AUTOPILOT AWRY

I have a piper Arrow and out of the blue...(no pun intended)... the
Autopilot banked hard left (30 deg.) for no apparent reason. It was on Hdg
Mode and the bug was centered.

Each time it was shut off and re-engaged it continued to bank hard left.

I have a long trip to FL planned over the school break and sure would like
to have the Autopilot working.

Does anyone know why this would happen?

Thank You All.
Bill Snow

First, what is the make and model of the autopilot?
For my respsonse, I'll assume you have a Piper autopilot,
which is based on the Edo-Aire / Mitchelle designs.

In general, a hard bank to one side indicates that that
roll servo is being driven that direction by the amplifier. (Obvious,
uh?). The question is why.

In my Piper Altimatic II, the roll servo is a motor that is driven by
an H-Bridge type of amplifier. This input to the final motor
driver amplifier is a left / right signal which is derived from the
difference in the heading bug and indicated heading, and if in
NAV, OMNI, LOC or REV LOC mode also the left / right
signal from the nav output. Therefore, your trouble could be in
the console amplifier itself or the input(s) to the amplifier.

If the alierons are driven to their mechanical limits, I'd suspect
a bad transistor in the console amplifier output stage (H-Bridge).

If the alierons are driven to their maximum bank position but not
the their mechanical limits, (usually about 30 degrees of bank), then
I'd suspect a stage before the final amplifier.

Since this happens in heading mode, the problem could the the
resolver in the DG, or any of the electronics in the autopilot console
amplifier or the radio coupler.

Ronnie


"William Snow" wrote in message
...
I have a piper Arrow and out of the blue...(no pun intended)... the
Autopilot banked hard left (30 deg.) for no apparent reason. It was on Hdg
Mode and the bug was centered.

Each time it was shut off and re-engaged it continued to bank hard left.

I have a long trip to FL planned over the school break and sure would like
to have the Autopilot working.

Does anyone know why this would happen?

Thank You All.
Bill Snow

Ronnie
It is an Autocontrol II B with HSI. I only tested it in the Hdg mode sine I
had a passenger and did not want to alarm him. When the weather clears I
will give it a test in the other modes to try and isolate the problems/mode.
Thanks for the tips.
Bill

"Ronnie" wrote in message
. com...
First, what is the make and model of the autopilot?
For my respsonse, I'll assume you have a Piper autopilot,
which is based on the Edo-Aire / Mitchelle designs.

In general, a hard bank to one side indicates that that
roll servo is being driven that direction by the amplifier. (Obvious,
uh?). The question is why.

In my Piper Altimatic II, the roll servo is a motor that is driven by
an H-Bridge type of amplifier. This input to the final motor
driver amplifier is a left / right signal which is derived from the
difference in the heading bug and indicated heading, and if in
NAV, OMNI, LOC or REV LOC mode also the left / right
signal from the nav output. Therefore, your trouble could be in
the console amplifier itself or the input(s) to the amplifier.

If the alierons are driven to their mechanical limits, I'd suspect
a bad transistor in the console amplifier output stage (H-Bridge).

If the alierons are driven to their maximum bank position but not
the their mechanical limits, (usually about 30 degrees of bank), then
I'd suspect a stage before the final amplifier.

Since this happens in heading mode, the problem could the the
resolver in the DG, or any of the electronics in the autopilot console
amplifier or the radio coupler.

Ronnie


"William Snow" wrote in message
...
I have a piper Arrow and out of the blue...(no pun intended)... the
Autopilot banked hard left (30 deg.) for no apparent reason. It was on Hdg
Mode and the bug was centered.

Each time it was shut off and re-engaged it continued to bank hard left.

I have a long trip to FL planned over the school break and sure would
like to have the Autopilot working.

Does anyone know why this would happen?

Thank You All.
Bill Snow

As you've been told, lots of things in the path between what's
happening and where it may have started. To isolate some of them out:


1. First try it in ROLL mode. If it still fails, then it's almost
certainly in the "computer" (the black box with the ROLL knob on it).
See if you can get the ROLL knob to have ANY effect in this mode.

2. If it works in ROLL mode, but not in HDG, then try bypassing your
NAV coupler. [This just requires unplugging two cables and plugging
one back in.] You also likely have an HSI - NAV adapter that could be
bad.

3. If the info from your AH is bad, then it will roll no matter what
(but you might be able to offset it, or at least slow it down, with the
ROLL knob in ROLL mode). Those Tyco (AMP) connectors are not very
reliable. Try plugging and unplugging the one between your AH and your
"computer" several times (and make sure it is plugged in good and
locked).

After all that, if even in ROLL mode it insists in rolling
oncontrollably, then you've got a bad computer. You might check for
broken wires on the rear connector housing, or even removing and
re-inserting it a time or to - but most likely it needs to go back in
for service. Century no longer services them, but several other places
do. It's just some simple analog transistor discrete stuff... nothing
high tech.

William,

The Autocontrol II is equivalent to my Altimatic II, the difference
being the altitude control in the later. I'm not sure how the Autocontrol
II B
differs from the orginal Autocontrol II, so some of this information may not
be correct.

The roll circuit is picks off bank information from the attitude
gyro via a RF pickup coil. This pickup is a small coil which
helps to determine the resonant frequency of an RF oscillator in
the roll amplifier. There is a "follow-up" capacitor on the roll
servo that indicates the position of the alierons. The turn trim or
roll knob on the console is another reactive component in this
circuit. With the wings level, attitude indicator level, and turn tirm
centered, all these reactive compoents are centered up and the circuit
is operating at it center frequency. If any one moves, the resonant
frequency of the roll circuit changes. There is a frequency discriminator
in the console amplifier that produces a DC output when the frequency
moves above or below the normal operating frequency. This DC output
is amplified and used to drive the roll servo motor left or right.

In heading mode, the radio coupler takes information from the resolver
(heading bug) in the DG and feeds that into the console amplifer as
a DC signal between the discriminator and the output amplifier stage.
When in NAV, OMNI, LOC, or REV LOC modes, the radio coupler
mixes in a left / right signal from a nav radio source. If the uncommanded
roll happens in heading mode, then none of the radio nav modes are going
to be any better. They will have the problem too.

Therefore, an uncommanded roll could be caused by:

1. Bad pick up in the attitude gyro.
2. Bad resolver in the DG
3. Bad radio coupler (yes, even in HEADING mode).
4. Bad console amplifier
5. Lack of feedback or "follow-up" signal from the roll servo
6. A loose connection between any of these components.

As others have suggested, check all the connectors by reseating
them a couple of times. If that does not work, you'll need to determine
which component has failed. Without a means to substitute a known
good unit or test signal for each of these, isolating which component has
failed bad may not be easy.

When I posted a query about my old autopoilot, the kinds folks on the
net had these recommendations for shops:

"Contact the folks at http://www.bevanrabell.com
Not sure about now, but they used to work on ALL the old stuff...

No longer works some of the older stuff, but may be able to point
you to a shop that does: Bob Ferguson, Autopilots Central at Sparks
Aviation, Tulsa, OK "


Good luck and let us know what you find out.

Ronnie

"William Snow" wrote in message
news
Ronnie
It is an Autocontrol II B with HSI. I only tested it in the Hdg mode sine
I had a passenger and did not want to alarm him. When the weather clears I
will give it a test in the other modes to try and isolate the
problems/mode.
Thanks for the tips.
Bill

"Ronnie" wrote in message
. com...
First, what is the make and model of the autopilot?
For my respsonse, I'll assume you have a Piper autopilot,
which is based on the Edo-Aire / Mitchelle designs.

In general, a hard bank to one side indicates that that
roll servo is being driven that direction by the amplifier. (Obvious,
uh?). The question is why.

In my Piper Altimatic II, the roll servo is a motor that is driven by
an H-Bridge type of amplifier. This input to the final motor
driver amplifier is a left / right signal which is derived from the
difference in the heading bug and indicated heading, and if in
NAV, OMNI, LOC or REV LOC mode also the left / right
signal from the nav output. Therefore, your trouble could be in
the console amplifier itself or the input(s) to the amplifier.

If the alierons are driven to their mechanical limits, I'd suspect
a bad transistor in the console amplifier output stage (H-Bridge).

If the alierons are driven to their maximum bank position but not
the their mechanical limits, (usually about 30 degrees of bank), then
I'd suspect a stage before the final amplifier.

Since this happens in heading mode, the problem could the the
resolver in the DG, or any of the electronics in the autopilot console
amplifier or the radio coupler.

Ronnie


"William Snow" wrote in message
...
I have a piper Arrow and out of the blue...(no pun intended)... the
Autopilot banked hard left (30 deg.) for no apparent reason. It was on
Hdg Mode and the bug was centered.

Each time it was shut off and re-engaged it continued to bank hard left.

I have a long trip to FL planned over the school break and sure would
like to have the Autopilot working.

Does anyone know why this would happen?

Thank You All.
Bill Snow

Hmmm... I have to admit, that's quite a bit different than a standard
Autocontrol (aka Century) for a normal Piper system - and I would ask
if you are sure of your information? The standard "wing level"
indication from the AH is indeed off a pickup coil, but it's about 5
KHz, no RF. And there is no frequency shift - the inside of the
Century AH/AI just has two coils and a metal plate behind them (looks
like an iron butterfly). As the thing shifts from one side to the
other you get a shift in amplitude, either in phase or out of phase
(180 degrees) with respect to the ROLL EXCITATION signal. [Or the HDG
EXCITATION signal, if it goes through the nav coupler.] With the AH
centered in roll, the amplitutde is zero. There is no frequency shift
regardless. It would be great if there was some sensor back from the
ailerons, but there isn't - so it has no way of knowing their position.

Having said all the above, your "list of the usual suspects" is the
same one I would come up with. #3 is the one that surprises most
folks, but if you have a nav coupler then it uses HDG EXCITATION to
drive the ROLL coils, so lose of HDG signal affects just operating in
the WINGS LEVEL mode. You can move the ROLL cable that goes from the
computer to the NAV Coupler to go to the AI instead and restore use of
ROLL EXCITATION for a test.

Hi James,



Thanks for your feedback. I think we are talking about two different

generations of Piper / Mitchell autopilots. Mine is an Altimatic II,

installed in a 1964 Piper Aztec. It obviously is original equipment,

factory installed. All the electronics are composed of germanium

transistors, so that also dates it.



My info all comes directly from the Piper Service Manual entitled

"Autocontrol I & II, Altimatic I & II Service Manual". The Piper part

number is 753 798 with a first publish date of June 1961. It contains

permanent revisions numbered PR710920 and PR721020, dated

September 20, 1971 and October 20, 1972, respectively. There are

also pages with various dates between these dates.



The 10Mhz frequency information is contained in the theory of operation

section and it describes in good detail the construction and operation of
the roll

vane in the AI. It shows it as being composed of an Iron Ferrite piece

on one side and of an Aluminum piece on the other. As the aircraft

banks to the Iron side, it says the inductance of the circuit is increased
and

when it banks to the Aluminum side, the inductance of the circuit decreases.

It also details a similar pickup in the AI for pitch and a variable
capacitor in

the DG, variable capacitors in the each servo's "follow-up" circuit and
variable

capacitors in the amplifier and control console. It goes step by step on
how

a change in bank and a change in pitch and altitude changes each reactive

component and how all of the these reactive components work together to

maintain a given resonant frequency.



Maybe the 10MHz figure is just an example - I have not measured the

frequency. However, the roll and pitch servos in my 1964 Aztec definitely

have the feedback or "follow-up" capacitors. I've seen them with my own
eyes

and checked their timing marks to see if they are lined up properly.



There are RF type variable capacitors in the control console and in the
console

amplifier to adjust the circuit reactance (level the wings and level the
pitch). There

is also an RF type variable cap in the altitude bellows assembly to detect a
change

in pressure altitude. The trim-turn and pitch controls on my control
console are

also variable caps, along with an altitude / pitch offset adjustment.



The amp has two cans which look like IF transformers with paper labels on
them

that state "Do not adjust discriminator". The whole thing is connected
together

with coax cables and the service manual is clear about using the proper
length

and type of cables otherwise the resonant frequency will be shifted. The
service

manual shows a drawing of these cable as being composed of a hollow tube
that

contains a curly wire on the inside and a braided shield on the outside.
The whole

thing is covered with an outer jacket of rubber or plastic material. It
reminds me

of the radio antenna coax used in automobiles in the 1960s.



When I went looking for an instrument shop to overall my AI, I ran into
several that

stated they would not work on "those old RF type gyros". All of this taken
together

makes me believe that the system does in fact use an RF signal as the method

to detect changes in pitch and bank and to detect the position of the
servos.



From the limited information I can find, it appears that the later Mitchell
/ Piper

versions switched to the pickup system that you describe. Maybe this was
the

Autocontrol IIB / Altimatic IIB and later versions? I don't know.



I wish I had a schematic of my console amplifier. My service manual is
lacking

in this area. It treats the amplifier as a black box. It does provide a
schematic

of the radio coupler, but no theory of operation information, or other
useful info

like the excitation frequency it supplies to the DG, voltage levels, etc.
Any

information you'd care to share along those lines would be greatly
appreciated.



Just FYI, the earliest Autocontrol and Altimatic systems used what the
service

manual refers to as a "0" heading directional gyro. This device did not
allow a

specific heading to be selected, but rather just indicated a change in
heading which

was used to keep the wings level. Also, these early models had amplifier
modules

attached to the back of the AI and to the back of the DG. There was no
central

"console" amplifier. When I first got the service manual and saw this, I
knew these

didn't exist in my airplane and thought there were missing components.
However,

as I dug in more, I finally figured out that these were consolidated in a
later version

and that the DG was changed to a resolver type wired to the radio coupler
rather

than the older RF type wired to the autopilot amplifier.



What is the excitation frequency of the radio coupler to the DG in the later
Mitchell /

Piper autopilots? Maybe it is the same in mine.



FYI, my coupler has a 4-pin Amphelnol connector for the radio input and a

9-pin Amphenol connector for connection to the DG and autopilot amplifier.

In my case, the coupler could not be eliminated from the system by simply

moving connectors around.



I'd be happy to learn all you're willing to teach me about these things. If
you have

any schematics or theory of operations on the early Autocontrol / Altimatic
I/II

amplifiers and/or on the early radio couplers, I'd really appreciate getting
a copy.



Ronnie



"jmk" wrote in message
ups.com...
Hmmm... I have to admit, that's quite a bit different than a standard
Autocontrol (aka Century) for a normal Piper system - and I would ask
if you are sure of your information? The standard "wing level"
indication from the AH is indeed off a pickup coil, but it's about 5
KHz, no RF. And there is no frequency shift - the inside of the
Century AH/AI just has two coils and a metal plate behind them (looks
like an iron butterfly). As the thing shifts from one side to the
other you get a shift in amplitude, either in phase or out of phase
(180 degrees) with respect to the ROLL EXCITATION signal. [Or the HDG
EXCITATION signal, if it goes through the nav coupler.] With the AH
centered in roll, the amplitutde is zero. There is no frequency shift
regardless. It would be great if there was some sensor back from the
ailerons, but there isn't - so it has no way of knowing their position.

Having said all the above, your "list of the usual suspects" is the
same one I would come up with. #3 is the one that surprises most
folks, but if you have a nav coupler then it uses HDG EXCITATION to
drive the ROLL coils, so lose of HDG signal affects just operating in
the WINGS LEVEL mode. You can move the ROLL cable that goes from the
computer to the NAV Coupler to go to the AI instead and restore use of
ROLL EXCITATION for a test.

Sorry if this is too obvious, but you are sure it is not actually in
wing-leveler mode with the wing trim turned off to the left?

--
Best Regards,
Mike
http://photoshow.comcast.net/mikenoel
"William Snow" wrote in message
...
I have a piper Arrow and out of the blue...(no pun intended)... the
Autopilot banked hard left (30 deg.) for no apparent reason. It was on Hdg
Mode and the bug was centered.

Each time it was shut off and re-engaged it continued to bank hard left.

I have a long trip to FL planned over the school break and sure would like
to have the Autopilot working.

Does anyone know why this would happen?

Thank You All.
Bill Snow

Autopilot banked hard left (30 deg.) for no apparent reason. It was on Hdg
Mode and the bug was centered.

Each time it was shut off and re-engaged it continued to bank hard left.

Does it keep the wings level when NOT in heading mode?
--
Jay Honeck
Iowa City, IA
Pathfinder N56993
www.AlexisParkInn.com
"Your Aviation Destination"

I had a failure like this. It wasn't the computer but one of two
transistors in the location of the servo. The avionics guy told me they
each process (?) turning signals and sort of balance each other out for
level flight. If one goes, the other is free to bank the plane without a
signal from the broken one to bring it back. Now his explaination wasn't
the best and this was many years ago.....but that was my experience. In
my situation, it would bank and never level out, I imagine I punched it
off about 30 degrees or so, but it would have kept going. On the ground
you can demonstrate the failure with the heading bug.

But I have a seneca with a king autopilot and the parts were under the
seats, computer in the tailcone.

In case it helps..

John

William Snow wrote:
I have a piper Arrow and out of the blue...(no pun intended)... the
Autopilot banked hard left (30 deg.) for no apparent reason. It was on Hdg
Mode and the bug was centered.

Each time it was shut off and re-engaged it continued to bank hard left.

I have a long trip to FL planned over the school break and sure would like
to have the Autopilot working.

Does anyone know why this would happen?

Thank You All.
Bill Snow