help me understand the technology behind ALT Hold in older AP

WannFly

Final Approach
Joined
Nov 28, 2016
Messages
6,553
Location
KLZU
Display Name

Display name:
Priyo
My introduction to AP was GFC 700 (with a G1000 set up), as i understand , its one of the best in GA , now before we get into garmin vs other other auto pilots.. thats not the intent of this post (though i am fairly certain it wont matter). anyway, after i got the century 2000 AP and flew around witt it, my first impression was it was broke. why? while flying in NAV mode it would do S turns trying to capture the magenta line or the green line. later on it dawned upon me that thats pretty much how the older AP works that do not have GPSS.

now, in GFC 700 when i used ALT hold, it actually did ALT hold, all the time. in century, it does hold but every now and then it raises/lower the nose for now apparent reason trying to go to a diff altitude, i have never actually waited to see where it settles. as i understand, this behavior happens when the pressure changes, most of the times i usually get a new altimeter from center around the same time. what i want to understand is the technology behind the ALT hold and why a newer AP compensates for the pressure change. is there some kind of altitude encoder connected to the newer digital APs?
 
On the Century III which may or may not apply to the 2000 the altitude hold is based on deviation from a voltage signal generated by a potential across a variable resistor. The autopilot "looks" at changes in the voltage from the "target" voltage set for any particular altitude. The resistor varies with altitude so it generates the change in voltage. With the Century III they hadn't figured out how to correlate voltage to a particular altitude since there was no altimeter setting input. Maybe they figured that out in the next generation.

As I understand it the newer autopilots do digitize the air pressure and use the altimeter setting to select the correct altitude.
 
Those old Piper autopilots need a fair bit of babysitting. Our club has an older but nice Archer with Century AP and while "it works" it needs constant babysitting to keep the DG aligned with the compass and that the Nav hold is actually on course... the S turns settle out after a while but I think that's just what you get with a 50 year old electric gizmo
 
Those old Piper autopilots need a fair bit of babysitting. Our club has an older but nice Archer with Century AP and while "it works" it needs constant babysitting to keep the DG aligned with the compass and that the Nav hold is actually on course... the S turns settle out after a while but I think that's just what you get with a 50 year old electric gizmo

Why does anybody expect a 50 year old anything to work if it's never maintained?
 
My introduction to AP was GFC 700 (with a G1000 set up), as i understand , its one of the best in GA , now before we get into garmin vs other other auto pilots.. thats not the intent of this post (though i am fairly certain it wont matter). anyway, after i got the century 2000 AP and flew around witt it, my first impression was it was broke. why? while flying in NAV mode it would do S turns trying to capture the magenta line or the green line. later on it dawned upon me that thats pretty much how the older AP works that do not have GPSS.

now, in GFC 700 when i used ALT hold, it actually did ALT hold, all the time. in century, it does hold but every now and then it raises/lower the nose for now apparent reason trying to go to a diff altitude, i have never actually waited to see where it settles. as i understand, this behavior happens when the pressure changes, most of the times i usually get a new altimeter from center around the same time. what i want to understand is the technology behind the ALT hold and why a newer AP compensates for the pressure change. is there some kind of altitude encoder connected to the newer digital APs?

You might find this interesting

https://mooneyspace.com/topic/2780-pitch-oscillation-with-century-2000-autopilot/
 
This post really strikes me as interesting, in that link above,

Well, problem solved! (For now...)


We had an instrument shop tighten the brushes that carry the pitch signal. The shop didn't think it would solve the problem, but we just completed a flawless test flight. So I am doubly convinced that the sharp, short duration pitch movements (jerky, and in both altitude and attitude hold modes) were due to the pitch signal coming off the AI!!


Beware - the AI bench-checked perfectly when it was not working in the aircraft, so I think the vibration may have something to do with the quality of the pitch signal and the need for the brushes to maintain good contact.
 
Why does anybody expect a 50 year old anything to work if it's never maintained?

Owned our Lance for nearly 4 years now and I've given the Altimatic IIIc attention here and there every one of those years. The autopilot has probably never been in as good of condition as it is right now. As I've never owned another aircraft with an autopilot, I don't have much basis for comparison; however, I am extremely happy with my old technology. You are correct though, you have to give it attention if you want it to perform correctly.
 
Why does anybody expect a 50 year old anything to work if it's never maintained?
I agree totally. The owner of this plane takes pretty good care of it, and when the autopilot works it's actually impressive how well it will follows a localizer or stays on the Nav heading.. but you can tell that it needs some more babysitting than what someone may expect, especially someone who uses a GFC 700

I do love how simple though the old Piper autopilots were

Incidentally, has anyone ever used that L / R roll feature? and why?

upload_2018-3-14_17-2-3.png

I can honestly say I've never used it.. but it does make the instrument seem nice and symmetric!
 
Classic control systems always "hunt". The unit reads that it is too far to the right, so it corrects to the left, it overshoots a bit, then corrects back to the right and so on.

The closer you set the point to correct, the more rapid the oscillation. If you correct when you are 3 tick marks off then you oscillate slowly. If you tighten it up to correcting when it is 1 tick mark off you oscillate faster.

Now there are fancy control systems that are more complicated than that. But I don't think its possible to completely eliminate "hunting".

If you want to learn more google "control system wiki".
 
Last edited:
I agree totally. The owner of this plane takes pretty good care of it, and when the autopilot works it's actually impressive how well it will follows a localizer or stays on the Nav heading.. but you can tell that it needs some more babysitting than what someone may expect, especially someone who uses a GFC 700

I do love how simple though the old Piper autopilots were

Incidentally, has anyone ever used that L / R roll feature? and why?

View attachment 60951

I can honestly say I've never used it.. but it does make the instrument seem nice and symmetric!
I had to use it once to see what it did when you used it. ;)

Seriously, I consider it primarily a backup in case the roll output from heading and NAV both fail.
 
Back
Top