G3X autopilot roll servo issue...

I've been working on the ground tests of the autopilot system prior to enabling it in flight for the first time. Today, I discovered that the roll servo only wants to roll left.

In other words, when the AP is turned on (with heading bug centered) in Hdg/Pit mode, it wants to roll left immediately and will not respond to turning the heading bug right. It's been suggested to me that I swap my pitch servo over to see whether the problem follows it or not. A reasonable troubleshooting step, though obviously a bit of a PITA.

My thought is that it's unlikely to be a servo failure, as they self-test and should annunciate a problem. It seems more like a wiring issue, possibly with a short or perhaps a pin not fully seated.

Has anyone run into this? Any troubleshooting suggestions that haven't occurred to me?

Thanks!

GSA 28 Roll Direction

kbalch said:

I've been working on the ground tests of the autopilot system prior to enabling it in flight for the first time. Today, I discovered that the roll servo only wants to roll left.

In other words, when the AP is turned on (with heading bug centered) in Hdg/Pit mode, it wants to roll left immediately and will not respond to turning the heading bug right. It's been suggested to me that I swap my pitch servo over to see whether the problem follows it or not. A reasonable troubleshooting step, though obviously a bit of a PITA.

My thought is that it's unlikely to be a servo failure, as they self-test and should annunciate a problem. It seems more like a wiring issue, possibly with a short or perhaps a pin not fully seated.

Has anyone run into this? Any troubleshooting suggestions that haven't occurred to me?

Thanks!

Click to expand...

Ken,

The purpose of this section of the on-ground checkout procedures is to ensure the servo is deflecting the control surfaces in the correct direction. This is actually in addition to another step in the process to ensure proper control direction, we want to make sure the servos are moving the in the right direction before you ever get engage the system in the air! If you hold the stick center and move the heading bug to the right in HDG mode, do you feel servo torque on the stick in the correct direction?

Thanks,

Justin

g3xpert said:

The purpose of this section of the on-ground checkout procedures is to ensure the servo is deflecting the control surfaces in the correct direction. This is actually in addition to another step in the process to ensure proper control direction, we want to make sure the servos are moving the in the right direction before you ever get engage the system in the air! If you hold the stick center and move the heading bug to the right in HDG mode, do you feel servo torque on the stick in the correct direction?

Click to expand...

I?ll redo this test tomorrow to confirm, but my recollection from today?s efforts is that nothing I tried would produce a roll or servo torque to the right.

In answer to yesterday's question, there's no servo torque to the right when performing the test mentioned.

Today's update: I swapped the pitch and roll servos and found exactly the same behavior as before, so it's extremely unlikely (impossible?) to be a bad servo. I also confirmed that the roll servo harness connectors are wired according to the master drawing.

It's beginning to feel more like a software issue, but I couldn't say of what sort, since all of the configurable settings are correct.

GSA 28 Roll Direction

kbalch said:

In answer to yesterday's question, there's no servo torque to the right when performing the test mentioned.

Today's update: I swapped the pitch and roll servos and found exactly the same behavior as before, so it's extremely unlikely (impossible?) to be a bad servo. I also confirmed that the roll servo harness connectors are wired according to the master drawing.

It's beginning to feel more like a software issue, but I couldn't say of what sort, since all of the configurable settings are correct.

Click to expand...

Hi Ken,

What is the roll servo torque setting you have programmed into the G3X Touch?

Thanks,

Justin

g3xpert said:

Hi Ken,

What is the roll servo torque setting you have programmed into the G3X Touch?

Thanks,

Justin

Click to expand...

Hi Justin,

It's 50% per the Van's instructions.

An update to close the loop:

After a week of multiple tests, successively eliminating hardware, wiring, configuration, and, eventually, software, it looks as though I?m finally good to go. After a couple of rounds with Garmin?s G3X tech support including videos and data logs, this is what we discovered:

The Garmin autopilot has a particular behavior in ROL mode: if banked <6 degrees, it?ll try to roll wings level. If on the ground, as one is during the required checks prior to initial in-air activation, the system is extremely sensitive. So sensitive, in fact, that my airplane was sensing the very slight amount that my hangar floor is off from dead level. So slight that it?s entirely imperceptible to me. Still, the AP has been trying to roll wings level and, because the ADAHRS reported that the airplane wasn?t actually moving (being on the ground), kept trying all the way to maximum aileron deflection. Once HDG mode is selected, it works fine.

I?ve pointed out to Garmin that the sensitivity of the system to extremely minor deviations from level (such as my imperceptibly banked hangar floor) should be made much more clear in the manual. Also, the single mention of the system?s behavior in ROL mode could be more effectively called to attention in the documentation. Believe me, I?m a very detailed reader of manuals (I probably read this one half-a-dozen times, cover to cover), but this escaped me.

Given that this is a normal behavior (it's a "feature," not a bug), it shouldn?t have been a mystery to Garmin. A little disconcerting, but at least we?re all on top of it now. Hey, on the bright side, I?m not racing to make SnF anymore!

In other words, when the AP is turned on (with heading bug centered) in Hdg/Pit mode it wants to roll left immediately and will not respond to turning the heading bug right

Click to expand...

Sounds like the only mystery here is why in your first post you stated you were in 'HDG' mode and now you want to blame Garmin for their system being to sensitive in 'Roll' mode?.

Walt said:

Sounds like the only mystery here is why in your first post you stated you were in 'HDG' mode and now you want to blame Garmin for their system being to sensitive in 'Roll' mode?.

Click to expand...

No real mystery at all. The AP comes up in ROL & PIT modes and, apparently due to my very slightly off-level hangar floor and the system?s barely documented behavior of attempting to roll wings level at very small angles of bank, it immediately began to deflect the stick. Once that movement had begun, switching to HDG mode, which I had done as reported, failed to arrest the stick movement.

I don?t want to ?blame? Garmin for the roll mode sensitivity, but would like to see some improvements in their documentation of the ?feature? as it may affect the ground testing of the system as called out in the manual.