|
DFC90 in R182 -- help me understand |
Post Reply 
|
| Author | ||
achimha 
Newbie 
Joined: 24 Mar 2013 Status: Offline Points: 25 |
 Post Options
 Thanks(0)
 Quote Reply
 Topic: DFC90 in R182 -- help me understandPosted: 21 Jun 2013 at 1:50pm |
|
|
I just got the DFC90 in my 1979 Cessna Skylane RG (R182) which previously had an Aspen PFD and S-TEC 60-2.
Well, it is not working and the shop doesn't know why. I have to say that they are very competent and experienced, Europe's largest by far. Support calls with Avidyne are in but maybe this forum can shed some extra light onto the situation? I'm the first 182 in Germany to get the DFC90. Apparently there were two done in the UK before but I don't have the contact details of the owners/shops. I also don't know anyone in the states that has it. No experience to draw from. The lateral operation seems to be OK but the vertical operation is totally messed up. Once a vertical mode is engaged (let's say ALT), the DFC90 starts immediately operating the trim servo which it should definitely not do and after a few seconds, the Aspen shows "TRIMMING UP" or "TRIMMING DOWN" in the middle of its AI which I think means there is a trim runaway situation. In my understanding, the autopilot should only operate the trim servo when there is excessive tension on the elevator cable but the primary means of controlling pitch should be driving the elevator servo. The tests were performed on the ground and in the air with the same results. On the ground, pressing ALT should do exactly nothing because the airplane is already holding the altitude. One should just feel that the servo is "blocking" the pitch axis. The cabling is correct, it was verified several times using an S-TEC 55X tester and the system works when swapping the DFC90 for a 55X (although that only provides limited functionality because the setup now uses RS232 for communication between the Aspen and DFC90 which the 55X does not understand). I have read that the original Cirrus setup where the DFC90 debuted used some weird approach of driving the trim as a primary means of pitch control. Maybe our DFC90 was setup incorrectly by the factory? A replacement box should be underway the shop told me. So apart form hoping for some bits of information on the situation my main question is whether anybody with a Cessna 182 successfully installed the DFC90 and whether there's anything to learn from it. I'm confident the issue will be solved and I trust both my shop and Avidyne to do what's required but I'm technically quite involved in everything that is done to my airplane so I am very keen on doing my own research on this. Thanks! Regards, Achim
Edited by achimha - 14 Jul 2013 at 5:44pm |
||
 |
||
|
AviJake
Admin Group 
Joined: 26 Mar 2009 Location: Lincoln MA Status: Offline Points: 2815 |
Post Options
Thanks(0)
Quote Reply
Posted: 21 Jun 2013 at 2:13pm |
|
|
Achim,
Trim will run all the time in flight - it flies pitch in parallel with the
pitch servo. On the ground the weight of the horiz stab will usually cause trim to run unless the pilot unloads that force manually. BTW, there have been a number of successful installs so far. That being said, we're very interested in hearing about any not-quite-right (or way wrong) experiences so we can address them. |
||
|
Steve Jacobson
sjacobson@avidyne.com |
||
|
||
|
achimha
Newbie
Joined: 24 Mar 2013 Status: Offline Points: 25 |
Post Options
Thanks(0)
Quote Reply
Posted: 21 Jun 2013 at 2:34pm |
|
|
AviJake,
thanks for your quick reply. Unfortunately I don't understand what you explained about the pitch servo. How can it run the trim in parallel with the pitch servo? The S-TEC pitch servo has 2 microswitches which trigger at a certain level of torque on the pitch axis and feed that information to the autopilot. The autopilot should then actuate the trim servo to remove the torque. That's what pitch trim is for, to remove elevator forces, isn't it? Are you saying that on the ground the weight of the elevator will get the trim going and given that the weight does not change, it will eventually show a trim runaway alert on the Aspen? We also saw the DFC90 operate the trim wheel when were were holding the elevator, i.e. when there was no moment. The trim actions appeared rather random to us. During flight testing we saw that the autopilot cannot stabilize the aircraft on the vertical plane, it will induce oscillations that become move violently over time. Also there are a few DFC90 config options in the Aspen setup and only some of them are mentioned in the installation manual. Is there a more complete documentation? Could a misconfiguration in the Aspen menu be the reason?
|
||
|
||
|
AviJake
Admin Group
Joined: 26 Mar 2009 Location: Lincoln MA Status: Offline Points: 2815 |
Post Options
Thanks(0)
Quote Reply
Posted: 21 Jun 2013 at 4:31pm |
|
|
The trim does run in parallel with pitch servo when the switches (Out-of-Trim or OOT) are open. Yes the
weight of the elevator is sufficient to close the the OOT limit switch in the
pitch servo when engaged on the ground. The surface is aerodynamically balanced
in flight. That all being said, it does sound like there's something wrong in your pitch loop based on the reported behavior in flight. However, the real kicker is the airplane type you are working with. The STC does NOT cover the turbo'd 182. I missed it on your first post but it looks like you've got a turbo retract. Retract is fine but turbo is not covered by the STC. |
||
|
Steve Jacobson
sjacobson@avidyne.com |
||
|
||
|
achimha
Newbie
Joined: 24 Mar 2013 Status: Offline Points: 25 |
Post Options
Thanks(0)
Quote Reply
Posted: 12 Jul 2013 at 3:56pm |
|
|
Did another test flight of the DFC90 in the R182. Whenever the autopilot is enabled, it immediately starts operating the trim servo and does so constantly. It appears to not operate the pitch servo as the primary means of controlling pitch. It blocks the pitch axis (need a lot of force to move it by hand) and appears to use the trim servo to adjust pitch.
All in all, the autopilot appeared to work but the behavior is odd. Is this working as designed, is it expected to constantly operate the trim servo? The previous S-TEC would do everything with the pitch servo and only operate the trim servo when there are excessive pitch forces (the OOT microswitch in the pitch servo triggers). I think in the Cirrus, the DFC90 does all pitch adjustments with the trim servo but is it supposed to do the same thing in the Cessna 182? Could it be that we got incorrect firmware? We received a replacement DFC90 but it shows exactly the same behavior. Also I noticed that the DFC90 struggles when applying flaps. I was flying 80 KIAS in HDG and ALT mode and then applied 10° and 20° it took the DFC90 quite some time to get things under control (speed down to 60 KIAS, 100 ft altitude gained). I know it's got microswitches in the flaps so it should be able to react. Again, I only saw it operate the trim wheel, not the pitch servo. VS modes were not impressive. I know from the S-TEC that it rarely does the VS one sets and was hoping that the DFC90 is better at that but it did +- 100fpm of what was set.
|
||
|
||
|
AviJake
Admin Group
Joined: 26 Mar 2009 Location: Lincoln MA Status: Offline Points: 2815 |
Post Options
Thanks(0)
Quote Reply
Posted: 13 Jul 2013 at 11:28am |
|
|
Just to be clear, the STC does not cover turbo 182s. Retracts are fine but the turbo is not covered.
That being said, the DFC90 flies the airplane using both the pitch servo and the pitch trim motor. Your description of the movement you are seeing sounds correct to me. As for the behavior you are seeing with flap movement, that isn't too far off the expected mark either. As I'm sure you know, the 182 has both a very high pitching moment and a slow running trim system. So, as you deploy/retract flaps, you get a significant pitch excursion and the DFC90 has to counter that by running the trim and since it's a slow trim motor and pitch servo, it takes a few seconds to catch up and get back on target alt or glideslope. You can demo that to yourself by not having the autopilot engaged and then deploying or retracting flaps and you'll note rather significant stick force changes - it's those forces, using a slow running trim motor, that the DFC90 is trying to deal with. As for VS performance, that too sounds normal. +/- 100 fpm in my opinion is a very small amount and is often within the error of the VSI/air data system accuracy and the trim system of the airplane. Note that I would NOT expect to see vertical porpoising at all when in ALT hold mode - if you're seeing that, there is clearly something not right. It should be rock steady, not moving one iota in ALT hold mode.
|
||
|
Steve Jacobson
sjacobson@avidyne.com |
||
|
||
|
achimha
Newbie
Joined: 24 Mar 2013 Status: Offline Points: 25 |
Post Options
Thanks(0)
Quote Reply
Posted: 14 Jul 2013 at 5:43pm |
|
|
Thank you very much. I think I am slowly gaining an understanding of this autopilot. I think it is rather important for a pilot to understand his autopilot well. I am now leaning towards the DFC90 working correctly but more test flights are required. I only briefly tested it because I feared that the constant trim wheel operation was a fault and it might destroy the servo... I can also confirm your observation about the 182's pitch up moment on flap extension. Even when flying manually, this is a challenge so I guess it's to be expected. I have to do some experiments to see how much airspeed I lose while the DFC90 tries to get the airplane under control and back onto its vertical path.
I have a few more questions: - Why does the DFC90 drive the trim servo without the pitch servo's OOT microswitch closed? The S-TEC would never do that. From how the servos are designed, I figure that the DFC90 has no means of measuring torque on the pitch axis and therefore has to use some internal model about how to drive the trim servo. Why doesn't it do it just like the old S-TEC: control pitch with the elevator servo and when there are excessive forces (OOT microswitch closed), operate the trim servo to remove the forces? What is the advantage of your method? I guess there must be one... - Related to my first question. When I experiment on the ground and e.g. set a climb, I do not observe the DFC90 pull the yoke but instead only drive the trim servo. Given that nothing happens it will show an error "TRIMMING UP" after some time. Why doesn't it pull/push the yoke? You said it uses a "combination" of both, I just never see that yoke movement so it must be small if any. The S-TEC would pull the yoke all the way back and not trim at all during this ground experiment because the torque was not enough to trigger the OOT microswitches. - Wouldn't this constant trim operation wear out the trim servo and the whole trim system with the spindle much faster than in a "classic" setup? I am not sure the 182's trim system is designed for being operated all the time and I don't think any other autopilot would do it that way. - When I landed the aircraft (AP disconnected), the DFC90 warned me about an imminent stall. Well, the autopilot was correct but that was my intention :-) The danger of bogus warnings is that one gets very used to them and might not assign the required importance when one should. Is there a way to make the DFC90 smarter about when it warns about stalls? PS: I changed the post topic so that you don't have a "not working" post on your forum which you probably didn't even deserve in the first place...
Edited by achimha - 14 Jul 2013 at 5:58pm |
||
|
||
|
AviJake
Admin Group
Joined: 26 Mar 2009 Location: Lincoln MA Status: Offline Points: 2815 |
Post Options
Thanks(0)
Quote Reply
Posted: 15 Jul 2013 at 12:58pm |
|
|
Reference your new questions, I'll number them 1 through 4.
1. We designed the system to fly the airplane with trim. Large corrections require the pitch servo but minor corrections use the trim. We also designed the system to attempt to put the pitch servo and the trim motor "at the edges" of their respective deadbands. This means the airplane will be immediately responsive to an autopilot input instead of first having to wind up enough voltage to get the servo moving out of it's deadband. Not doing that, means the first few fractions, or more likely, whole seconds following a command input are spent trying to get the servo moving which results in a laggier and less responsive autopilot. 2. Ground-based tests with no aerodynamic loads on the flight control surfaces are considered poor tests to understand how the system will really perform in flight. No aero loads (or simulated aero loads) equals no valid test. 3. We don't know of any operational degradation in trim motor life expectancy due to this type of use. 4. Careful on how you're using the system. You'll note one of the published limitations is no autopilot use of any kind below 200' AGL. This includes flight director. Now, that being said, it sounds like you are experiencing Envelope Alerting (EA) functionality. We will alert via a "Caution, Underspeed" aural alert at or below 1.2 Vs. The definition of Vs changes based on flap (and possibly other) configurations. Wiring in of the flap position helps the autopilot know which definition of Vs to use (full flaps, partial flaps, no flaps). Additionally, we suppress the Underspeed alert if we think you are in the landing configuration in order to minimize potential nuisance alerts. So, if flaps are set to full, we suppress Underspeed alerting (when the servos are not coupled). Thanks for the thread title change. |
||
|
Steve Jacobson
sjacobson@avidyne.com |
||
|
||
|
achimha
Newbie
Joined: 24 Mar 2013 Status: Offline Points: 25 |
Post Options
Thanks(0)
Quote Reply
Posted: 15 Jul 2013 at 2:04pm |
|
|
Thank you AviJake, this is very helpful!
I did more testing today and I am now generally very happy with the autopilot. A lot of thermal turbulence and it did a very good job at flying approaches, much more stable than the previous S-TEC 60-2. 1. You're saying that it should not output the "caution, underspeed" alerts when the flaps are fully extended (40° for the R182). Of course the AP was disengaged prior to reaching 200ft AGL. Unfortunately, it does give the alert all the time for me, I get it on every landing about 1 second before touchdown. Does this mean the flap position switch is wired incorrectly? I thought it was working because I see it react immediately when I operate the flaps. Is there a good testing procedure? 2. I understand about the issue with ground tests. The S-TEC was easy to test on the ground, its control loops were very simple to understand but the DFC90 is much more sophisticated... 3. I observed the DFC90 give me "TRIMMING UP" and "TRIMMING DOWN" alerts on the Aspen several times, probably due to turbulence on approach. The alerts went away after about 2-3 seconds and they did not seem to impact operation. Is this to be expected or something that should never happen? 4. Why did you choose to beep 5 times on AP disconnect? I don't see that as an improvement over the 3 beeps of the S-TEC and it's a highly annoying tone. Disconnecting the AP is part of the normal operating procedures and in my opinion does not require such a massive alert. I know that volume can be adjusted but I find that the disconnect alert is rather loud compared to the envelope alerting voice. What do you think?
|
||
|
||
|
AviJake
Admin Group
Joined: 26 Mar 2009 Location: Lincoln MA Status: Offline Points: 2815 |
Post Options
Thanks(0)
Quote Reply
Posted: 15 Jul 2013 at 5:48pm |
|
|
1. Yes, you should not be hearing the Caution Underspeed alert on short final/flare/roll-out with full flaps so I would have the flap position wiring checked out. I'll have to check the install manual to see if we provided the proper voltage expectations in there. I know the expected voltage for the various flap positions are included on the installation drawings. Do you have access to those?
2. Roger. 3. That doesn't sound defective to me. Those annunciations appear when trim has been running for more than 4 consecutive seconds. It was put in to alert you to a potentially runaway trim situation. When the servos are really working hard (e.g. very turbulent conditions, flap deployment at the Final Approach Fix, etc) then this can often be observed. If you are seeing this during benign flight conditions, then there is likely a problem. 4. For better or worse, the high annoyance factor was intentional. Too many pre-DFC90 aircraft mishaps involved pilot unawareness of autopilot disconnect. A 2nd push of the autopilot disconnect button will silence those beeps.
|
||
|
Steve Jacobson
sjacobson@avidyne.com |
||
|
||
|
oskrypuch
Senior Member 
Joined: 09 Nov 2012 Location: CYFD Status: Offline Points: 3061 |
Post Options
Thanks(0)
Quote Reply
Posted: 16 Jul 2013 at 11:40am |
|
The more annoying the better. All on board need to be very aware when the A/P is disengaged -- sometimes it is as a result of abnormal operating procedures. It happened to me once, an item fell and hit the A/P disconnect button. I was unaware until moments later when I noticed the beginning spiral. I now have a tone injected into the audio panel, so that won't happen again. Of note, if you are sitting quite forward on a airliner, you will usually hear the A/P disconnect warble all the way into the passenger cabin. Boeing has a similar outlook. * Orest Edited by oskrypuch - 16 Jul 2013 at 11:43am |
||
|
||
|
achimha
Newbie
Joined: 24 Mar 2013 Status: Offline Points: 25 |
Post Options
Thanks(0)
Quote Reply
Posted: 16 Jul 2013 at 11:47am |
|
|
Orest, given that I can silence the beep by pressing AP disconnect again, I think it's great. A warning that cannot be missed but with a convenient way to stop it in normal operation. Somebody clearly spent some thought on this. I like the safety focus of the DFC90, that is rather unique.
I just found the information about how to silence the alert in the DFC90 POH. I thought I had read it well but obviously I haven't!
|
||
|
||
|
oskrypuch
Senior Member
Joined: 09 Nov 2012 Location: CYFD Status: Offline Points: 3061 |
Post Options
Thanks(0)
Quote Reply
Posted: 16 Jul 2013 at 1:44pm |
|
|
Best of both worlds.
* Orest |
||
|
||
|
94S
Senior Member
Joined: 06 Mar 2014 Location: Bismarck, ND Status: Offline Points: 163 |
Post Options
Thanks(0)
Quote Reply
Posted: 07 Oct 2014 at 5:38pm |
|
Steve, I am curious why the STC doesn't cover the Turbo 182 RG (aka TR182). I own one, and was strongly considering the DFC90 for my future autopilot but this thread has me thinking otherwise now. The R182 and the TR182 are identical except for the engine and service ceiling. Was not including the TR182 in the STC intentional or an oversight? And more importantly, can it be added? I currently have an IFD540 on order through the pre-buy program and plan to upgrade the rest of my panel to Avidyne in the future. It sure would be nice if I could include the DFC90. Sincerely, David |
||
|
||
|
94S
Senior Member
Joined: 06 Mar 2014 Location: Bismarck, ND Status: Offline Points: 163 |
Post Options
Thanks(0)
Quote Reply
Posted: 07 Oct 2014 at 5:46pm |
|
Now I'm confused. I just read the "DFC90 for For Cessna Skylane" brochure (copyrighted 2011) available under the thread "DFC90 Available for the Barron, Bonanza, and Skylane" where it lists the TR182 in the "Models Supported" Edited by 94S - 07 Oct 2014 at 5:47pm |
||
|
||
|
AviJake
Admin Group
Joined: 26 Mar 2009 Location: Lincoln MA Status: Offline Points: 2815 |
Post Options
Thanks(0)
Quote Reply
Posted: 09 Oct 2014 at 7:07pm |
|
|
The brochure is wrong then.
We don't have the turbo part of the turbo-retracts due to missing flight test points. We were required to fly several test points for the turbo at the top of its altitude envelope but didn't have access to a turbo airplane during the flight testing, hence the inability to include the turbos on the STC.
|
||
|
Steve Jacobson
sjacobson@avidyne.com |
||
|
||
|
94S
Senior Member
Joined: 06 Mar 2014 Location: Bismarck, ND Status: Offline Points: 163 |
Post Options
Thanks(0)
Quote Reply
Posted: 10 Oct 2014 at 8:55am |
|
|
Is there any chance the Turbos will be added to the STC in future? I'd volunteer mine for the flight tests, but mine doesn't have an Aspen PFD or a S-tec autopilot at this point. And won't for quite some time. That brings up another question. Can the DFC-90 be installed as a stand alone install, or does it have to "plug & play" with an existing S-tec?
|
||
|
||
|
AviJake
Admin Group
Joined: 26 Mar 2009 Location: Lincoln MA Status: Offline Points: 2815 |
Post Options
Thanks(0)
Quote Reply
Posted: 10 Oct 2014 at 9:43am |
|
|
There's a chance but if it happens, it'll be a while. It of course needs a viable attitude source (Avidyne PFD or Aspen PFD).
Until our servos are done and certified, you will need to be "P&P" with STec so that you have a usable set of servos on-board. Edited by AviJake - 10 Oct 2014 at 9:44am |
||
|
Steve Jacobson
sjacobson@avidyne.com |
||
|
||
|
94S
Senior Member
Joined: 06 Mar 2014 Location: Bismarck, ND Status: Offline Points: 163 |
Post Options
Thanks(0)
Quote Reply
Posted: 10 Oct 2014 at 11:37am |
|
|
Thanks. It will be many years before my budget will allow this upgrade. Maybe by then, the turbos will be included, the servos will be certified, and the PFD4000 will be available.
|
||
|
||
|
richholt
Newbie
Joined: 17 Aug 2015 Location: Sierra Vista AZ Status: Offline Points: 5 |
Post Options
Thanks(0)
Quote Reply
Posted: 18 Aug 2015 at 7:58pm |
|
I too own a 1981 Cessna TR182 which has the original Navomatic 300A autopilot and no altitude hold. I am anxiously awaiting the certification of the Avidyne ASA575 servos and the addition of the Turbo RG 182 to the STC for the DFC90/100. Is it any closer? |
||
|
Richard
|
||
|
||
|
AviJake
Admin Group
Joined: 26 Mar 2009 Location: Lincoln MA Status: Offline Points: 2815 |
Post Options
Thanks(0)
Quote Reply
Posted: 19 Aug 2015 at 7:50am |
|
|
I wish it were but I'm afraid not. It's not an active project at this time and is not on the books for anytime soon.
|
||
|
Steve Jacobson
sjacobson@avidyne.com |
||
|
||
|
Post Reply
|
|
Tweet
|
| Forum Jump | Forum Permissions You cannot post new topics in this forum You cannot reply to topics in this forum You cannot delete your posts in this forum You cannot edit your posts in this forum You cannot create polls in this forum You cannot vote in polls in this forum |
Topic Options