Earlier last week I was planning on making a trip to the furthest
airport in my phase one fly off area. The air was cool and crisp, clear skies
and very light winds …. a perfect day to begin working on seriously attacking the
flight test cards.
After getting the oil temperature up and performing the preflight
run-up, took to the active runway and decided on a 65 – 70 knot climb out. The
engine was running good down the runway and developed over 5200 RPM as the
climb out began. At about 250 feet off the deck I caught a whiff of fuel and a
moment later there was a partial loss of power. I had lost approximately 1,200
RPM … I quickly throttled back some then advanced the throttle a little but
the engine seemed “flat” so it was throttled to keep the engine around 4300 RPM
where the engine was running OK and seemingly smooth , decided it best to abort
the flight so called the tower for clearance to return to terra firma to
conduct a “little FBI Double-O-Seven-type investigatin” to coin a lyric from a
Jr. John song.
One of the really nice features of the Dynon SkyView is there is a built
in flight data recorder that constantly records a plethora of parameters … input from all of engine and fuel sensors,
GPS, transponder, autopilot, airspeed, altitude, attitude, ect., ect. By
default, the sampling rate is 16 times per second (this can be changed in the
setup menus) which provides approximately 2 hours of very detailed data
capture. If the sampling rate is lowered, then much more than 2 hours of
storage can be obtained … as an example, a sample every 10 seconds will result
in 150 hours of flight data storage.
Decided it best to make a data dump of the SkyView's log to a flash drive and try to analyze what happened ... this is done by saving the SkyView's data logs onto a flash drive plugged into the USB port.
But before delving into that, I should mention the USB port under the
instrument panel base connects to one of the USB ports on the back of the Dynon SkyView. The under the panel USB port is used for a plethora of tasks such as upgrading the SkyView’s firmware and allowing for the monthly
updating of navigational data bases. In addition, third party chart software can
also be utilized by the SkyView to display FAA sectionals, IFR approach plates and
detailed airport diagrams called “Safe Taxi”. Plus, the USB port allows flight plans to be
loaded into the SkyView or saved from the Skyview along with allowing the pilot
to make backups of all the SkyViews internal settings and the biggie for me at the moment,
exporting of the SkyView’s user data logs. All of this can coexist on one flash drive ... it just requires plentiful storage space on the flash drive.
Some of the above features such as FAA sectionals, IFR charts and “Safe Taxi”
airport diagrams require leaving a flash drive in the under the panel USB port so the SkyView
can access the data as needed. This can present a problem, because a typical flash drive
would extend down from the bottom of the instrument panel and could very
easily be snapped off from contact with the knees or an arm. So one of the many niceties the
DOG Aviation procurement department has recently purchased consists of two tiny
PNY 32 Gig Elite-X USB flash drives. They are small in stature (yet big on available storage space) … they are just long enough for the fingers
to get on them to pull them out of the USB port. At 32 Gig, there is plenty of
storage for the software revisions, monthly updates, third party charts, flight plans, saving internal
settings and the biggie at the moment, saving the user data log recordings … with plenty of room left over. Two of
the flash drives were purchased so one flash drive can be at the house to get updated
with the latest maps, ect, then taken to the airport and swapped for the one in
As one can see in the above photo, compared to a rivet, the PNY 32 Gig Elite-X
Fit flash drive is tiny and will not be subject to breakage from the knees
because it barely extends beyond the USB port.
OK, back to the user data log dump … after making a safe landing and taxing
back to the hangar, the flight data logs were saved onto one of the PNY flash
drives to take home for analysis. Saving the user data logs to a flash drive plugged into the USB port is accomplished by pressing and holding
buttons 7 & 8 on the SkyView to enter into the Setup Menu. The first item
in the Setup Menu list is System Software … selecting this will bring the user
to a sub-menu that has an entry of Export User Data Logs. Selecting Export User
Data Logs will save the data logs onto the flash drive plugged into the USB
port on the bottom of the RV-12’s instrument panel. The files are saved as a .csv
file which is a computer file standard meaning … comma separated values. The
Microsoft Excel spread sheet is capable of opening and displaying the contents
of this type of file.
A tip for those builders using Excel to view the SkyView’s data log: Do yourself a favor, after opening the data log file with Excel, click on the View tab
on the top of the window and click on Freeze Panes … then select Freeze Top
Row. What this does for you is freeze the top row which is where the name for
each column is located … such as RPM, EGTL, EGTR, Fuel Pressure, Fuel Flow, ect.
Freezing the top row allows the spreadsheet to be scrolled to where the event
being analyzed takes place without losing the name for what the column
represents. The user has to analyze what the numbers mean and form a picture in
“your mind’s eye” as to how the numbers correlate with one another. So a lot of
back and forth scrolling is needed … but there is a better way.
Getting high tech … deciphering the SkyView data log can be taken to the
next level by using a service provided by a company named Savvy Analysis. Savvy
Analysis offers a service for pilots that allows the Dynon SkyView’s user data log
to be uploaded and converted into a graph chart which is visually much more
meaningful …. various parameters can be selected by the user to be charted simultaneously, very cool!! And
best of all, the service is free … unless you want them to analyze the data log
results for you. Below is a copy of the Savvy Analysis chart showing the “event”…
once I had the chart on the computer screen, I converted it into a .jpg file to
post here on the Blog.
Uploading the Dynon SkyView’s data log to Savvy Analysis for conversion
into a chart will yield an easy to look at chart of the event… which greatly
aids in analysis of the event. Here the sudden loss of RPM can be seen
accompanied with a cooling of the left cylinder’s exhaust gas temperature (EGT)
while the right EGT remain unchanged. Looks like the carburetor for the left
side went rich, probably due to the carburetor’s float needle valve getting
stuck on a piece of debris.
As can be seen in the above chart, the Rotax 912ULS is turning out a
steady 5,200 RPM during the climb out then instantly drops to roughly 4,000 RPM
the exhaust temperature on the left side of the engine also takes a drop
suggesting there was a rich mixture coming from the left carburetor. Also of
note, at the time of the “event” the fuel pressure dropped from roughly 5.5 PSI
to around 5 PSI which, to my way of thinking, would correlate with a stuck
needle valve in the left carb reducing the fuel pressure because of the
momentary excess flow of fuel. It is not a fuel pump issue because the minimum
pressure for the Rotax 912ULS engine is 2.2 PSI so that is not an issue.
There are only a few things that would contribute to a carburetor going
rich … excess fuel pressure (not the case here) choke being applied (it was
full off) which leaves … floats sinking in the bowl, debris preventing the
needle valve from closing, bad needle valve or float level adjustment being off
… all of which require removing the bowls from the carburetors. So the bowls on
both carburetors were removed for inspection. All the floats appeared to be
floating nicely. Just to be sure, each float was weighed at 3 grams and each
pair weighed 6 grams … so they are spot on weight wise. The left carb bowl (suspected
problem one) had just one little speck of debris in it. I decided to turn on
the master switch and use the electric fuel pump to flush out the float needle
valves while moving them up and down with the float bracket. Next a check was
made to see if the needle valves would stop fuel from flowing when closed … all
seemed well. Also checked the arms on the float bracket which should parallel
the body of the carburetor when in the closed position … both float brackets
appeared to be correct. Interestingly, the right carburetor had much more
debris in the bowl than the left carb did and also had a very small flake of
what appeared to be a flat black paint or coating that appeared to come from
the float brackets. This got me off on a short tangent thinking perhaps the
wiring for the EGT probes got switched or the probes were attached to the wrong
port on the engine monitor… I used a heat gun on the right exhaust pipe near
the EGT probe to verify the Dynon SkyView data log file recorded the heat on
the correct sensor … fortunately, it did so the “event” was a left carburetor
event for sure.
After the carburetor bowls were placed back onto the carburetors,
decided to make a test flight with a shallower climb out of 80 knots. The test
flight reveled no engine problems. Engine RPM remained constant around 5,250 on
the climb out until I pulled back on the throttle … so for the next few
flights, will plan on staying in the traffic pattern for a handful of takeoffs
and landings so I can incrementally steepen the climb outs to make sure the
issue has been resolved. Below is a couple of videos of the takeoff and landing
after working on the carburetors. The videos are heavily edited to meet the
under 100 Meg requirements of Blogger. Unfortunately, the camera switched files
just as I landed so there is a little skip in the video at that point as I
landed with a slight thump. Also will have to make improvements on mounting location
for Mike’s camera to get it a little sturdier so the camera does not rattle.
A takeoff video made after working on the carburetors all seemed to go
well … had a steady 5,250 RPM on climb out although it was a shallower climb out
than when I had the “event”.
During the landing, I kept the RV-12 purposely high until on final then
slipped a bit to lose the excess altitude which can be seen during the period
of time when the horizon looks canted while on final to the runway.
I would have flown a few more loops around the pattern but needed to get
on the ground to install a winter heating system that requires temperatures
around 70 degrees to cure the epoxy properly … and there are only two warm days
left before the winter plunge. More on the heating system in the next post.