Dynon Autopilot Problems

Can someone please help solve two Dynon problems? I have an RV8 with a Dynon FD180, HS34, pitch servo and roll servo.

The FD180 worked fine until the HS34 and servos were installed, along with the 5.1 version of software.

The first problem's symptoms are:
1. Just as power is increased for takeoff, the DSAB fail warning comes on, typically followed by the "Lost HS34" warning after takeoff. Both warnings can be acknowledged, and the FD180 continues to function. Usually before departing the pattern the "Lost Roll Servo" then "Lost Pitch Servo" warnings come on, and can be acknowledged.

2. Shortly after takeoff, everything operates normally: The HS34 functions properly; the autopilot can be engaged in both axes.

3. Everything seems fine until arrival at the next airport, with all of the symptoms in #1 returning. The data displayed on the FD 180 is still correct, but sometimes no other display can be selected (as in shifting from EMI to HSI 1/3 page). On one occasion, no warnings could be acknowledged, and the FD180 unit refused to turn off on landing. On the next leg the FD180 recovered, functioning normally, even after the DSAB and HS34 warnings reappeared!

What we have done so far to try to solve the problem:
1. For trouble shooting, the DSAB and power wires to the servos were disconnected. The HS34 functioned normally, and no DASB, HS34 or Servo warnings were present.
2. Then the DSAB wires were reconnected, but not the power wire to the servo. All of the problems in #1 reappeared.
3. The HS34 DSAB connection to the FD180 was changed to a twisted pair (to reduce the effect of RFI). All of the problems remained.
4. The wiring to the servos is by multi-strand, shielded wires. The third thing tried was to ground the shield at the servo end only. All of the problems in #1 remained.

Therefore, it appears that the problem lies in the DSAB wiring going to the servos. But what is the problem?

The second problem is instability in the autopilot. The sensitivity has been adjusted to a level that seems to minimize the pitch and roll excursions, but the heading continues to wander +/_ 10 degrees from the assigned heading. Flying in "Course" helps a bit, but the heading never settles out. With only one person aboard, the pitch control is reasonable but not perfect. However, the addition of a rear seat passenger results in a diverging pitch oscillation which becomes more violent and with greater pitch oscillation with time. In other words it appears to be negatively damped. Roll oscillations are also worse than with only the pilot aboard. Can anyone help with this?

Craig Taylor N5678C

Craig,

Have you contacted Dynon? They will help you work through this.

Also, have you udpated the servos and the HS34. (Note, it sounds like you do not have the AP74, autopilot controller.)

I woulnd't worry about the stability issue until you have the DSAB working properly.

Did you update the EFIS/AP software with the new sport program available from Dynon's site or did you use the old program?

Again, check with Dynon, they will be able to work you through your issues.

I occasionally have a similar problem with the DSAB, but it only happens on engine start. I think what is happening if the battery is weak when I crank the engine, the system voltage drops to less than 10 volts. When that happens DSAB connections sometimes fail. If I get a DSAB failure, recycling power to the Dynon system fixes it. It has never happened when the system is up and running along with the engine running.

Regarding the AP, I changed out the servos to the larger 42's because for the pitch servo in particular, I was getting a lot of slippage and the AP did not do a good job. Right now I have torque to the pitch servo set at 85% which is more than the 32 can put out. The AP performs well in smooth air. In turbulence the heading may wander +/- 4 or 5 degrees, and the altitude may wander 50'. In smooth air heading is within a degree or two and altitude generally within 20'. Overall I'm satisfied with the Dynon. I have never flown a Tru Trak so can't compare performance, but from what I hear the TT may do better in turbulence than the Dynon. I think Dynon still has some tweaking of their software left to do, but again the system is working pretty well for me.

I don't know what to tell you about the electrical problems you are having, but the stability issues you are experiencing may not have anything to do with it. If you have the new servos with the shear screw mod (or you did the mod yourself) and are still having the problem with roll and pitch excursions - welcome to the club. You can eliminate some of it with sensitivity adjustments, but it is still not that good. My pitch excursions seem to get worse with each correction too. I currently have my roll sensitivity set at 5 (any more than that and it rocks back and forth continuously) and my pitch is set at 20. I increase from 12 to 20 yesterday and the altitude excursions went from about 50-60 feet each way to about 30 feet. In roll using heading it is not stable, but in NAV or TRK is does OK in smooth air. In turbulence you'll basically zig-zag your way to your destination. Dynon says there are RV-8's with good performance from their autopilot, so I don't know what the difference is. If you call them at (425)402-0433, they will help you. That's what I am doing and hoping there is a fix for this. I only know personally of one other guy with the Dynon autopilot and we are both experiencing the same results. They'll want you to set your D180 up for data logging so they can troubleshoot, so you might want to do that and collect some data. Dynon support has a forum, but you need technical help that goes deeper than forum support, so I would suggest, as WR did, that you just call them. They'll bend over backwards to help you.

Scott

Process of elimination - ship's power??

Craig:

What's interesting is that you experience anomalies when you power up and power down. Does the FD180 have an internal battery? If so, can you test w/o ship's power by pulling the breaker/fuse to the FD180?

I build/repair lots of pc computers and in the past, alternators & generators. Aberrant power can cause all kinds of crazy events via noise and spikes so it would be good to eliminate one possibility.

I assume all devices are grounded to one central place and all grounds are rock solid. My tech counselor was adamant on that issue. His technological explanations were over my pay grade but the results of his advice has been nothing short of excellent.

The datalog records voltage so that would be a good thing to check. Consider setting it for 1 second intervals but transient spikes & sags can be a lot shorter than 1 second.

ROLL: I also have my roll sensitivity set at 7 on my D180 / AP74. Like the man said, any more than that and it rocks back and forth too much.

Hope this helps.

Barry
Tucson

dynon problems

Bill, Jim, Scott and Barry,

Thank you all for your thoughts and inputs. I think that Barry may have found a possible lead. In response to your thoughts, here is what we have done so far:
1. We have been in continuous phone contact with Mike at Dynon, and he is baffled. I have posted the problem letter on the Dynon support forum, but have heard nothing yet.
2. I do not have an AP74, as I felt that for occasional autopilot use that the D180 buttons would be adequate. I do like the simple control offered by the HS34, through. All units (FD180, HS34, Roll Servo, and Pitch Servo) have been updated using the current procedure on the Dynon site.
3. I have the internal battery in the FD180, so I shut off the Avionics Master and alternator field during starts to prevent voltage spikes. The internal battery maintains the FD180 very nicely.
4. We have changed the 32 servo for the pitch axis to the 42 unit because f slippage in the 32. The bigger servo if anything has too much authority. We have the linkage set to the second hole from the end of the bellcrank.
5. The 32 roll servo seems quite adequate for roll.
6. I probably misled Barry by using the "power up" phrase; I should have said when I push the throttle forward for takeoff I get the DSAB and/or HS34 fail.
7. Barry, the one thing that I have not checked is the single point ground. The servo wire shielding is grounded at the servos, so maybe I should change the ground to the front using the D180 ground. I will also check for a single point ground for all of the other avionics. Thank you for this suggestion.
Any other ideas out there? Thank you all for your inputs!

Craig Taylor
N5678C

gear1 said:

...
7. Barry, the one thing that I have not checked is the single point ground. The servo wire shielding is grounded at the servos, so maybe I should change the ground to the front using the D180 ground. I will also check for a single point ground for all of the other avionics. Thank you for this suggestion.
Any other ideas out there? Thank you all for your inputs!

Craig Taylor
N5678C

Click to expand...

Craig,

FYI, I have been flying with the Dynon AP as a Beta tester for more than 100 hours and a year now. (Up until last weekend's incident.) In all that time I never ran into the problems you are seeing. My servos, D100, & AP74 are all on an avionics master and get powered up shortly after engine start. (That includes multiple test servos I swapped out during testing.)

If I leave the D100 powered up on its internal battery, then yes, I do get a DSAB error when the other units are off line. Once I turn the avionics master back on, the DSAB self clears as the other network devices come on line.

There is no need for shielded wire and in my installation, I don't even have twisted pair wires, and it works just fine. (Worst case installation was what I was going for.)

I agree, check the grounds and all your connectors. Some connection is dropping out. If you used DB9 connectors as I detailed on my web site, check to make sure all the pins (male and female) are seated properly. Also check the crimps.

Next time you get a DSAB error, check the configuration screen and see which device is not communicating. That might help you isolate your issue.

It is possible that insulation on one of the wires is compromised and is shorting out when you accelerate.

flickroll said:

I occasionally have a similar problem with the DSAB, but it only happens on engine start. I think what is happening if the battery is weak when I crank the engine, the system voltage drops to less than 10 volts. When that happens DSAB connections sometimes fail. If I get a DSAB failure, recycling power to the Dynon system fixes it. It has never happened when the system is up and running along with the engine running.

Click to expand...

If you have battery backup in the EFIS, this is a normal occurrence. It is not the DSAB failing but rather the HS34 shutting itself down for low voltage. Once it comes back on it takes a moment for the DSAB to "wake up."

More on alternators and wiring

Craig:

You said: Barry, the one thing that I have not checked is the single point ground. The servo wire shielding is grounded at the servos, so maybe I should change the ground to the front using the D180 ground. I will also check for a single point ground for all of the other avionics.

During my 18 month build, I was very fortunate to have three electrical engineers guiding me. Two of them were repeat RV builders. Gil A was adamant that all leads ground to the same large copper block & each ground had its own nut plate. It had something to do with impedance and potential differences in the current flow to and fro. Like I said, it was over my pay grade but I know when to listen to the combined experience of three very smart fellas.

Second, an alternator spinning at engine rpm 2500 can be a whole different animal than an alternator spinning at engine rpm 700 so don't discount what a partially defective alternator could contribute at different rpm. Ergo, testing via battery alone makes sense.

Finally, I also used the best aviation grade wire that the three EE's recommended. They didn't care how much it cost & neither did I given the application.

Stick with it and report the results of each test so we can collectively help you solve the prob. Many people helped me debug my 9A. I'm simply returning the favor.

Bottom line: these airplanes are just an absolute joy to fly. After 155 hours, I still walk away after each flight muttering something like "UNBELIEVABLE".

Have you tried pulling the alt breaker and make a quick flight to eliminate alternator noise as a cause?

Gregg and Barry,
I think you may have hit on something. The aircraft that I have had not been flown much in the past three years; when I first started flying it regularly, there were two flights that the generator showed some voltage instability. I will be flying on Saturday, and will shut down the alternater field and pull the breaker for a few takeoffs and landings. (The entire avionics package only pulls about 8 amps, so I could fly for a long time on the Concord battery...)
I will post on Saturday as to what I found! Again, thank all of you for some great ideas.

Craig

PS: I still have not heard anything from Dynon since I posted the same letter on their forum early this AM.

If you have a digital volt meter in the airplane be sure to monitor it. Like I mentioned earlier, if the voltage drops below 10v, you'll likely get a DSAB error.

Craig,

I can't speak to the DSAB error messages (been lucky in that I haven't seen those), but concur with the others' thoughts.

A couple things do come to mind, so thought I'd mention them to add to your troubleshooting regimin:

- You mentioned version 5.1 in your OP. If you have not updated to 5.1.1, you might try that, as it seemed to fix some attitude fluctuations (not your observed issues, but who knows, you may have stumbled on a new glitch in 5.1, that 5.1.1 may help alleviate). I went straight from 5.0 to 5.1.1, and have seen no issues.

- For roll issues, you might try a fresh compass cal, as that did wonders for my early roll control issues. I've actually done one after each software upgrade, to be sure the upgrade didn't trash the compass cal (might be overkill). Heading hold is still not as solid as NAV or TRK, but is now at about +/- 2-4 deg, and is pretty stable (SV-32, 100% Tq, sens at 6). I don't have the wing-waggle issue mentioned by others, but if I go much higher it does get a little twitchy in roll. Where I have it set now, the only time it gets a bit wiggly in roll is after some fuel burn-off, and I'm a bit out of aileron trim (AP fighting the out of trim condition). You're probably running your tests in good trim, but thought I'd mention it, just in case.

- My pitch performance improved after the servo screw and castellated nut service, and the divergent oscillations I saw at the outset of testing have pretty much gone (though it took a lot of futzing to get it to improve, and I've had about 15-20 good hours on it, so it's not a long history to speak from). Concur with Scott on looking there as well. FWIW, 15-16 seems to be the sweet spot for pitch sens in my set-up...a bit less in bumpy air (and bumpy air performance is not stellar...still some oscillations that I'd like to get rid of).

Of course, mine is a 6, so there isn't the CG shift with the extra pax, so there may be other issues at play in the 8, as Scott and Jon are still fighting this too, like Scott said.

Hope you make some progress, and like the others, I'm very interested in hearing about your progress. Best of luck!

Cheers,
Bob

DSAB error messages also, but no big deal at all.

I also get DSAB messages, but only when starting the engine. Since the EFIS has a backup battery and the other connected hardware does not, I suspect that is causing the issue when the voltage drops (when starting). As soon as the engine is running, the problem disappears.

Autopilot performance (it's an RV-8): ok on Roll, more or less 'ok' on pitch. We have the SV-32's & AP 74's. If I get the airplane on course, trim and then enable autopilot, all is well. Deviations to different altitude are sometimes OK, but most often result in tremendous overshoot, both higher or lower altitudes and sometimes seem not to smooth out, unless I disable the autopilot, level off manually and re enable. Course deviations are much better and smooth out pretty quickly in all three modes (NAV, TRACK & HEADING).

The -8 is very sensitive in pitch trim regaring W&B, load and power changes. This may complicate the matter?

Dynon has been SUPER in support, and I have to admit that I did not further contact them regarding our pitch issue.

results of test flight

Dear Bill, Jim, Scott, Barry, Gregg, Bob and Arjan,

On Friday afternoon I received a call from Dynon (Mike and Lawrence) who added a raft of ideas to try in addition to all of the great suggestions that you all have made. No, the problem is not yet solved, but I believe that we are homing in on it. One of the suggestions that Lawrence made was to verify that the angle of EDC-E10A compass module match the faceplate angle of the FD180. I happen to have a Lucas Angle finder which measures to 0.010, so I measured the ange of the two components and found ti to be > the 1.0 degree specified by Dynon. I adjusted the EDC-E10A to match the FD180 to be within 0.1 degree. This is the reason for the Roll/Rudder/Heading stability check noted in the summary. Another check recommended by Lawrence is the pull-push VSI check is to verify that the static system is operating normally. If this data gives anyone any further ideas, your comments would be greately appreciated! We will be installing the 5.1.1 software in the next week or two, and checking the wiring.

Thanks again to all of you for helping to trouble shoot and sort this problem.

Craig

RV8 N5678C AUTOPILOT TEST
JULY 11/12 2009

1. Preflight
a. EFIS>Autopilot>Setup>Status: Servo Position Pitch
i. Stick Full Forward: 69
ii. Stick Full Aft: -123
b. EFIS>Autopilot>Setup>Status: Servo Position Roll
i. Stick Full Left -120
ii. Stick Centered -14
iii. Stick Full Right 30
c. Data Log Set 1 sec

2. In Flight
a. Took runway with alternator off and CB pulled
i. Initially got over-current warning (!)
ii. Upon addition of TO power up got DSAB
warning (acknowledged) (Normally get HS34
warning first.)
iii. Turning downwind got HS34 warning
(Acknowledged)
iv. Alternator on, no problems voltage stable
and normal.
v. After HS34 and DSAB warnings disappeared,
the Autopilot worked both axes, but heading
and course wandered. (see 2.e.)

c. Compass Roll and Rudder heading stability check
within 2 degrees both directions
d. Vertical Speed on pull indicates up (no negative VS)
e. Vertical speed on push indicates down with very
slight delay.
f. Engaged autopilot pitch (Note: aft cg, big passenger)
i. Pitch servo reduced torque to 80%
ii. Reduced sensitivity to 8
iii. held altitude reasonably well, but if oscillations
started, it would diverge.
iv. Did not try climb or decent intercept on this flight
f. Engaged autopilot roll
i. Roll Servo reduced torque to 80%
ii. Reduced sensitivity to 5
iii. Heading - wandered badly
iv. Track worked somewhat better than heading
v. Nav - wandered same as track, but did
intercept course.
vi. Increased roll sensitivity to 10. More abrupt
corrections, but wander amplitude decreased.
vii. The course (magenta triangle on compass tape)
seems to wander +/- 5-10 degrees, which is then
is chased by autopilot. This wander is not
reflected in the GPS course.

3. Post Flight
a. Serial Number of HS 34: 01659
b. Data not logged yet (operator not trained yet)