Auto pilot pulsing

[Flyenforfun]

I have an RV-8 with the Dynon Skyview system that I have been flying for three years now and I have a problem with the altitude hold that I've had since day one. In smooth air the autopilot pitch axis pulses and it drives me nuts. I've gone through the tuning procedure many times and tried different setting but still no luck. The only thing the various setting seem to do is speed the pulse up or slow it down just a little. At any rate nothing seems to help I'm starting to wonder if its mechanical or something electronic in the ADAHRS or pitch servo.
Roll axis seems to work fine. Any suggestions would be appreciated.
Matt

 

[jakej]

Matt

Have you turned off everything you don't need for flight eg Transponder & strobes etc ?

What type of coax antenna cable do you have ?

Is there any free play in the elevator linkages ?

 

[dynonsupport]

Matt,
If you trim slightly up or down when the AP is engaged, does it stop? If so, you have some slop somewhere that the autopilot is hunting against.

 

[Flyenforfun]

Yes if I trim slightly up or down it does settle down a bit but it does seem to be hunting

 

[Flyenforfun]

Matt

Have you turned off everything you don't need for flight eg Transponder & strobes etc ?

What type of coax antenna cable do you have  ?

Is there any free play in the elevator linkages ?

No, I haven't tried that and the linkage seem good with no play. Are you suggesting some kind of electronic noise?

 

[jakej]

Anything is possible but usually can be any of these -

1. Free play in linkage.

2. Poor design/ location of static ports

3. Cheap antenna coax eg RG58 - IMO RG400 ( that's what I use) or equivalent is the way to go.

4. "Noise" - from Transponders & Strobes & high current wiring being close to servos.

5.  Software - sometimes a 'bug', but not likely in this case.
These are not in order of importance, just main items to look at in diagnosing issues. 

The easiest, cost effective, troubleshooting method I use is what I'd mentioned before, turn off everything electrical not absolutely needed for the test flight.  Then turn on the various avionics, one at a time, (wait to see if the issue manifests itself) & note the results.

 

[Flyenforfun]

Are used RG 400 but I will try the other things turning everything off when it starts pulsing again I didn't think of that but that's a good idea.
thanks.

 

[dlloyd]

I have had an occasional similar pulse with the roll servo, never in pitch. Only noticed it a few times in several hundred hours flying. Next time it happens I will check to see if the trim or stobe could be a factor.

 

[Flyenforfun]

I think what I'm going to try next when the weather gets better is to put it in expert mode and go through the indicated airspeed control procedure step-by-step I'm wondering if I have an aronious airspeed set in the skyview and the autopilot is pitching up and down trying to achieve the airspeed but the altitude hold is not allowing it to do that so the plane just starts to Oscilate. The other suggestion someone gave me was turned off all electronics like the strobe and the transponder maybe it's getting electronic noise interference. Either way I'll let you guys know what happens.
Matt

 

[jabarr]

There are many complaining of the same issue on the Vans Air Force website. This has been happening from day 1--- through many firmware upgrades. I wouldn't describe it as pulsing but rather hunting plus and minus in 100 foot cycles.

 

[Flyenforfun]

Mine is real quick forward and aft movments of the stick. I can look back at my elevator and see it pulsing. Actually it's kind of unnerving to watch the elevator while its doing it. I have tried many times tunning it out. Im going to reset everything back to default and try the tunning procedure again of that doesn't work im going to start talking to Dynon.

 

[vlittle]

Casting my memory to my Control Systems courses from decades ago, what you are seeing is the characteristic of what is called a bang-bang controller. It is characteristic of a system that has too much free play/hysteresis.

One way to check is to turn off auto-trim during cruise flight. Slightly de-trim the aircraft and see if the autopilot exhibits the same behaviour when re-engaged.

Many aircraft have non-linearities built into their control systems deliberately to balance control forces for minor corrections. Good for people, not so good for computers.

Vern

 

[krw5927]

I'm one of the ones who also posted on Vans Air Force.  No auto trim, and even when the pitch control is loaded by putting it slightly out of trim I experience some mild pitch hunting with Skyview AP.  Not +/- 100 feet or anything like that, just a perceptible pitch hunt that you can feel.

This has always happened, and only occurs in smooth air.  My autopilot is tuned for my RV9 pretty great with just the default settings.  I tried messing with some settings to eliminate the slight hunt, but each change adversely affected other operations like capture/hold altitudes or capture/follow glideslopes, etc.

 

[vlittle]

This is speculation, but I have zero hunting on my Harmon Rocket, so let's look at the differences between, say an RV-9a and an HR-II. I have built and flown both.

I found the RV-9A to be very sensitive to trim and very light in pitch control for small pitch excursions. The HR-II, despite being light in the controls is more progressive in pitch controls. The Rocket has the RV-4 style tail, with exposed elevator horns, whereas the Rv-9a has the enclosed elevator control horn tips which have a higher break-out force.

This break-out force is a 'non-linearity', added to aircraft to make them more pitch stable. There are other differences, but this may be one source of error, in addition to normal backlash.

The other difference is that I designed my own auto-trim system for the HR-II. It exhibits no hunting and the autopilot is stable. Perhaps my integration filter is set up differently so that it does not fight with the Autopilot. Combine the aerodynamic non- linearities with the dynamics of the autopilot control system plus the auto-trim control system and things get complicated.

It might be useful if this hunting can be correlated to elevator design.... For example, do RV-4s have better results than RV-8s?

Vern

 

[jakej]

Has anyone experiencing this issue disconnected the static line & just left it sampling cabin air ? 
That's a tecnique used to determine whether a poorly place/functioning static system or port has any impact on Alt hold in other autopilot systems.  This method immediately eliminates one possible factor simply.

Subtle differences in static ports, & placement, can have an impact - we are 'experimental' so it can happen unlike 'certified' which are proven before release.

 

[krw5927]

Has anyone experiencing this issue disconnected the static line & just left it sampling cabin air ? 

Nope, don't want to invalidate my aircraft IFR cert.  Too expensive and time consuming for simple troubleshooting.  The good news is, in the Midwest I rarely find smooth air.   :

Vern may be on to something. The RV9 is indeed very pitch sensitive. I monkeyed with a lot of settings though, and nothing seemed to totally eliminate the pitching. Everything I changed seemed to also have a negative impact on the autopilot performance in other, more important (to me) areas. So I live with it. It's really not that bad, just noticeable. Something you can "feel" rather than any measurable altitude excursions.

 

[led]

Hi,

There have already been a number of good suggestions in this thread, so I don't have a lot to add, but I wanted to mention some useful tricks and clarify what people are seeing.  If you're having a hard time chasing down a problem, the best thing is to get in touch with tech support so we can give you individual attention.

There are two different issues being discussed here
- Altitude hunting
- Pulsing on the stick

Before I get to those it's important to always check two things when troubleshooting the AP:
- Make sure there are no servo SLIPS (annunciated on the PFD screen)
- Check the linkages from the control surface all the way to the servo arm for SLOP (more on that in a minute)

If you have either of those things you'll save a lot of time by addressing them before trying anything else.

On the pitch axis, the advice to adjust the settings per the settings guide to get airspeed hold right is the key to good performance.  One word of advice: don't over do it.  If increasing settings stops improving things, stop there and maybe back off a little and you'll be right on.

Low frequency altitude hunting (many seconds per cycle) is usually a result of having ALT_GAIN set too high relative to VSI_GAIN.  The default settings work well for the majority of airplanes:  ALT_GAIN of 0.6 and VSI_GAIN of 1.5 is pretty close to the middle.

As mentioned, slop can cause hunting too.  If you can trim against the servo slightly and the hunting goes away, that's a good indication of slop in the system.  When we improved the trim sensing and introduced autotrim a number of airplanes turned up with slop which had previously been hidden by the preload due to being out of trim.

Another cause of low frequency pitch hunting is bumping up against the airspeed limiter.  In that case you wont hold altitude well at all.

(Note that the AP either tries to hold airspeed or altitude - it can't do both withtout access to the throttle. )

Higher frequency pitch movements can have a number of causes.
- The AHRS can be bouncing around on a flimsy mount
- The airspeed input is noisy (this will make the AHRS pitch jump around)
- Many of the pitch settings are turned up too high.

The first two are difficult to diagnose without more detailed information; the latter can be addressed by running the airspeed tuning process.

Hope that helps some,

Lawrence Doan
Autopilot Guy