Roee Kalinsky's RV-7A Project

2009.05.04: (0.0) I decided early in the project that I will install a 2-axis autopilot.  Before going into further detail, let's get some basics out of the way.

WHY AN AUTOPILOT

Several people, both pilots and non-pilots, have asked me: "Why would you want an autopilot?  Isn't the whole point of this airplane that you will enjoy flying it?"  Fair question.  First of all, yes, the main purpose of this airplane is that I'll enjoy flying it.  But "flying it" doesn't exclusively mean "stick-and-rudder flying it".  The RV-7A has flying characteristics that make it a very fun stick-and-rudder airplane, but it is also a very versatile airplane, and an autopilot helps make full use of that versatility.  More specifically, the RV-7A's handling characteristics are light and quick, which is a lot of fun, but also requires a high level of attention from the pilot to maintain a desired flight attitude, heading, and altitude.  These handling characteristics are ideal for fun VFR flight and sport aerobatics, which are at the center of its mission profile.  But the airplane is also generously "long-legged" such that cross country travel and IFR operations are also among its primary missions.  On long cross country flights, and especially in IFR, those same handling characteristics increase the pilot's mental workload and fatigue.  An autopilot allows the pilot to temporarily offload the task of physically controlling the aircraft in order to tend to other tasks, which is a great asset at certain times during a flight.

WHY A 2-AXIS AUTOPILOT

The Warrior that I currently fly has a single axis autopilot, roll only.  This autopilot can hold a heading, and even track a course on a VOR, ILS, or GPS receiver.  But it has no control over pitch, which makes it mostly useless in practice.  In perfectly smooth air, one could, in theory, perfectly trim the airplane and expect it to hold altitude.  But reality doesn't usually work out that way, and especially in bumpy IFR conditions when the autopilot would be of most value.  My experience has been that having to manually control pitch but not roll does not significantly reduce my workload compared to manually controlling both, especially if there's any turbulence at all.  Nor does it free up my hand.  So that's my final verdict on single-axis autopilots.  Mine will be a 2-axis autopilot (pitch and roll), which will afford me true hands-off flying when it's engaged.

So why not 3-axis?  What about the rudder?  Well, the RV-7A's ailerons exhibit so little adverse yaw under normal cruise conditions that sufficiently coordinated flight, including standard rate turns, can be maintained even without rudder inputs.  So within the autopilot's operating envelope, it can get by without rudder control.  So 2-axis is sufficient, mostly.  The "mostly" is that the rudder should still be in trim for coordinated flight for the steady state flight condition.  This I plan to achieve using a separate rudder trim system of my own design (none is included in the stock airframe).  I may even provide for automatic control over it using AHRS data, but that'll be further down the road.

2009.06.27: (0.0) I settled on a TruTrak 2-axis autopilot.  Why TruTrak?  TruTrak's line of autopilots is the best in class (in fact just about the best there is) in performance and reliability, and has a long proven track record in hundreds if not thousands of aircraft.  A full line of control heads is available from the most basic to full featured, and the company offers a very flexible upgrade option.  The TruTrak autopilots also integrate very well with a variety of other avionics (including Advanced Flight Systems EFIS), but are not dependent on those avionics and are capable of operating stand-alone.  TruTrak, run by the legendary Jim Younkin, also has earned a solid reputation for standing behind its products, which by the way are all under lifetime warrantee.

I purchased just the servos for now through SteinAir.  The servos will work with any of TruTrak's 2-axis autopilot heads, including the upcoming Advanced Flight Systems customized version of the TruTrak DigiFlight II VSGV, which is my current plan.  The servos themselves are TruTrak's "B" size servos, which are the smallest ones, and are what's generally used for the 2-seater RV's.  The "B" servos have a torque rating of 30 in-lb, and a residual torque (i.e. inactive resistance) of approximately 0.8 in-lb.  The roll servo is a standard servo, and the pitch servo is a version that has a trim sensing feature.  The servo purchase also includes all the necessary mounting hardware to install the servos in the RV-7A.  The pitch servo sits just aft of the elevator bellcrank in the aft fuselage, and the roll servo sits just outboard of the aileron bellcrank in the wing.  Both servos connect to their respective bellcranks using small pushrods.  I opted to mount the roll servo in the right wing for better lateral balance, but installation hardware is available for either wing.

2009.07.10: (0.0) I received the servos and installation hardware.  Everything appears well designed, robust, compact, and elegant.

I did encounter a couple of issues though upon closer inspection.  The pitch servo had a small "ding" in the housing, as if it had been dropped or slammed against a hard surface.  The ding was pretty small, but it did prevent the servo from sitting flush against the mounting bracket.  It also was cause for concern about the possibility of the servo having suffered internal damage by the same event that caused the ding.  I don't have the control head yet, so I have no way to test the servo.  But also, the servo arm exhibited a fair amount of free play, both rotational and in-out.  This is true of both servos, but much more so on the dinged pitch servo, so that also caused me concern.  I spoke with Paul at SteinAir about it, and he instructed me to return the servo for a replacement.  I did, and a new one is on its way.  After the servo in question had arrived back at SteinAir, I spoke with Stein about it.  He indicated that the amount of free play in the servo arm was normal, that the trim-sensing pitch servos always have more free play than the standard roll servos.  And he would have suggested that I simply file down the ding and retouch it with an alodine pen.  I would have actually been ok with that, so long as it wouldn't affect my warrantee.  But anyhow, a new servo is already on its way, so that's that.

The installation hardware for the RV-7 is well designed, simple and robust.  All very straight forward, but just a few minor gotchas worth noting:

Gotcha 1: The accompanying drawings, which are annotated to be 1:1 scale, were actually printed slightly scaled down.  However, they're also available as PDF's on TruTrak's web site, and I was able to print them out myself at the correct 1:1 scale.  Anyway, this only matters for setting the roll servo pushrod length, which does not have a dimension call-out on any of the drawings.

Gotcha 2: The drawing for the pitch servo installation shows the F-729A rib having its lower flange on the starboard side, whereas in fact it is on the port side.  The drawing was originally done for the RV-6, so maybe that's the source of the discrepancy.  Anyway, not a problem.  The servo mounting bracket still installs on the starboard side of the F-729A rib, opposite and F-729A's bottom flange, and the mounting bracket's bottom flange gets riveted directly to the bottom skin.

Gotcha 3: The install hardware includes most of the necessary fasteners (not quite enough washers), but the stack-ups of washers, etc. are nowhere shown on the drawings or any of TruTrak's other documentation that I could find.  Another builder on the VAF forum posted that he had obtained this information from TruTrak by phone.  Unfortunately he then didn't list the stackups, but he did post a couple of photos that show it, more-or-less.  It appears to be as follows:
- The servos are fastened to the mounting brackets with an AN3H-3A drilled-head bolt, AN960-10 standard washer, through the mounting bracket, and into the threaded mounting hole in the servo.  Then the bolt heads should be safety-wired.
- Pitch servo arm to push rod (left to right): AN3-7A bolt, servo arm, rod end bearing, AN970-3 large diameter washer, AN364-1032A thin nut.
- Pitch servo push rod to elevator bell crank (right to left): AN3-10A bolt, AN970-3 large diameter washer, rod end bearing, two AN960-10 washers, elevator bell crank, AN960-10 washer, AN364-1032A thin nut.
- Roll servo push rod to roll servo arm (top to bottom): AN3-6A bolt, AN970-3 large diameter washer, rod end bearing, servo arm, AN960-10 washer, AN364-1032A thin nut.
- Aileron bellcrank to roll servo push rod (top to bottom): AN3-12A bolt, aileron bell crank, 0.523" aluminum spacer (parts list says 0.563", but actual measured 0.523"), rod end bearing, AN970-3 large diameter washer, AN364-1032A thin nut.
I will likely tweak these stackups slightly, but that's a good starting point.

ROLL SERVO INSTALLATION

The roll servo mounting brackets couldn't be easier to install -- they simply bolt on at the location of Van's standard aileron bellcrank brackets.  (These appear to be cadmium-plated steel, but I will still likely give them a coat of epoxy primer for increased corrosion protection).

Measuring from the drawings, I came up with a pushrod length of 4-7/8" bearing center to bearing center, which in fact worked out perfectly.  When the aileron bell crank is in the neutral position (determined using the aileron bell crank jig), the servo arm is right at the 12 o'clock position, i.e. 90° to the pushrod.  The aileron bellcrank can also be swung all the way in either direction (until it hits the wing spar in one direction, or hits the servo in the other direction) without binding or over-centering the pushrod geometry.  The actual range of motion will be far less, as limited by the aileron stops, but it's nice to know that the servo installation is safe from over-centering or any other binding issues well beyond the ailerons' maximum throws.

Above: neutral ailerons (notice the aileron bellcrank jig).

Above: extreme right ailerons, well beyond aileron stop (notice aileron bellcrank hitting the spar web).

Above: extreme left ailerons, well beyond aileron stop (notice the aileron bellcrank hitting the servo).

I also decided to modify the fastener stackups slightly to get better alignment, and better security.  The new stackups are:
- Roll servo push rod to roll servo arm (top to bottom): AN3-7 drilled-shank bolt, AN970-3 large diameter washer, rod end bearing, AN970-3 large diameter washer, servo arm, AN960-10 washer, AN310-3 castle nut, AN380-2-2 cotter pin.
- Aileron bellcrank to roll servo push rod (top to bottom): AN3-13 drilled-shank bolt, aileron bell crank, 0.523" aluminum spacer, AN970-3 large diameter washer, rod end bearing, AN970-3 large diameter washer, AN960-10 washer, AN310-3 castle nut, AN380-2-2 cotter pin.

Above: view showing the pushrod fastener stackups (and my toes).

Note that the large diameter washers are playing the role of "capture washers", which is standard practice when it comes to rod end bearings (although Van's doesn't always adhere to it).  Their job is to retain, or "capture", the rod end in the unlikely event that the bearing fails catastrophically and disintegrates.

PITCH SERVO INSTALLATION

The pitch servo mounts to a bracket (provided by TruTrak) that is riveted to the ...