Honeywell KFC225 autopilot roll servo failure

[Peter]

I've just had my 6th failure of the roll servo, since 2002. This time
I got a video of it, showing that the power-up tests pass even with a
dead roll servo:

http://www.zen74158.zen.co.uk/misc-files/kfc225-failure-2005-1.mpg

Evidently Honeywell have not fixed this - even though the KFC225 has
failed on every aircraft whose owner I have managed to contact over
the last few years.

Can anyone offer me any info on this servo, e.g. schematics? I already
have the KFC225 main unit schematics so I have the interface (which is
very simple).


Peter.

 

[martin griffith]

I've just had my 6th failure of the roll servo, since 2002. This time
I got a video of it, showing that the power-up tests pass even with a
dead roll servo:

http://www.zen74158.zen.co.uk/misc-files/kfc225-failure-2005-1.mpg

Evidently Honeywell have not fixed this - even though the KFC225 has
failed on every aircraft whose owner I have managed to contact over
the last few years.

Can anyone offer me any info on this servo, e.g. schematics? I already
have the KFC225 main unit schematics so I have the interface (which is
very simple).


Peter.

Then why do you keep buying them?


martin

 

[[email protected]]

I've just had my 6th failure of the roll servo, since 2002.

"Doctor, it hurts when I do this."
"Then don't do that."

Can anyone offer me any info

Try
http://www.rockwellcollins.com/products/cs/br/flight_controls/index.html
maybe?

Matt Roberds

 

[Mark]

I've just had my 6th failure of the roll servo, since 2002. This time
I got a video of it, showing that the power-up tests pass even with a
dead roll servo:

http://www.zen74158.zen.co.uk/misc-files/kfc225-failure-2005-1.mpg

Evidently Honeywell have not fixed this - even though the KFC225 has
failed on every aircraft whose owner I have managed to contact over
the last few years.

Can anyone offer me any info on this servo, e.g. schematics? I already
have the KFC225 main unit schematics so I have the interface (which is
very simple).


Peter.
--

Did you contact Honeywell in Phoenix Arizona?

Mark

 

[Peter]

Did you contact Honeywell in Phoenix Arizona?

H are looking at it; I have no idea how far up the organisation any of
this has gone.

The indications to me (I am an experienced h/w and s/w design
engineer) is that nobody has done any analysis on the failures, and
that the failed units are simply repaired and recycled into the
production process, or perhaps into "warranty exchange" units, as is
normal in the avionics business.

In avionics, one tends to get items which are often several years old,
when getting a warranty exchange unit. I've had a 10 year old KI229
RMI supplied in this way.

 

[Peter]

[email protected] wrote

Try
http://www.rockwellcollins.com/products/cs/br/flight_controls/index.html
maybe?

Well, yes, I am sure if I went through the massive paperwork exercise
and installed something like that, or S-TEC, it would work just fine.

It has to be said that the KFC225 does a super job of controlling the
aircraft, even in pretty bad turbulence. It just keeps packing up. I
am on my 3rd main unit, too (the built-in altimeter and pitch gyro are
prone to packing up)

 

[Mark]

H are looking at it; I have no idea how far up the organisation any of
this has gone.


By "IT" do you mean they are repairing your one particular unit or are
they looking at the more general issue?

I would contact QC or engineering, not just the warrantee repair
people.

Mark

 

[Scott Skylane]

That WAS the PLAN, back a couple years ago when the article was written, but
I recall (for what that's worth) that the last one was retired (for carrier
operations?) and the advanced Hornet was accelerated into operations. Wasn't
that a thread here just a few weeks ago?

Matt,

See: http://makeashorterlink.com/?Q36E21BDB

It's the Google group archive of J. Honecks thread. His first post has
a link to an article that indicates the last F-14 squadron was just
deployed to the Gulf.

Happy Flying!
Scott Skylane

 

[Matt Barrow]

Matt,

See: http://makeashorterlink.com/?Q36E21BDB

It's the Google group archive of J. Honecks thread. His first post has a
link to an article that indicates the last F-14 squadron was just deployed
to the Gulf.

I read Jay's post of "What an amazing run the F-14 has had. It will be
terribly sad not to see
them flying anymore.... ", but not the article. :~(

"The venerable F-14 Tomcat, a mainstay of naval aviation, Hollywood movies
and air show awe since the 1970s, will retire for good after a final combat
deployment, which began Thursday. "
....
"The Tomcats require between 40 and 60 hours of maintenance for every hour
in the air, LaBranche said. For the F/A-18s, it's more like 10 to 15."

 

[Peter]

By "IT" do you mean they are repairing your one particular unit or are
they looking at the more general issue?

Honeywell won't get involved as a company; here in the UK they work
only via avionics dealers. I have to fly the plane to one of these and
they will change the servo (plus the main unit which is on its 4th
failure; it randomly changes the pilot-selected VS value during a VS
climb or descent...) and what if anything Honeywell do with what comes
back to them I have no idea.

Based on past experience I doubt anybody at Honeywell will look at the
faulty parts to see what the reason for the fault might have been. If
they do, they sure keep it very quiet (and nothing gets done about it
in the way of modifications).

I would contact QC or engineering, not just the warrantee repair
people.

If you have a contact, I would appreciate it! My email is at the
bottom.


Peter.

 

[Joerg]

Hello Peter,

The indications to me (I am an experienced h/w and s/w design
engineer) is that nobody has done any analysis on the failures, and
that the failed units are simply repaired and recycled into the
production process, or perhaps into "warranty exchange" units, as is
normal in the avionics business.

If they really did that they would be racking up a huge liability
portfolio. In the world of med electronics the FDA would pretty quickly
shut down everything from production to sales, followed by a "root
canal" type audit of the QC process.

If you have the impression that the problem hasn't percolated to upper
management you could help that process along by a letter to the CEO.
Sometimes that does wonders.

In avionics, one tends to get items which are often several years old,
when getting a warranty exchange unit. I've had a 10 year old KI229
RMI supplied in this way.

That may be ok if it's a full refurb like it is done on engines, IOW
where you get a zero-hours paper with it. But from an ethics point of
view they should tell you if it's a refurb.

I wonder if there is some kind of lemon law for those parts. If you buy
a car in California and repair attempts failed x many times you can
demand a full refund.

Other than that, could there be spikes caused by either the servo motors
or by some other gear that fry something?

Regards, Joerg

 

[[email protected]]

[email protected] wrote

Well, yes, I am sure if I went through the massive paperwork exercise
and installed something like that, or S-TEC, it would work just fine.

It might be a silly amount of paperwork, but then you'd more than likely
have a reliable autopilot. As it stands, you are getting warranty
replacements, but there is a time and hassle factor of getting them
swapped out. Somewhere there is a break-even point.

It has to be said that the KFC225 does a super job of controlling the
aircraft, even in pretty bad turbulence. It just keeps packing up. I
am on my 3rd main unit, too (the built-in altimeter and pitch gyro are
prone to packing up)

Have you checked out the rest of the electrical system on the aircraft?
Maybe there is some other problem that is burning this stuff out. This
might even involve hooking up some kind of data logger and flying around
for a while; some spikes might not show up in ground testing or on short
flights.

Matt Roberds

 

[[email protected]]

Hmmm...RC try to be customer-responsive these days, but I CAN remember
when they upgraded the FMS 800 and the roll command (used to steer to
the desired track), was reversed from its previous polarity - made for
interesting waypoint/turn capturing!

I based my suggestion on a few different observations. I used to work
at a flight simulator company and it was a lot easier to make the
Collins radios work right than the King ones. At that job, and since
then, I have also noticed that the "cheap" avionics package often
involves King radios, while the "upgrade" packages involve Collins
radios. Finally, I went on a job interview with Collins in 1996 and was
totally impressed by the amount of shiny! stuff at the plant in Cedar
Rapids. (Got an offer, too, but chose another offer in Dallas over
Cedar Rapids. Probably should have picked Cedar Rapids.)

Matt Roberds

 

[Brian Whatcott]

///. I used to work
at a flight simulator company and it was a lot easier to make the
Collins radios work right than the King ones. ///
Matt Roberds

Hmmm.... radios are one of the items where simulators invariably use a
front panel mock-up with switches and pots signaling the host.
They are not designed or provided by the avionics makers.
Making a mock-up work doesn't have much relation to the
avionics OEM, seems to me??

Brian W

 

[Peter]

That may be ok if it's a full refurb like it is done on engines, IOW
where you get a zero-hours paper with it. But from an ethics point of
view they should tell you if it's a refurb.

It's a refurb; the weakness is in the process that is used on the way
there.

A unit comes back with a "defective" tag on it. It is going to go to
the person who originally designed it, and be checked out with lots of
instrumentation? Of course not. It will go back into the factory test
process, and if it passes the factory test then it is classed as
working.

If there is an obvious fault that will get fixed, and statistically
that will likely be the only fault, so that is OK. The problem is with
faults that are intermittent, or faults that don't get picked up by
the factory test. I've been in electronics design/mfg for 25+ years so
know this problem well; in my business we scrap anything that comes
back, just to make sure.

My warranty avionics bill totals something like US$100,000 and that is
mostly Honeywell avionics, added up at list prices, over 2 years.

 

[Joerg]

Hello Peter,

A unit comes back with a "defective" tag on it. It is going to go to
the person who originally designed it, and be checked out with lots of
instrumentation? Of course not. It will go back into the factory test
process, and if it passes the factory test then it is classed as
working.

That would be a normal process.

If there is an obvious fault that will get fixed, and statistically
that will likely be the only fault, so that is OK. The problem is with
faults that are intermittent, or faults that don't get picked up by
the factory test. I've been in electronics design/mfg for 25+ years so
know this problem well; in my business we scrap anything that comes
back, just to make sure.

If intermittent is flagged this should cause more diligence in the
repair process. If the test folks can't duplicate the error there should
be further investigation, at least in cases that involve a lot more than
one unit. That's where the QC system should kick in with its database
information. Or in the med biz, that is where the QC system has to come
in on a mandatory basis.

My warranty avionics bill totals something like US$100,000 and that is
mostly Honeywell avionics, added up at list prices, over 2 years.

I wonder how they survive. I ran a business for several years and
warranty overhead was factored in on a "per product" basis. My boss
would probably have fired me with cause if I had ever failed to see an
epidemic trend in one of the lines we offered. It never happened but we
watched that stuff like hawks.

On the part of the user I'd be mighty concerned about what happens after
the warranty period ends. After that, every time the servo hangs you'd
hear that slurping sound coming from your bank account.

Regards, Joerg

 

[Peter]

If intermittent is flagged this should cause more diligence in the
repair process. If the test folks can't duplicate the error there should
be further investigation

Absolutely right! But you are assuming that there is a process for
feeding back the info from the original pilot. In this case, the unit
works its way back through the Honeywell return chain. It's now been 3
years from my earliest roll servo failure, and there were plenty
before me, on the KS270C servo. They just keep packing up....

I wonder how they survive. I ran a business for several years and
warranty overhead was factored in on a "per product" basis. My boss
would probably have fired me with cause if I had ever failed to see an
epidemic trend in one of the lines we offered. It never happened but we
watched that stuff like hawks.

I guess they survive because in a piece of avionics listing at $2000
the materials come to $100-$200. So the actual cost of supply of a
replacement is just the latter figure. This ratio would be true for
all avionics I've seen - right up to any IFR GPS.

Also all the failed stuff gets recycled. The bit that costs real money
is the metalwork. The electronics cost peanuts, around $10 for a $2000
servo. I've seen what's in there - $10 max. The metalwork doesn't go
wrong and gets re-used.

On the part of the user I'd be mighty concerned about what happens after
the warranty period ends. After that, every time the servo hangs you'd
hear that slurping sound coming from your bank account.

My warranty ended over 1 year ago. Honeywell offered me an indefinite
extension on the KFC225, all the time it keeps packing up. I also
purchased their extended warranty on all my other Honeywell kit -
$4000 for two extra years. I got that back within 6 months when the
KI229 RMI packed up (its 3rd failure) - that lists here in the UK at
about US$7000.

Most serious owners go for litigation after that. The trouble is that
litigation would not fix the electronics.

 

[Joerg]

Hello Peter,

Absolutely right! But you are assuming that there is a process for
feeding back the info from the original pilot. ...

Quite scary. In medical, if there isn't a process for that the FDA or
another agency shuts the place down.

I guess they survive because in a piece of avionics listing at $2000
the materials come to $100-$200. So the actual cost of supply of a
replacement is just the latter figure. This ratio would be true for
all avionics I've seen - right up to any IFR GPS.

I am not a pilot. My guess is that a roll servo failure usually doesn't
lead to a crisis situation but I don't know. If that happens on gear
that can cause grief the liability can be huge. $200 won't even pay for
a single attorney hour.

My warranty ended over 1 year ago. Honeywell offered me an indefinite
extension on the KFC225, all the time it keeps packing up. I also
purchased their extended warranty on all my other Honeywell kit -
$4000 for two extra years. I got that back within 6 months when the
KI229 RMI packed up (its 3rd failure) - that lists here in the UK at
about US$7000.

They really need to have someone look into that. Whenever I do that for
a client they usually break even on my fees within a month or so, just
because of reduced field returns. Others shy away from hiring a
consultant or other experts and try to hang on. Long term that won't work.

Just imagine if I would realize my dream of becoming a pilot. I'd
probably rent because of the cost but what if I didn't and needed an
autopilot? After reading your story here would I buy a Honeywell
autopilot? Probably not. That's the hidden cost of sub-par designs.

Regards, Joerg

 

[[email protected]]

Hmmm.... radios are one of the items where simulators invariably use a
front panel mock-up with switches and pots signaling the host.

Not always. Many of the radios I worked with (business jet and up) were
the kind where there was just a control head in the panel and the actual
guts (tuner, etc) were someplace else. We used the real control heads
and just dispensed with the remote guts. For the Collins radios, you
could read the spec, set up the control and data lines to the control
head, and get it right fairly quickly. For the King radios, reading the
spec was only the first step of a long journey into undocumented
goofiness. This pattern repeated itself with other hardware where we
used the actual aircraft item complete, such as a FMS.

Matt Roberds

 

[Peter]

Just imagine if I would realize my dream of becoming a pilot. I'd
probably rent because of the cost but what if I didn't and needed an
autopilot? After reading your story here would I buy a Honeywell
autopilot? Probably not. That's the hidden cost of sub-par designs.

You may find that the manufacturer of whatever you buy has already
made a similar decision