Autopilot sensitivity

T

thito01

Pre-takeoff checklist
Joined
Aug 10, 2020
Messages
224
Location
Central Florida
Display Name
Display name:
Tony T
Aircraft: PA32R-300, GNS430W, Sandel SN3308, CenturyIII.

When I first purchased the aircraft in 2005, it had a non-WAAS GNS430 and the autopilot tracked the course well in all selected options (NAV, OMNI, HDG, etc). When I had the GNS430 upgraded to a GNS430 (2009), I started seeing sensitivity issues. When in NAV or OMNI, the wings would rock back and forth in a very annoying manner. In HDG mode, the wing rock was there, but not too bad. The shop that did the WAAS upgrade 'checked everything' and could not come up with an answer. The 'final answer' was to reduce the sensitivity on the CenturyIII using the left bank and right bank screws (A and C on Fig 6.2 in the manual). Nobody ever explained why a WAAS upgrade would cause this. (FYI, the required audio-transformers between the EHSI and the Century were verified and tested.) I was told to 'find a Century expert' but that they knew of none they could recommend.

While it helped a lot, HDG worked acceptably but I have not really been able to use NAV or OMNI since. They track, but the flight is not pleasant. I have just gotten use to always using NAV mode and dialing in my own wind correction.

I am getting back into IFR currency and have been watching a lot of GNS430 videos on YouTube. In one of those, they discussed that the flight CDI sensitivity is reduced down to tighter margins in all flight modes with the WAAS upgrade. Down to just 1 mile each side within 30nm of a waypoint and only 2 miles outside of 30nm. This triggered an explanation that the radio shop never could provide. For the first time, I now think this is not just a problem specific to my installation. Maybe someone else has seen this problem and knows a better correction other than playing with the left/right bank adjustments.

Anyone seen this? Was there a resolution?

Thanks.
 
Welcome to my world...I've been battling a Century III problem of "left/right bank", back and fourth, for 6 months. No one can find the problem. I'm $9000 in the hole (rebuilt servos, rebuilt gyro, rebuilt head unit) but the problem is still there. For $20k I could have had a new AP. Take my hindsight and don't spend A PENNY on that old crap. Just replace it.
 
Your symptoms remind me of one that was resolved:
My Century2(Piper Autocontrol 3B) was behaving that way in heading mode while following a heading, after I put in my Aspen, but not before. I wondered if this was the new normal and tried to read everyone’s suggestions about adjustments and servo rebuilds but did not act-as it turned out, glad I did not. The funny thing was that the autopilot was beautiful and smooth when in GPSS mode(connected to heading mode from in the autopilot)from the GPS signal from my 430W, so I lived with it. However, after a flight with my CFI friend, he insisted that was definitely not OK. After checking further with Aspen, it was explained that
there is a installer fabricated circuit board for the "Fine tune excitation input to adjust the gain" of the autopilot. I did not know beforehand that this was to be part of the install. It turned out that just changing the resistor in that circuit to an appropriate resistance made all the difference. It was easy, and in my case, something that was never done properly in the install.

I would suspect that the newer digital autopilots may not need that interface, but the legacy analog ones, like mine do. So your installer should have just such a circuit, and if similar to my situation, the value of the resistor in that circuit may need to be changed. If that is it, you won’t have to curse out your legacy autopilot, which in the case of the Century 2, is very capable and worth keeping. YMMV
 
I have this same problem. In my case, the roll oscillations developed after having the AI overhauled. I have thrown a little bit of money at it by letting my mechanic re-check all the connectors for loose pins/bad connections, but that made no change in the behavior. I'd like to fly it to Autopilots Central in Tulsa, but it's tough to carve out a window of time to do that.
 
Century IIB here. Mine was always a little sloppy with a slow oscillation that was particularly bad in heading mode when tracking the heading bug. I then put a 650 and a GI275 in, and it tracks a GPS line via GPSS (also in HDG mode) absolutely perfectly. I set about trying to improve the heading mode oscillations (when tracking the heading bug) by increasing the sensitivity or “gain” of the instrument. This was a lot harder to do than expected, tbh, but eventually I was able to increase the gain enough to nearly eliminate the rock in heading bug mode, but the downside is that I’ve introduced a high-gain (too sensitive) hunting in GPSS mode.

I pretty much figure... oh well. I’ll turn the gain down a bit to restore the really quite good operation in GPSS mode, as that’s the more important one. Frankly, it does that so well that it’s hard to be very frustrated.

@MajorTurbulence you may have described my exact problem. That interface you mentioned - is that the “radio coupler” perhaps? Or maybe for Garmin based systems it would be the GAD29B (I think?). Any more info you can provide would be greatly appreciated as that might put my little issue to bed permanently.

Now... if someone can explain a low-cost option for adding altitude hold to my system...
 
After reading the replies, I went back and reviewed the wiring diagram for the Sandel SN3808 to Century connections. I have a 1C-388-2 Radio Coupler which requires (3) 10K-10K audio transformers. Two of which also need a 10K resister between the output sides. While this box was tested when I first had the problem, I was thinking that maybe a different transformer might help. But(!), then I read the notes on the page and found that there is an settings screen in the SN3808 that can be used to make adjustments to the handle overshoot, undershoot and stability issues. And, it's intended to be adjusted in flight by trial and error starting with some default settings. "MajorTurbulence" comments were very helpful.

I will try to get in the air in the next few days and give it a try.

AA5Bman, the altitude hold is based on a billows assembly mounted behind the glove box. I have been told that if it ever failed, I would have to live without altitude hold until I replaced the A/P because they are hard to find and costly when found.
 
Adjusting the Left/Right bank limit pots will not remove a roll oscillation issue. It will adjust the maximum bank limit the system is able to produce. Depending on the unit Century 2 VS Century 3 there is a way to adjust the feedback in the roll circuit called Roll THD. On the Century 3 it is a adjustable pot in the main amplifier typically a 1C515-1. For the Century 2 if the unit has the adjustment it is behind the face plate. Normally in flight you would adjust the pot just until a roll oscillation is present then back the pot off until the oscillation stops. Nav and OMNI modes are the weakest tracking modes of a Century 2/3 system.

Depending on the install if isolation transformers are used there is typically a voltage divider resistor circuit that will adjust the "gain" of the output error. In the Aspen and Garmin IM there are specific values mentioned for the voltage divider. I have seen were the advertised values did not meet spec for heading/course gain. Century maintenance manual says 10 deg of heading error should be equal to 10 +2/-1 deg of attitude error. Both Aspen and Garmin provide a means to adjust the GPSS gain without having to adjust the resistor voltage divider. However, if the system can not meet spec of 10 deg heading error to 10 deg of attitude error adjusting the GPSS gain will not resolve a heading issue and the voltage divider must be adjusted.

As for the alt hold issues. Parts are no longer available from Century so we are left to finding usable parts in the field from parts units. The transducer assembly can still be repaired if the shop has the necessary parts.
 
AA5Bman said:
Century IIB here. Mine was always a little sloppy with a slow oscillation that was particularly bad in heading mode when tracking the heading bug. I then put a 650 and a GI275 in, and it tracks a GPS line via GPSS (also in HDG mode) absolutely perfectly. I set about trying to improve the heading mode oscillations (when tracking the heading bug) by increasing the sensitivity or “gain” of the instrument. This was a lot harder to do than expected, tbh, but eventually I was able to increase the gain enough to nearly eliminate the rock in heading bug mode, but the downside is that I’ve introduced a high-gain (too sensitive) hunting in GPSS mode.

I pretty much figure... oh well. I’ll turn the gain down a bit to restore the really quite good operation in GPSS mode, as that’s the more important one. Frankly, it does that so well that it’s hard to be very frustrated.

@MajorTurbulence you may have described my exact problem. That interface you mentioned - is that the “radio coupler” perhaps? Or maybe for Garmin based systems it would be the GAD29B (I think?). Any more info you can provide would be greatly appreciated as that might put my little issue to bed permanently.

Now... if someone can explain a low-cost option for adding altitude hold to my system...
Click to expand...

I don’t have any expertise on this matter other than my personal experience dealing with the problem.
Excluding the most recent knowledgeable commenters, it would seem that too many installers do cookie cutter things to reduce their time, but fail to really understand all the nuances.

But this brought back some memories of additional info that some might find helpful from a necro-thread from 2/2018 “Aspen Evolution vs Dual Garmin G5's”, but fails your need re the GAD29B interface.
From that discussion:
(http://deklintech.com/ea100-interface-unit.html)
The TR1/TR2 unit seems to be available as shortcuts for the Aspen installer dealing with legacy Century autopilots in leu of fabricating their own circuit board to "Fine tune the excitation input "
Again, probably not helpful to you, but maybe useful for others.

Regarding your/ and my wish for a cheap add-on altitude function for our autopilots, wouldn’t that be nice ( not holding my breath)

PS: Oh, I almost forgot. I even have electric trim, if it were only that simple.
 
Last edited:
You must log in or register to reply here.
Back
Top