Thursday, July 23, 2015

The Flight Review

Every couple of years (give or take a few days), I have to take a ride with a certified flight instructor (CFI) in what id called a Biennial Flight Review (BFR), the purpose of which is somewhat murky to me. It would make sense to say to a lapsed pilot that you have to take some level of refresher training, but for someone that has been flying regularly and has managed to stay alive, well, it seems like more of a formality, especially when you consider that it isn't a check ride - there is no pass or fail.

That said, a CFI can refuse to sign off the BFR as complete if you're completely inept, but nothing stops a person truly dedicated to flying himself into an early grave from simply finding a more pliable CFI.

Loath to waste the dollars paid to the CFI to just bore holes in the sky for an hour, I usually pick something to practice and/or learn. This time around I thought it would be a good opportunity to see if the Dynon SkyView and autopilot were capable (as opposed to 'legal') of flying a GPS instrument approach. I chose MadCo, primarily because its very close to home base. The forecast winds were showing runway 27 as the preferred runway choice, so I downloaded the RNAV (GPS) RWY 27 approach plate from the internet. I actually have these approach plates built into the SkyView, but I wanted a chance to get a good look at it before flying it. It has been a very long time since I was instrument current; so long, in fact, that I have never seen or flown a RNAV (GPS) approach.

Here is the approach plate:

The unmolested plate doesn't have the bright red line on it - I drew that on so you could see the course we would be flying.  I'm lazy, so the intersections aren't rounded as they were when the route was depicted on the SkyView.

Note that the SkyView is not approved for this type of usage. That said, while I can't see myself ever screwing up badly enough to ever need to do this for real, it doesn't hurt to be prepared. There are plenty of other stupid things that I've done despite believing that I would never be stupid enough to do them, so.... better safe than sorry, and whether I would ever fly an IFR approach in my plane isn't really the point - just spending time flying with the Foggles is worth the effort.  

Even if I never stray into a big cloud, I may still find myself in a situation where I need to be able to control the plane by reference to the instruments. Have you ever driven into a setting sun on a hazy day?  Yeah, that.

Because the SkyView isn't approved for instrument approaches, it doesn't have the niceties that make the routine use of approaches easy. For example, I could simply load the approach, which would automatically load the waypoints into a flight plan. Instead, I had to manually create a flight plan for the UHAGY → AKPUC → FOFWI → JADMU progression. Even that was more difficult than usual because Dynon seemingly suppresses the automatic completion of a waypoint entry for these types of fixes - I had to enter all five letters. That was easy while sitting on the ground, but would have been much trickier in the air in even moderately bumpy conditions.

I decided to fly the first attempt more or less by hand, and see how well the autopilot could do it after I had my chance at it. I say 'more or less' by hand because I did use the flight director.

"That what??!?"

The flight director is related to the autopilot system. It displays a guide on the artificial horizon, which shows the attitude of the airplane, but does nothing to control the plane. The guide represents a reference of an airplane attitude that will follow the parameters set for the autopilot. The pilot can manually fly the plane directly where the flight director indicates, and by doing so the plane will follow the parameters set for the autopilot.

It looks something like this:

The two purple trapezoidal bars (positioned horizontally in the picture) are the flight director indicators. The yellow bars nestled up underneath them are the indicators for the current bank and pitch angles being flown. When the purple and yellow bars are tightly nestled like that, the plane is oriented exactly the way the autopilot would have done it.

What this would look like to be as I was hand-flying the approach would be, assuming I need to make a left turn to stay on course (or intercept a new course) would be the purple bars rolling into a bank to the left while the yellow bars stayed behind. I would move the control stick to the left which would roll the actual airplane into a left turn. This would cause the yellow bars to roll to the left also, with their angle being commensurate with the bank angle of the airplane. If I were to 'fly' the yellow bars such that they stayed matched up with the purple bars, the airplane would follow the precise path determined by the flight plan.

Letting the autopilot do it would be the easiest way, but if for some reason I wanted to hand fly (too bumpy, or whatever), using the flight director would be the 2nd best choice.

It went well, but the RV-12 is a light plane and the weather wasn't super smooth, so there ended up being a lot more lag between the purple bars telling me what to do and me doing it.

The autopilot logic did a very good job of giving me plenty of notice for upcoming turns. The very tight hairpin turn to the right at UHAGY was the one that I thought might be a bit of a challenge, but in the event it was no steeper or any more difficult than a normal turn, albeit a bit steeper than a standard rate turn. I looked into making a setting in the autopilot configuration screen, but there is no such option. This seems odd as it would be easy for the Dynon to make the required calculations and adjustments to make standard rate turns.

After a not-so-great landing, we went around again and let the autopilot fly the entire approach. I took the Foggles off for that - I wanted to keep an eye on what it was doing. It did a great job. All I had to do was manage the step-downs in altitude, and even that was simply a matter of turning a knob. The landing was no better than the first one, though.

We went on to do a couple of pro forma slow flight maneuvers and such, then it was back to base. I let the CFI make a couple of landings, thinking that his would surely be worse than mine, but it wasn't to be.

Oh well.

It was worth doing, I suppose. I learned a little bit more about the capabilities of the autopilot and the practice time with the Foggles was definitely worth doing, so I have that going for me for the next two years plus a few days.

Sunday, July 12, 2015

Chuff Chuff Chuff

You meet the most interesting people as a pilot, and while it may be personal bias talking, I think you meet even more interesting people when you build your own airplane. Part of this is because you can become something of a mentor to people that bought their airplanes already built, like I did with my RV-6.  One of those people, who we will call The Judge, bought an already built RV-12, had it painted, and now keeps it nearby over at MadCo, which you may remember as my airport-of-choice for buying avgas.

I was helping do a little work on his plane one day when he suggested that we fly somewhere together. I had just the destination in mind: I thought he light like to fly out to Darke Co. to visit with the CEO of Schmetterling Aviation. It took awhile to get it done, what with rainy day after rainy day after... you get the drift. It's been a lousy summer for flying, unless you're a mosquito; they're eating it up!

We decided that The Judge would lead the flight from MadCo to KVES. We had a brief wait at the end of the runway while another plane landed.

Once in the air, I tucked in just behind, below, and to the left.  This worked out well, but as we were cruising along, I was able to take brief peeks at my panel to ensure that nothing was in the process of going awry whilst I wasn't looking. One thing that really got my attention was a traffic indication of an airplane flying from either KVES or Piqua to MadCo. He was right on our course line, albeit being reported as flying 200' below us. That's far too close for comfort, so I asked The Judge to make a 20 degree turn to the left. We flew off the Direct-To course line for a few minutes to get some clearance, then returned to course. As we did. we saw that plane pass in front of us no more than half a mile away.

Thank you, ADSB!

Being the youngster in the group, I rode in the kiddie seat.

As it turns out, this was the day of the Antique Farm Implement show at the county fairgrounds. I like old hardware in general, and tractors in particular.  Still, while walking through the flea market I saw this old propeller for sale.

I would be more inclined to believe this story had I not later seen a beat up old guitar being advertised as John Lennon's first guitar, and carrying a $19,000 price tag.

The "for complete story go to" thing leads to precisely nowhere.

As it turns out, though, it could be true. The 'Heath' mentioned above started a airplane company that became best known for the Heathkit line of electronics and kits:
Their products over the decades have included electronic test equipment, high fidelity home audio equipment, television receivers, amateur radio equipment, robots, electronic ignition conversion modules for early model cars with point style ignitions, and the influential Heath H-8, H-89, and H-11 hobbyist computers, which were sold in kit form for assembly by the purchaser.
Heathkits were influential in shaping two generations of electronic hobbyists. The Heathkit sales premise was that by investing the time to assemble a Heathkit, the purchaser could build something comparable to a factory-built product at a very significantly lower cash cost. During those decades, the premise was basically valid. Commercial factory-built electronic products were constructed from generic, discrete components such as vacuum tubes, tube sockets, capacitors, inductors and resistors, and essentially hand-wired and assembled. The home kit-builder could perform the same assembly tasks himself, and, if careful, to at least the same standard of quality. In the case of their most expensive product, the Thomas electronic organ, building the Heathkit version represented very substantial savings.
The Heath Company was originally founded as an aircraft company in 1912 by Edward Bayard Heath with the purchase of Bates Aeroplane Co, soon renamed to the E.B. Heath Aerial Vehicle Co. Starting in 1926 it sold a light aircraft, the Heath Parasol, in kit form. Heath died during a 1931 test flight. The company reorganized and moved from Chicago to Niles, Michigan. In 1935, Howard Anthony purchased the then-bankrupt Heath Company, and focused on selling accessories for small aircraft. After World War II, Anthony decided that entering the electronics industry was a good idea, and bought a large stock of surplus wartime electronic parts with the intention of building kits with them. In 1947, Heath introduced its first electronic kit, the O1 oscilloscope that sold for $50 - the price was unbeatable at the time, and the oscilloscope went on to be a huge seller.
It probably would have been worth $250 just for the hub.

My affinity for old tractors probably comes from learning to drive one way back in my pre-teen years. While we lived in a city, we also had a farm that we would go to every weekend. We had a Farmall M and a Ford Jubilee, but that doesn't limit me to looking at only those two brands.

Which isn't to say that I don't like the old Farmalls the best!  When I first started driving one, I had to get off the seat for my short little legs to fully depress the clutch.

Unable to resist the lure of the CHUFF   CHUFF   CHUFF of an old single piston engine, we finally tracked it down. Even with all of the mechanicals exposed and a fairly decent understanding of engines, we still took awhile to figure out where all of the requisite parts were, as you will hear in this short video:

The oil is a consumable, just like gas. It doesn't recirculate in the way modern engines do.

This old tractor caught my eye, primarily because of its color.

This is my favorite picture of the bunch. Although I was too small at the time to get up to eye level with it, this is the way I remember our old tractor - not as a pristine, better-than-new restoration, but as an aged implement kept in a dusty old barn.

As The Judge and The CEO are both interested in equine sports, we passed through one of the horse barns.  Other than the radio, this is a picture I could have taken sixty years ago.

As we were leaving the fairgrounds, I thought my eyes were deceiving me - it seemed like I was looking at a twin engine airplane parked in a field.

I was!

Even stranger was this immense croquet court. I can't even manage a guess as to what this is used for!

Back at the airport, my favorite statue ever is still there, expectantly watching the eastern skies waiting for the westward bound air mail plane to arrive. I have no idea, nor can I find an record of, KVES having anything to do with the old airmail routes, but there he is, incognizant of his place in history, but waiting and waiting just the same.

The two birds look quite sharp together.

The new autopilot panels have simplified the flying to the point where I can kick back and be a passenger, more or less, should I chose to. With a 67 degree OAT, it was quite a comfortable ride.

Piqua, OH.

The bugs were out in force, so I had some clean up to do when I got back.

The Windex and microplush towel are cleaning supplies. The other bottle holds a type of repellent: I drink it to keep me from jumping right back into the plane and going flying again!

Thursday, July 2, 2015

Automating the Autopilot

I love my SkyView.

I remember back in my RV-6 days ogling the demo units at Oshkosh, mentally adding up the costs that would be associated with retrofitting such a system into Papa Golf, and rapidly deciding that I was not likely to ever enjoy flying with modern avionics of that sort.

I say this because I don't want the following to be viewed as criticism of this wonderful piece of kit.

So here it is: the autopilot is something of a persnickety pain in the posterior to operate.

You have a choice between Simplified Controls or Complex Controls.

With the simplified controls, there are only two pertinent soft buttons: TRK+ALT and HSI+ALT. They both do the same thing, more or less, with the only salient difference being whether the directional control is to be based on the Nav source (typically GPS in an RV-12) or a magnetic heading as set with a little 'bug' configured through the use of a knob on the SkyView.

In typical operation, what happens with the SC ("simplified controls" in a simplified typing style) is that I already have a GPS course set in as I taxi to takeoff. After takeoff and with a 1,000' of air below me, I will press the HSI+ALT button. That will cause two things: first, the autopilot will enter a turn to capture the GPS ground track, and it will lock the altitude at my present height, regardless of what altitude I want to climb to.

Here's where the limitation of the stock SkyView comes in: it only has one knob for adjusting things like autopilot altitude, vertical speed, GPS track/mag heading, barometric altimeter setting, etc. To set the function that is to be changed, you press up on the knob (it acts as both a rotary setting knob and a little control joystick) to select the function, then press to the right to set that function, then rotate the knob to adjust the value.

It's somewhat burdensome.

I had gotten into the practice of setting the knob function to ALT just before takeoff - this would make it easier to press HSI+ALT and very quickly reset the target altitude to my desired cruise altitude, Later, if I wanted to set the directional control to TRK/HDG, I would have to go through the knob function setting again. This is not uncommon; I often use the TRK setting to offset my course when I see traffic heading right at me, or when I want to position my arrival at the destination airport to a more suitable spot in the sky for a pattern entry.

While enroute, the autopilot still needs a lot of attention if there is any vertical lift/drop component in the current weather (which is almost always the case) as it tries to maintain a constant altitude. If it detects that it's holding control pressure against a rise or drop for any appreciable period of time (just a few seconds), it will post an alert asking for TRIM UP or TRIM DOWN. It generally alternates between the two. A more sophisticated autopilot would know how to manage the trim setting by itself.

These are, of course, minor irritations when compared to having no autopilot at all, but still... if only there was a way....

There is.

Dynon doesn't consider the SkyView to be suitable only for little LSA airplanes like mine. In fact, if it wasn't for the cost/benefit ratio being to dominantly weighted towards benefit, it wouldn't make much sense to use such a capable device in a little bird like mine.  But when you consider, say, an RV-10, which is likely to be used for lengthy trips and in inclement weather, having more granular control over the autopilot is a must.

For this reason, the SkyView also provides a more complex set of menu items, but this complexity comes at the cost of requiring much more interaction with that single little knob. This would not do, so a set of optional panels was created.

These little panels provide the buttons and knobs required to have much easier access to the complex settings, and additionally interface with the trim system.


You can buy these from Dynon or any of a plethora of vendors, but what you will get is a box containing the panel and the pins and backshell for making the required connectors. You can also buy them from Van's for only a little bit more money (Van's is very, very fair in their markups) and get a complete cable harness that is already built to work with the big junction box and the two panels. You also get the required pieces/parts to install the panels in the, well, panel. Plus... instructions!!

Counter to my normal process, I actually read the instructions all the way through before starting. This was intended mostly to learn what tools and supplies I would need. On the tool front, I was going to need a Dremel-style rotary tool and a reinforced cutting disk.  The rotary tool was under $20 from Harbor Freight, but I went name brand for the cutting disks.

I've had bad experiences with these cutting disks - they are, in fact, the scariest tool I ever use. The disk is normally held onto the mandrel with a little screw. Don't tighten it enough and it can work loose and release the disk into flight.  That's not good. Get it too tight and you can crush the area of the disk under the head of the screw and cause it to fail. Again, a flying disk.

Dremel has a new way:

You can see that I left some of the mandrel exposed - the cutting disk is NOT a precise tool in use and the areas where I would be cutting don't offer a lot of room. Leaving the mandrel out a little bit would make it easier to get the cutting edge where I needed it.

DO NOT DO THIS!!  Get the mandrel all the way down into the chuck of the rotary tool.

You'll see why.

Here's everything I thought I would need, minus what Van's calls 'PVC Tape,' which no one at Lowes had ever heard of**.

Mildly suspicious that this might be another METRIC CRESCENT WRENCH thing, I just went with Gorilla Extra Tough Duct Tape as an alternative. The 'LIGHT' package is disposable plastic drop "cloth."

The instructions start with positioning the template for the two cut-outs that will have to be made in the panel.

Before doing that, I went ahead and removed the avionics cover.

I shouldn't have - the template needed two of the nutplates to hold it in place.

Then the radio is removed. That's a very easy thing to do.

Then the SkyView screen is removed. I never did figure out why as it wasn't really in the way of anything.

The drop "cloth" is used to create a catch for all of the metal shavings and stuff that will be flung about during the cutting of the panel holes.

And then... the first cut (in a technical sense, I suppose)

Rats. The bit walked a, uh... bit on the first hole. I hope that doesn't turn out to be important.

Then you go through a convoluted process of using a scribe to mark the outer boundaries of the slots, followed by using a Sharpie to draw the outline. The possibility exists that the same result could have been achieved more easily simply by drawing boundaries outside of eight of the twelve holes, although that could lead to disaster if you selected the wrong eight.

This is where the mythical PVC tape has been replaced by the Gorilla tape. I don't know if it made any difference. I suspect the PVC tape might have protected better against accidental contact with the cutting disk than the duct tape did. The blue tape is to provide a color contrast when cutting.

Despite their power and convenience (or because of it!), these cutting disks are frightening to use. I would have much preferred using something more protective than just safety glasses. You know, something like a welding mask, but with better visibility. I've had enough of these disks come flying off that I have become heartily distrustful of them.

It didn't take much cutting to figure out why the disk/mandrel should be fully pushed down into the chuck. All it took was relatively high RPM and one little inappropriate twist...

Luckily, the disk flew away from me, not towards me.  I went back to Lowes and bought another mandrel, and you can bet that I got that thing all the way down into the chuck this time.

Like I said, it is NOT a precision tool.

After dressing up the rough edges and cutting a little more off of the bottom, the first panel fit in.

The drop cloth did exactly what its name implies: it dropped all of the crud right into the seat. To be fair, Van's depicted using more tape to hold the bottom of the drop sheet up higher, but I didn't think I would be able to adequately control the cutting disk while standing outside the airplane and leaning in. I had to sit in a seat to get a workable angle.

Except for the stuff that attached itself to my arms.

Once both of the panels are in, more holes are drilled and nutplates are installed.  That was the worst part of the entire project. There wasn't a great deal of room to work, and there were areas where my lack of precision with the cutting disk had left too little edge distance for the rivets that attach the nutplates to the panel.

I didn't use the screws provided by Dynon. They're the same style as those that hold the SkyView screen in and the hex head in them is too small and, at least in the case of the SkyView display, trying to get them into nutplates (which are in essence lock nuts) can cause the hex edges to get rounded out. I had at least one of those that ended up being difficult to remove.

I decided to use the same size screw (6-32) with larger cap heads. These are typically used in R/C airplanes, so I just ran up to Hobby Town to get them.

A little rant here: as I was trying to find enough of the size I needed (they had only one pack of four in each size), I overheard the store owner griping to another customer about how online stores are killing his business. Well, yeah. But here I was, more than willing to pay a little more than I would from a web store for the convenience of getting them right away rather than having to wait for shipping, but he didn't have them in stock! I'll bet Tower Hobbies does...

Also, not a rant but more of a mea culpa: that Gorilla duct tape holds onto stuff like a pit bull protecting a nice juicy bone. Removing it took more of the powder coating off in the area that was damaged by the paint shop.  That said, the flanged areas of the panels were wide enough to hide the ugliness of the panel cutouts.

I do wish the hobby shop had had black screws instead of silver.

For the wiring, the old autopilot plug gets disconnected. I could probably cut it out of the plane entirely, but as it weighs only a few ounces and would be hard to replace... I just tied it down and left it.

The SWITCHES connector has to be modified. The new autopilot panel becomes the King of Trim, so the three existing trim switch wires have to be removed. That means opening the backshell and removing pins.

I hate pulling pins. I hate it even more when I'm supposed to remove pins 15, 18, and 19, but remove 15, 17, and 18.

I really need to adjust to the encroaching failure of my short-term memory.

By the time I realized my mistake, I had three unmarked white wires dangling loose with no way of figuring out which one should be returned to its rightful home in spot 17.  I had to go back home and consult the wiring chart to determine which pin at the other end of the wire should have continuity with pin 17.

It's pin 21 for those keeping score at home.

The three old wires just get covered in heat shrink and tied back out of the way. I later combined the three loose bitter ends of the three wires into a wider diameter piece of heat shrink in the interest of neatness.

The plugs are easy-peasy to install.

Then there are changes to be made to the WH-00007. This is the worst part of all of the wiring. It was problematic during the install (due to a small wiring mistake by Van's that I've encountered a couple of times now as I work on other peoples planes). Problems like this come from being one of about 40 builders that got caught between the transition from the older Dynon D-180 to the newer SkyView.

Don't get me wrong - the transition to the newer avionics was a godsend and Van's is to be commended for not just leaving us hanging with the older technology. It adds a little bit of complexity long-term, but that is a very small price to pay.

That said. the connector is located in a horrible location under the panel, and nothing good ever comes from touching it. This resulted in yet another trip home to consult the wiring diagram as the colors of my wires didn't match the more festively colored wires used in later kits.

The Van's instructions for the configuration updates required to make in the SkyView software are very precise. So precise, in fact, that I was lulled into thinking they would also be very complete.

They were not.

They skipped an essential step.

That caused no small amount of consternation and self-doubt as the panels didn't work when I first applied power to the system and attempted to calibrate the auto-trim function.

Naturally I suspected I had done something wrong in the always suspect WH-00007. Sometimes past history does indicate future performance. I've been frustrated with that connector and its wiring so many times that it has become an ingrained habit to always assume that any autopilot problem must be caused by that little four-wire devil.

I could find nothing wrong, so after far too long I finally resorted to the what-should-have-been-obvious fallback of running the Network Detect function. This is akin to the CTL + ALT + DEL three-finger salute that PC users are so accustomed to. It should have been the first thing I tried, but after having been lulled into complacency by the robust step-by-step instructions thus far....

Finally, though, the light bulb (EPA-compliant CFL) went on over my head and I launched the detection process.

Ta-da!!  There they are at the bottom of the list.

Having flown with the panels and taken some time to learn the differences between the simplified and complex modes of operation, I am ready to proclaim that this upgrade was worth every bit of the investment of dollars, time, and self-induced frustration. The auto-trim provides exactly the kind of 'automatic' that you expect from an autopilot, and having separate knobs (for everything but vertical speed, but the UP/DOWN buttons are close enough) is very convenient and cuts way down on the fiddling around that is the hallmark of using the single control knob on the SkyView screen.

I did run into a problem caused by using the Van's config files for the SkyView. While they were convenient for getting everything operational, they had an unintended side effect. Van's now supplies a little microswitch that detects if the canopy latch is, uh... latched before takeoff. The Van's config file expects that switch to have been installed and views its absence as an indication of the latch not being correctly latched. It responds to this with a loud screeching alarm just after takeoff, and that proved to be quite distracting as I could find no way of shutting it up. Ironically, being able to engage the autopilot while I dithered around looking for a way to make it stop was extremely helpful.

A very smart individual on Doug Reeves' Van's support forum gave me a hint (well, explicit and correct directions*, to be honest) as to how to turn it off in the SkyView settings, so all is good now.

* "Go to Setup/EMS Setup/Sensor Input Mapping. Scroll down to the line that starts with C37 P12. Read across and it should say "Canopy" on the right. Select that line using the select button, then change the function from contact to unused."

** "PVC tape" seems to be another way of saying "electrical tape."