14 November 2018

Maintenance: First annual inspection

With 87.5 hours on the airframe following eleven months from the first flight, the first annual inspection was completed over the course of four days.  For the occasion I created a detailed checklist (using Tim Olson's N14YT checklist as a starting point).  My inspection checklist can be found below and on my page with my POH, flight checklist and flight test cards.  Be sure to customize the checklist to your own aircraft, should you choose to use it.

.pdf .odt .doc
Annual Checklist  v1.0 v1.0 v1.0

There were no findings of particular concern.   Here are the findings of interest:
  • The oil filler neck was loose.  The safety wire prevented it from working completely out.
  • The fuel injectors showed stains from leaking fuel.  An investigation into the cause revealed that this is normal because the injectors take in ambient air to atomize the fuel (if air can go in, it can go out).  Image of injector #2.  Fuel stains visible.

  • The FF-00076 Sniffle Line Bracket continues to break (as first identified during an oil change).  A closer inspection shows that it simply melted (pure Al melts at 1,221 °F which is below the 1,500+ °F EGT) as shown in right image.

    I replaced it with a steel bracket which I fabricated from 4130 stock.

  • Brake pads appeared okay at 87.5 hours and 89 landings.

  • Apparently I neglected to grease the nosewheel bearings when I installed the wheels back in Feb-16! However, remarkably there was no indication of wear of any kind.  I'll take that as evidence of excellent technique in keeping the nosewheel up as long as possible on takeoff and landing. :)
  • The nosewheel pull out force was at 12 pounds (per page 40A-07, it is supposed to be 26 pounds, as I set it back in Feb-16).  I have never experienced shimmy.  Upon greasing the nose wheel fork, I could only get either 17 or 42 pounds of breakout force (due to cotter pin alignment).  I settled for 17 given that I had no shimmy at 12.  I don't recall, nor did I document, which grease I originally used during construction.   For this round, I used Aeroshell #6.  For the next round, I'll use a thicker grease.
  • The spark plugs looked reasonable for the hours imposed on them.

  • Added Spi Wrap to the elevator trim wire in the empennage and coming out of the trim motor.

  • The cylinders indicated the honing is still intact, so they are not glazed.  Cylinder #3 shown below.

  • Timing remains at 25° though the engine data plate specifies 20°.

Maintenance: SB 18-09-17 Aft bottom skin cracking

Van's released SB 18-09-17 to address some cases of the aft fuse bottom skin cracking (examples of the cracks in flying aircraft can be found here and here). Following inspection of my airframe, I found no evidence of cracks.  However, I chose to install the stiffeners as described in the service bulletin.  The process took less than an hour with two people: One in the empennage and other underneath.
Here are the parts associated with the service bulletin.

Because the F-14148A Skin Doubler (the diamond) is affixed to the exterior of the dimpled F-01478 Bottom Skin, the dimples would remain unfilled and preserve voids under the six rivets which attach the Skin Doubler. So I chose to create fillers for the dimples by using drilled out rivet heads for the purpose.  Thus, I drilled out about 25 rivets on scrap to make sure I could obtain 6 rivets with well-centered holes.  I then sanded them smooth.  These six fillers were cleco'd into place on the Skin Doubler prior to cleco'ing on the skin (after all, we're working upside down here!), as shown below.

Removing the F-01406F Baggage Bulkhead Corrugration, placing boards and towels in the baggage area and a blanket in the tail, it was trivial to access the area required to install the stiffeners. 

A 12 inch #30 drill bit was used to match drill the bottom holes in each of the F-14147 Stiffener Clips.  There was adequate space available to squeeze the AN470AD4-4 universal rivets with my longeron yoke.   Tungsten bucking bars were used to buck the AN426AD3-x rivets.  It's worth noting that I increased the length of each AD3 size rivets by 0.5 as compared to what the SB calls out.

Following drilling through the F-01408-L/R Side Frames to attach the F-14147 Stiffener Clips.

Shop heads of the rivets, on the aft side of Side Frames, holding the Stiffener Clips.

Forward side of the Side Frames showing forward interior stiffeners in place.

Aft side of Side Frames showing single aft interior stiffener in place.

Exterior views of diamond Skin Doubler.  I need to clean up the area with acetone as I chose to prime the associated surfaces.

13 November 2018

Avionics: Minor change in electrical bus topology

Minor change in my electrical bus topology.  First, a functional explanation of my three buses is in order:

  • Master:  Critical devices for flight such as lights, electric fuel pump, flaps, trim, ELT, alternator field that can also remain powered during engine start.  This bus should always remain on in flight.
  • Avionics 1:  Critical flight instruments such as left side EFIS, AHRS, GTR-200 transceiver, transponder, backup AI and low fuel level indicator.   This bus can be shut off only in VFR flight.
  • Avionics 2: Less critical instruments and devices to include right side EFIS, GTN-650 navigator, capactive fuel indicator, autopilot, GMA-245 intercom and ship's power LEMO.  This bus can be turned off any time to reduce energy demand (it should remain on in IFR to stay legal).
With that out-of-the-way, when I installed the canopy camera power, I used the "unassigned" fuse position that was on the Avionics 1 bus.  However, there is no need for the cameras to be on Avionics 1.  And, back when I designed my bus topology, I decided to put the GMA-240 (now a 245) on Avionics 2.  However, I should have kept the intercom on Avionics 1 so I can keep Avionics 2 off (primarily during taxi) and still hear my passenger.  So I decided to swap the two positions:  Intercom and canopy camera.  This required a lot of contorting, cutting of wire ties and losening of screws.  Here is the back side of my fuse block, which I designed (you can learn more about what I did here).  I had to remove a wire from the left and center rows and swap them.

Below is my new fuse legend.

I wish I had put the LEMO power on Avionics 1 so I can maintain power to the headsets with Avionics 2 off.  It isn't practical to change that now and the LEMO versions of my headsets understandably are not provisioned for batteries.

12 November 2018

Maintenance: Wing jack

First use of my new jack.  I obtained the Alpha Aviation model 326 jack.  It has a 26-43 inch effective range, representing a 17 inch ram travel.   I coupled that with Bogert Aviation's RV jack pads.  Together, they worked great!

The jack:

The jack pad:

In use on an RV-10 (with a different jack):


04 November 2018

Avionics: Installed Kanardia Horis 57

On 12-Sep-18, I received my Kanardia Horis 57.  It is to replace my TruTrak ADI2 on my panel.  The ADI2 is a solid instrument, however I like the additional information provided by the Horis.  It came to $1,040 plus $70 shipping minus $64 for exclusion of GPS antenna (can use the one from my ADI2) and remuneration for potential import fees.

It's a very capable device.  The following are just some examples, rather than a complete list:
  • Attitude 
    • Roll, pitch
  • Airspeed
    • IAS and TAS
  • Vertical speed
  • Altitude
  • OAT
  • Altimeter setting
  • G meter
  • Heading rate (1 minute and 30 second turn)
  • Tracking course
    • True or magnetic (with optional magnetometer)
    • Includes additional separate directional indicator function
  • Wind speed (with optional magnetometer)
I installed it on 4-Nov-18.  I placed the OAT sensor on the right wing on the inboard inspection plate (I didn't want to drill into the wing skin) and ran the cable through the upper aft bushings, under the right seat pan, up the tunnel and over to the instrument.

For reference, here is the OAT sensor pinout diagram.  It uses a 4C4P RJ-10 (also known as RJ-9 and RJ-22) connector.  I had to lop it off and replace it in order to negotiate the cable through the airframe.  I also had to buy a new crimper for the little connector.

Here is the rear of the instrument as installed.  The Horis uses 5 mm ID connections.  So I snagged some 8 mm OD, 5 mm ID nylon tubing and 5/16" to 1/4" couplings (5/16" is 7.9375 mm) to connect to my 1/4" OD pitot static system tubing.

Here it is in flight.  I need to calibrate the altitude as it's about 300 feet off.

03 November 2018

Avionics: Center console USB power socket change

I decided to replace my center console dual USB power plug only because I found this neat one

Not only does it support QC3.0 (3A per port for compatible devices) and has a snazzy red anodized aluminum body (to match my red accented interior), it also has a cool voltmeter on it.  So far, it has not caused hash on the GTR-200 transceiver like my original USB power socket did.

Here it is installed in the airfame.  It did require me to drill out the hole in the F-14123 Power Outlet Bracket to 1.125".  That hole is now round and no longer has the flat "locking" sides.

29 September 2018

Avionics: Adding a lot more voice alerts on GRT EFISs!

GRT EFIS Audio Alerts

Update 16-Apr-19:  Added alarm #288.

Executive summary: At least 80 undocumented audio alerts are available on most GRT EFISs. This post explains what they are and how to enable most of them.


The GRT HXr, HX, Sport and Horizon EX, Sport SX, Mini-GA and (I think) Mini Series EFISs have an audio output to provide audio alerts to your intercom using audio files that you must provide.  The output is found on pin B24 with ground on B25 for all but the Minis.  GRT has disclosed only 17 special alerts which use predefined filenames (shown in parentheses below) which are typically copied on to the EFIS from a USB drive using the Set Menu on the EFIS:
  • Minimum oil pressure (oil_pressure.wav)
  • Minimum fuel pressure (fuel_pressure.wav)
  • AoA tone near stall, repeats more frequently closer to stall (filename unknown)
  • Obstacle alarm (obstacle.wav)
  • New traffic within alert criteria (traffic.wav)
  • Autopilot disconnect 
    • Manual (ap_disconnect_advisory.wav)
    • Automatic(ap_disconnect_warning.wav)
  • Maximum oil temperature (oil_temperature)
  • Maximum CHT (cht.wav)
  • Low/High voltage (low_volts.wav and high_volts.wav)
  • Maximum fuel flow (fuel_flow.wav)
  • Vne exceeded (speed.wav)
  • Trim alert from GRT pitch servo, repeats every 30 seconds (trim.wav)
  • SAP height above runway at 500, 200, and 100 feet (500.wav, 200.wav,1500.wav)
  • Minimums if decision altitude is set (minimums.wav)
  • Approaching minimums when 200 feet above decision altitude (approaching_minimums.wav)
  • Waypoint sequence (waypoint.wav)
However, I recently became aware of this post on GRT's forums which explains that "every unique alarm can have its own audio file using an internal alarm number".  The format that GRT uses for those files is alarm####.wav where #### is the four digit number of the alarm.  Unfortunately, GRT did not originally provide documentation of the alarm numbers since releasing the initial firmware update for audio support in 2015 and to date, still has not.   But, that information lead me to identifying nearly 60 undocumented alarms.

Below I...
  1. Explain how the EFIS manages audio alerts.
  2. Explain how to add the additional alerts.
  3. Show the additional undocumented alerts I've identified thus far
  4. Provide audio files for you to use.
  5. Explain the process I used to determine the undocumented alarms.

GRT EFIS Alert Process

First, it's helpful to understand how the EFIS alert annunciation process works.

Short version:
  1. Alarm occurs.
  2. EFIS looks for associated file.
  3. EFIS plays the file.
Long version:
  1. Alarm occurs with unique internal four-digit alarm number..
  2. EFIS looks for the associated alarm####.wav or predefined file in the Alerts folder first on the USB drive, then on the EFIS itself.  
  3. If the file is found, it is played.  
    • If there are duplicate files then the USB file is played first, immediately followed by the EFIS file resulting in a double alert. 
      • This would occur with the predefined special alarms like obstacle.wav which is also alarm0091.wav.
    • To avoid double alerts, don't provide both files.
      • For example you may have already placed obstacle.wav on your EFIS, so don't provide alarm0091.wav also unless you want to hear both wav files:  "obstacle, obstacle".
How to Add Your Own Voiced Alerts

You can find the undocumented alarms in the table shown below in the section "Alarms I Have Identified".  Here are the steps to expand your voiced alarms on your EFIS:
  1. Decide what message you want your EFIS to voice for each alarm you're interested in having.
  2. Create the voices using an online text-to-voice service.  
  3. Save each message with the proper filename (alarm####.wav - must use four digits).
  4. Copy your new alarm####.wav files onto your USB drive into a folder named Alerts.
  5. Plug the USB drive into your EFIS. 
  6. You're done.  Verify it works on the ground with engine off, prior to flight.
    • I don't suggest copying the files to your EFIS since I'm not clear on how to delete them without wiping all settings on the EFIS.
  7. Consider carrying the full set of alarm files to help identify additional alarms as explained in the "Files You Can Use for Alerts" section below.
Additional items to consider:
  • If you decide to copy your alarm files to your EFIS, your collection of selected audio files should not exceed 100 MB in size as the EFIS will not copy more than that.
  • Keep in mind that the EFIS will play the entire file when the alarm triggers, so very terse messages are probably best.
    • For example, "hey, pilot in command, your cylinder head temperature on cylinder 2 is too high, do something about it!" is not a wise choice for alarm 76.  
  • Not all alarms get a preceding tone (a "blip" or "whoop").  If you want your alarms to have a preceding tone, you may need to add the tone to the wav file you plan to use.
  • My AUX inputs probably don't match yours.  Be sure to use messages that match your configuration.
  • I make mistakes.
Alarms I Have Identified

If you find additional alarms or errors, contact me.

  • Alarm:  Alarm number (remember, the number in alarm####.wav needs to be four digits).
  • Message:  My suggested voice message for the alarm.  You can use something different if you prefer.  
    • Note:  Your AUX inputs are probably different than mine so you will need to change the associated messages to match your configuration.
  • Description:  What the alarm represents.
  • Predefined:  If it is also a predefined alarm, the associated special filename. 
Alarm Message Used Description Predefined?
0 Avionics data, unavailable. Serial and EthLink no data
2 Flight time limit Flight time limit
3 Interval elapsed Interval alarm
5 Oil pressure high. Max Oil Pressure oil_pressure.wav
6 Oil pressure, low. Min Oil Pressure oil_pressure.wav
8 Oil temperature high. Max Oil Temp
9 Oil temperature, low. Min Oil Temp
10 RPM high RPM High
11 RPM low RPM Low
12 Total fuel low. Min Fuel
13 Manifold pressure high. EIS1 Max AUX1
14 Left fuel tank high. EIS1 Max AUX2
15 Right fuel tank high. EIS1 Max AUX3
16 Alternator current high. EIS1 Max AUX4
17 Battery current high. EIS1 Max AUX5
18 Fuel pressure high. EIS1 Max AUX6
19 Manifold pressure, low. EIS1 Min AUX1
20 Left fuel tank, low. EIS1 Min AUX2
21 Right fuel tank ,low. EIS1 Min AUX3
22 Alternator current, low. EIS1 Min AUX4
23 Battery current, low. EIS1 Min AUX5
24 Fuel pressure, low. EIS1 Min AUX6
25 Coolant temperature high Max Coolant Temp
26 Coolant temperature low Min Coolant Temp
27 EIS voltage high Max EIS Voltage
28 EIS voltage low Min EIS Voltage
29 EFIS bus one voltage high Max EFIS Voltage Bus 1 high_volts.wav
30 EFIS bus one voltage low Min EFIS Voltage Bus 1 low_volts.wav
31 EFIS bus two voltage high Max EFIS Voltage Bus 2
32 EFIS bus two voltage low Min EFIS Voltage Bus 2
33 EFIS bus three voltage high Max EFIS Voltage Bus 3
34 EFIS bus three voltage low Min EFIS Voltage Bus 3
36 Exhaust 1 temperature high. Max EGT1
37 Exhaust 2 temperature high. Max EGT2
38 Exhaust 3 temperature high. Max EGT3
39 Exhaust 4 temperature high. Max EGT4
40 Exhaust 1 temperature, low. Min EGT1
41 Exhaust 2 temperature, low. Min EGT2
42 Exhaust 3 temperature, low. Min EGT3
43 Exhaust 4 temperature, low. Min EGT4
72 EGT span high EGT Span Too Large
75 Cylinder 1 temperature high. Max CHT1
76 Cylinder 2 temperature high. Max CHT2
77 Cylinder 3 temperature high. Max CHT3
78 Cylinder 4 temperature high. Max CHT4
79 Cylinder 1 temperature, low. Min CHT1
80 Cylinder 2 temperature, low. Min CHT2
81 Cylinder 3 temperature, low. Min CHT3
82 Cylinder 4 temperature, low. Min CHT4
85 Cylinder 1 cooling fast EIS1 CHT1 Cooling Too Fast
86 Cylinder 2 cooling fast EIS1 CHT2 Cooling Too Fast
87 Cylinder 3 cooling fast EIS1 CHT3 Cooling Too Fast
88 Cylinder 4 cooling fast EIS1 CHT4 Cooling Too Fast
91 Obstacle Obstacle obstacle.wav
92 Altitude deviation Check Altitude
121 CO guardian unit failure CO Guardian Unit Failure
122 CO level unsafe Carbon Monoxide Level Unsafe
123 Cabin pressure unsafe Cabin Pressure Unsafe
124 Check bio data Life Guardian periodic bio data check
288 Roll trim required Roll trim required
302 RPM range limit TachHi/Lo alarm

*Note:  "Traffic", "Speed", "Autopilot disconnect", G maximum limits and the "AoA" tone do not appear to have an associated alarm number.

Files You Can Use for Alerts

Here are the files with voiced alarms in a 6.2 MB zip file (let's call these the messaged alarms) and the files for the predefined alerts in a 1.7 MB zip file that I use, if you wish to use them too.

How You Can Help Identify Additional Alarms

I'm also giving this link which has the 10,000 alarm files that I assembled in a 318 MB 7z file that just voices the number of the associated alarm (let's call these the numbered alarms).  For example, alarm1293.wav simply says "one thousand two hundred and ninety three".  To decompress a 7z file you will need to use 7zip (for Windows) or Keka (for Mac).

Why do I provide those files?  If you use them, when your system triggers an alarm I haven't documented yet (e.g., I don't have three props, a carb, TIT sensor, etc., so I don't plan to find what their alarm numbers are), you'll hear the number and you can contact me with the information so I can add it to the list. 

To use these numbered alarms and still keep your messaged alarms:
  1. Copy the 10,000 numbered alarms onto the USB drive's Alerts folder.
  2. Copy the messaged alarm files.
    • Overwrite the associated numbered alarms from Step 1 so you still get the messages you want instead of hearing numbers.
  3. Delete any file with an associated alarm you're not interested in hearing.
    • For example
      • Maybe you don't want to hear when your RPM gets too low so delete alarm0011.wav.  
      • You already copied obstacle.wav to your EFIS so you don't need alarm0091.wav.
  4. Plug the USB drive into your EFIS. 
  5. You're done.  If you find additional alarms, please contact me.
The Process I Used To Determine Alarm Numbers

Short version:
  1. Created 10,000 wav files, each which voices the numbers 0-9,999.
  2. Placed them in the Alerts folder on a USB drive connected to the EFIS.
  3. Artificially caused as many alarms as I could.
  4. Wrote down the number I heard for each alarm.
  5. Noted duplicates (i.e., when predefined alarms were followed by a number too).
  6. Created alarm files with the messages I want to hear.
  7. Updated numbered alarm files with messaged alarm files.
  8. Keep numbered alarm files on USB drive to facilitate identification of alerts in the future.
Long version: 

Armed with the file format, it was trivial to decode most of the alarms that were relevant to my installation and setup.  All I needed to do was create 10,000 audio files that announce numbers, then each alarm would annunciate as a number and I could simply write down what each corresponded to.

So, to save myself the time of manually creating one thousand wav files I downloaded pre-record number files (0-99, x00, x000).  These represented only 118 of the files I need.  I needed the remaining 9,882 audio files.  But with those 118 files, I could use ffmpeg and some simple basic DOS scripts that I quickly wrote to automatically create and assemble files "alarm0000.wav" through "alarm9999.wav" giving me the complete collection of 10,000 files, each of which annunciates the associated number.  I further used ffmpeg to ensure each had the file format Windows PCM, 44,100 Hz, mono, 16-bit (as that's what I used for the voice alerts defined above, so I knew that format would work) and was amplified 5 dB to make them louder without too much clipping.

Next, knowing that the EFIS will first look for audio files on the USB drive and then use the internal ones, I created an "Alerts" folder with those 10,000 alarm wav files on a USB drive and connected it to my EFIS.  I then sat in the plane in my hangar and forced as many alarms as I could (by adjusting the criteria in the various settings pages), and as I heard each number in my headset, I wrote down what alarm each corresponded to.  Next, I flew the aircraft and forced the remainder of the alerts.  Then I assembled what I learned in a spreadsheet.

Finally, I went home and decided which alarms I actually wanted voiced, typed out what I wanted for each then had them voiced by ivona.  I then took those voices, dumped them into separate wav files with the correct filenames, as determined by their associated alarm numbers, and placed them on the USB drive that stays with my EFIS (it also has the charts and approach plates).

Whilst I would have preferred GRT's documentation, they gave enough information for me to create it for myself.

12 August 2018

Avionics: Canopy USB power change

I use a GoPro Hero 6 Black mounted inside the canopy.  To power it, I designed a simple board to provide an available 4 Amps from two USB ports.  I mounted that board to the canopy frame.  However, when connected to that power source, the Hero 6 camera will shut off (actually, it crashes and turns back on in band aide mode), without fail, as soon as I taxi into position on the runway.

Originally, I had a Powerwerx dual USB port between the seats.  It powered the Hero 6 without issue.  However, it created a lot of hash on my GTR-200 transceiver

So I replaced it with one from Blue Sea.  This one did not create hash on the GTR-200.  But, the Hero 6 again would shut off at odd times.

After doing a lot of testing at home, I determined that the Hero 6 is extremely picky on which power source it can accept without crashing.  It even won't accept power from my computer's USB port without crashing.  Since the Powerwerx socket originally worked, I decided to replace my USB board with the Powerwerx socket at the canopy frame.  Using some spare aluminum and angle, I made a quick mount for the socket and attached it to a tooling hole on the C-01405-L Aft Canopy Frame with a nylon screw.

Now the Hero 6 stays on the whole flight and there is no hash on the GTR-200.  That may be because the ground wire for the Powerwerx socket goes to the main ground bus, whilst the ground wire for the Blue Sea socket goes to the aircraft chassis.  So perhaps when the Powerwerx socket was originally between the seats, that grounding setup was the cause of the hash.

View from camera in daytime:

In the nighttime: