have a reasonable distance still to run...if the remaining distance is under a nm then don't be surprised to see track start to vary significantly as you count down to 0) is all you should see. If the NMEA data is out of spec then the AP will simply be in No GPS mode. It would be a rare fault that makes the NMEA Data valid, but wrong, there are protocol checks at the serial level to ensure that wrong data is known to be wrong.....

 

In the manual para 3.2.3 deals with the GPS off

 

rhtrudder where is home? Trio have an Australian Agent slightly south of Adelaide city at Hallet Cove from memory...Im at Coffs/Grafton and can probably help if your ready to burn something....I was a RAAF Radio Tech on P3's and F111's.....

 

All GPS's are not equal when it comes to NMEA, the better GPS's have a faster NMEA refresh rate than others but to be honest when your still a long way from a waypoint the slower updates shouldn't really reflect that much in degrade performance because rel;atively not much is changing from one update to the next.....closer to the waypoint where things are changing relatively quickly then it can matter. I did a quick google search for garmin 296 NMEA problems and there wasn't a groundswell of complaints....so not sure what to make of that.... Is the AP the only user of GPS NMEA Data in your aircraft? I ask in that a lot of glass panels also want it so I was wondering if there is an issue with the GPS system driving more than one reader.....

 

Can you publish your GPS NMEA Setup settings and also confirm if the route info appears stable on the AP display....

 

 

[DrZoos]

All GPS's are not equal when it comes to NMEA, the better GPS's have a faster NMEA refresh rate than others but to be honest when your still a long way from a waypoint the slower updates shouldn't really reflect that much in degrade performance because relatively not much is changing from one update to the next.....closer to the waypoint where things are changing relatively quickly then it can matter.

So Andy what's the technique here for smoothing out GPS induced oscillation...have waypoints as far apart as possible???

 

 

[Guest Andys@coffs]

No....the point I was making was that if GPS A provides a complete NMEA refresh of data only once per second as compared to GPS B that provides it 4 times a second then when you are a long way from your waypoint then the relative change second to second is relatively minor until you get up close where the 1 second per change may make the changes appear jerky....for GPS B with its 4 times a second update the changes appear smoother...... but for both when you are at least a Nm away there wont be any appreciable difference between them.

 

Waypoints are places, that generally we want to fly over or to, that being true they are all going to be close at some stage and when that occurs the changes will be noticeable...That said, if you fly for 2 hours over 4 way points then in 120 minutes we might see the jerkiness for 2 mins max of that 120min flight......

 

<WARNING Techspeak ahead!!!>

 

For a description of NMEA as it relates to Garmin this website is useful, albeit quite old http://www.gpsinformation.org/dale/interface.htm In this link about 1/2 way down there is a discussion on the NMEA protocol. In it, to summarise, a single NMEA update is one sentence long transmitted over a serial interface at 4800baud (although I'm pretty sure I use 9600 on mine) the sentence is transmitted in RS232 so it has its own built in methods of ensuring that data isn't corrupted in the transmission, then at the NMEA Protocol level there is a checksum so that ensures that the NMEA sentence is also not corrupted (redundancy). Each sentence is updated X times per second as defined by the GPS OEM. the refresh rate isn't something that is probably easy to find in the GPS Blurb....but a email/call to the OEM will usually result in the required info......

 

I note that for the 296 in software update 2.6 to 3.0 the update included "Added output rate configuration to the Advanced NMEA setup page." ( https://www8.garmin.com/support/download_details.jsp?id=889 ) I would be interested to see what the setting for this is on the troublesome install......

 

I seem to recall that 1 a second refresh was the norm back in the day of the first mapping GPS's but I also note that the Garmin equivalent of the Bad Elf used for ipads etc has a NMEA sentence refresh rate of 10 per second......

 

....anyway all this discussion of NMEA is relevant in the whole picture but at this time is no more surely "the problem" than any other part of the system. Further fault finding required.

 

Andy

 

 

[dutchroll]

I remember being in the RAAF on a C130H model flying to Malaysia and noting that the heading hold AP had the aircraft constantly rocking Left and Right with a cycle time of about 15 seconds....it was noticeable visually, but not at all by closing your eyes and using our inbuilt balance detection etc....

You sure you weren't witnessing my namesake? Dutch roll?

I flew C130Hs and they were very prone to it. Actually I flew all the Herc models the RAAF ever had except the "A" which was well before my time and they all did it. The Boeing 767 and the Airbus 330 do it too, but it's pretty subtle and you can barely notice it.

 

Usually the autopilot would be in "Nav" mode coupled to the INS and flying the ground track but it didn't matter what mode you had it in, it would still dutch roll. If you looked out at a wingtip it would prescribe a small circle as the plane was rolling and yawning.

 

This is an aerodynamic roll-yaw coupling and the autopilot just spends the whole time trying to correct it (almost like a dog chasing its tail). It could often be dampened out to a good degree by adjusting the rudder trim.

 

 

[Guest Andys@coffs]

Perhaps a point of clarity...turning off the GPS while the AP Servo is engaged and following a GOS Route will not likely result in the servo remaining engaged, I would expect the AP to disengage the servo in what it would clearly determine is a fault condition.... However if you then press the servo engage button on the head unit, I expect the servo to re-engage, the head unit to display "NO GPS......." and the rest as per manual.....

 

I was concerned you would take me literally and expect the servo to remain engaged.....

 

For the purpose of determining if the behaviour you are seeing in the AP is or isn't GPS driven that test will do fine.

 

To determine if the refresh rate is an issue create a route that has you flying a small square with each leg about 1nm long. If refresh is an issue then see how abrupt the changes are and how long it takes for each leg to settle down. If refresh is an issue you may not achieve equilibrium in the 1nm you have between waypoints..... Another is what happens when you reach the final waypoint.... The system is supposed to circle atop the last waypoint, does it do so?

 

I haven't done these tests in mine because under normal circumstances most enroute waypoints for long distance work result in only minor heading changes ( not at all like 90 degree changes) and when you do get to the stage that 90 degree changes are required your in the circuit and will have reverted to hands on flying.

 

Make sure you have plenty of height....and take a second pilot who can fly the plane while you are head down in the guts of your AP/GPS.....now isn't the time to rediscover why " 1st fly the plane" is an unarguable reality.....

 

Andy

 

 

[Guest Andys@coffs]

You sure you weren't witnessing my namesake? Dutch roll?I flew C130Hs and they were very prone to it. Actually I flew all the Herc models the RAAF ever had except the "A" which was well before my time and they all did it. The Boeing 767 and the Airbus 330 do it too, but it's pretty subtle and you can barely notice it.

 

Usually the autopilot would be in "Nav" mode coupled to the INS and flying the ground track but it didn't matter what mode you had it in, it would still dutch roll. If you looked out at a wingtip it would prescribe a small circle as the plane was rolling and yawning.

 

This is an aerodynamic roll-yaw coupling and the autopilot just spends the whole time trying to correct it (almost like a dog chasing its tail). It could often be dampened out to a good degree by adjusting the rudder trim.

Dutchy

 

It was likely as you say, the reality is that the self loading cargo doesn't exactly have celubrious window space for contemplating the aerodynamic behaviours

 

In this case however we are talking only single axis AP, or perhaps like mine he also has the alt hold module so pitch and roll only

 

Rhtrudder can you confirm we are talking only single channel and I if not, that the original behaviour occurs on only roll AP servo selected on. Are your feet still on the rudder pedals and the aircraft is still balanced ball centre?

 

Andy

 

 

[jetjr]

Dynons issue with "NAVSRC lost" was isolated to Garmin 296 (or maybe all x9x range) so there must be a specific issue there

 

data rates are pretty good and adjustable, to 9600. This speed didnt alter the issue

 

As the problem is oscillation or hunting id suggest reaction sensitivity is set too high - maybe nothing to do with data source.

 

 

[rhtrudder]

I can slow it down and it rocks the wings slower ,to much gain and we are out of control, tried every setting, every flight seems different , sometimes it disconnects and slowly banks to the left have to turn it off and start again. I need to take a experienced pilot for a spin for another opinion, unsure what to do next. soon as I get a chance I will switch off the 296 and hope that makes a difference.

 

Thanks for all the help

 

 

[Guest Andys@coffs]

when the headunit disconnects the servo, what message does it display, Headunit disconnecting the servo is an error condition and I would expect some feedback on why. Is it possible that you have the servo settings incorrect so that the servo instead of driving to correct a condition actually makes it worse by going the wrong way? That would likely cause a disconnect.....

 

 

[rhtrudder]

Sorry andy, can't remember will take note next time