Completing Safety Directive SD-00001 Replacement Of Trim Servo Motor

A potential safety issue has been identified with RV-12’s that have the Ray Allen ES MSTS-T3-7A-2 trim servo motor installed. The issue involves the threaded portion of the trim servo motor shaft just forward of the AN315-3R lock nut that secures the AN665-21R clevis. The shaft is bending and, in some very rare cases, snapped altogether. The service directive requires mandatory replacement of the ES MSTS-T3-7A-2 trim servo motor with an upgraded Ray Allen ES MSTS-B6-7T-165 trim motor if the shaft is determined to be bent … or after 1000 hours of flight time.

Drawing from Safety Directive SD-00001 showing the location of the fatigue area.

The replacement Ray Allen trim servo motor has a part number of ES MSTS-B6-7T-165 and MUST to be installed with a bushing (Van’s part number BUSH-BS.188X.313X.222) that slips over the short threaded area of the shaft preventing the shaft from bending. Van’s also suggests when replacing the trim servo motor to also replace the F-1287A servo mounting tray with the newer F-1287A-1 mounting tray. The new F-1287A-1 mounting tray supports the DB9 connector Van’s has switched to … replacing the micro-Molex connector (problematic for some builders not having the proper crimping tool for the tiny pins) used on legacy RV-12’s.   When switching to the new F-1287A-1 mounting tray, you will also need to order an additional AN315-3R nut and Bag 2670 from Van’s which contains the male and female electrical pins, bodies for the new DB9 electrical connectors along with the necessary mounting hardware, plastic bushings and some rivets.

Drawing from Safety Directive SD-00001 showing the new ES MSTS-B6-7T-165 trim motor, bushing BUSH-BS.188X.313X.222 and AN665-21R clevis.

The decision was made that it would be better to take the approach of being proactive rather than reactive so decided to change out the trim motor …. Especially because, to Ray Allen’s credit, the company is offering RV-12 owners a substantial rebate if exchanging the old trim motor for a new one.

The DOG Aviation RV-12’s trim servo motor did not have a bent shaft. However, decided to make the exchange anyway and not need worry about the trim servo motor shaft bending issue anymore. Plus, I liked the idea of switching over to the new F-1287A-1 mounting tray which supports a DB9 connector to replace the micro-Molex connector.

New F-1287A-1 mounting tray which needs to have the doublers cut away and riveted onto the servo tray. Also, my finger is pointing to the new mounting flange for the DB9 connector.

After separating and smoothing the edges of the servo tray parts, they were primed with SEM primer and top coated with white paint for extra protection. Van’s calls for LP4-3 rivets to assemble the doublers onto the servo tray, but I used solid AN470AD4 rivets instead for the assembly. Next the plastic bushings are installed and filed down so they just clear the servo tray. (This is the same process as used on the original servo tray … in fact, I probably could have gotten by using the plastic bushings from the old servo tray).

Below is a photo of the new Ray Allen trim servo motor on the right and the old trim servo motor on the left. The three most notable differences on the new motor are … the round brass portion of the actuator that the new brass bushing will seat against, only four mounting holes as opposed to the six on the older unit and if one looks closely, the four corners of the case are scalloped a little.

Original Ray Allen ES MSTS-T3-7A-2 trim servo motor on the left and the replacement Ray Allen ES MSTS-B6-7T-165 trim servo motor on the right.

On the original trim motor instillation, Van’s covers the trim motor’s mounting flanges with doublers … but they are not called for on the new trim motor. I like the idea of the doublers because the trim motor’s mounting flanges appear to be just a tough plastic. Unfortunately, because of the scallops on the case of the new trim motor the old doublers can’t be used. I tried to file the correct profile in one of old doublers and was not happy with the results … because of the center hole in the old doublers, it leaves a very thin doubler in the area of the center hole.  Although not called for, I decided it would not hurt to just make a pair of doublers as can be seen in the photo bellow.

After trimming the old doubler (top of the photo) to fit the new trim servo motor, one can see how thin the metal is around the center hole (which is not on the new Ray Allen trim servo motor). So a new doubler (bottom of photo) was fabricated to mount the new trim servo motor in the F-1287A-1 mounting tray.

The new Ray Allen ES MSTS-B6-7T-165 servo trim motor installed in the F-1287A-1 mounting tray with my handmade doublers added for good measure.

For the final assembly, the forward threaded shaft temporarily receives two AN315-3R nuts which are tightened together or "double nutted", as they say. The nuts are temporarily used so a wrench can be used to hold the shaft from twisting and torqueing the internals of the trim motor when the AN665-21R clevis is tightened against the bushing. Prior to final assembly Van’s wants the clevis to be 15° from vertical when dry fitting the parts together using only fingertip pressure.

Two AN315-3R nuts are used to double nut the servo motor’s shaft so a wrench can keep the servo motor’s shaft from twisting and possibly damaging the servo when the bushing and AN665-21R clevis are tightened together.

The bushing in the above photo is slightly longer than necessary. To insure a proper fit, Van’s recommends using a drill press with some sandpaper to remove a little material at a time from the bushing so the bushing’s edges remain square. This is one place you don’t want to do any hand filing because Van’s wants the bushing to be a tightly mated fit between the trim motor and the clevis. Material is removed from the bushing until the slot in the AN665-21R clevis is approximately 15° degrees BEFORE vertical when the clevis is hand tight to the bushing. Note: Go slow! … only remove a few thousands at a time because it doesn’t take removing much material to make quite a difference in the positioning of the clevis ( I almost over did it on the second cycle to the drill press where I removed quite a bit more material than I did the during the first cycle). When a finger tight dry fit 15° shy of vertical is achieved, the bushing is ready for final assembly. Permanent red Loctite thread locker is applied to the aft threaded portion of the trim servo motor shaft and while holding a wrench on the double nuts to prevent the shaft from twisting, the clevis is threaded on further beyond the 15 degree point where it should become snug as the clevis reaches its proper vertical orientation.

Completed trim servo assembly ready for electrical connections and final instillation. Note, the two AN315-3R nuts are not yet removed from the forward threaded portion of the Ray Allen servo motor. The two nuts need to be removed at this point prior to installing the assembly back on the RV-12.

As previously mentioned, the new F-1287A-1 servo tray is designed for use with a DB9 connector. Female pins are attached to the wires coming from the trim servo motor and male pins are attached to the wires exiting the tail cone. There is a small change in the colors of the two power wires going to the new trim servo motor (the three trim position wires remain the same colors). The old trim servo motor power wires are both white … the new trim servo motor uses a white and a gray wire. Before permanently installing the wires into the DB9 connector bodies, I thought it best to use a 9v battery to make sure the new motor moves in the same direction as the old motor did … doing this insured the two white wires (servo trim motor power) exiting the tail cone will be connected to the new trim servo motor so motor movement is the same as the old trim motor. After the correct motor movement was established, the trim motor power wires exiting the tail cone were marked. Next, the pins were inserted into the DB9 housings. Van’s suggests sealing all the wires with silicone RTV so that was done prior to final assembly.

After the silicone RTV cured, the DB9 connectors were installed onto the F-1287A-1 trim servo tray. While trial fitting the connectors together, I noticed the DB9 connector was not fully seated. Upon a little investigating it was determined that because the male DB9 connector from the tail cone rests on top of the F-1287A-1 servo tray which has approximately a .060" or so of thickness, the standard sized threaded barrels are a tad too long … so I removed .060" from each threaded barrel and now have a fully seated DB9 connectors (If one chooses not to do this, it is not a big deal. However, I wanted my connectors fully seated and it was easy to accomplish just by removing a little material from both threaded barrels).

The DB9 connector #4 mounting hardware is NOT a piece of cake to install. Access is limited and the use of the tiny #4 MS21042 all metal hex stop nuts makes instillation a real hassle. The hex stop nuts are slightly egg shaped so they really grip the pan head screws … the problem is they grip just a little too tight when trying to install in such close quarters using a 5/32" wrench and only being able to tighten one flat at a time. I finally resorted to placing the head of the screw in a vice and running the metal stop nuts on and off a couple of times using CorrosionX as a lubricant to reduce the bite. The other issue I ran into was my long thin Philips screwdriver was 70 miles south at the southern outpost, so I needed to cobble together another way of accessing the #4 Philips screw heads from above the stabilator while using the 5/32" wrench from underneath the stabilator. Standard Philips screwdrivers are too short to accomplish this task when working by yourself. Below is a photo showing a drawing of how the DB9 connector is to be mounted onto the F-1287A-1 servo tray and the Rube Goldberg use of tools to fashion a way to get on the heads of the #4 Philips mounting screws from above.

Lacking immediate access to my very long thin Philips screwdriver, a Rube Goldberg assembly of various tools from the tool box was used to devise a way to hold the #4 Philips screws from above the stabilator. The photo also shows the drawing for mounting the DB9 connector.

Completed instillation of the new Ray Allen ES MSTS-B6-7T-165 trim servo motor, F-1287A-1 servo tray with the DB9 connector in place. Per Van’s instructions, silicone RTV is applied to the wires to seal the connector.

With the exception of dealing with the #4 mounting hardware for the DB9 electrical connectors, swapping out the RV-12’s trim servo motor went smooth and is not a daunting task. However, one does want to be careful when using the drill press to remove material from the bushing. Go slow and only remove a little material at a time to creep up on that 15° sweet spot.

RV-12 Rudder Pedal Block Extensions

Thus far, I have been very pleased with flying the DOG Aviation RV-12 … it is truly a nice flying airplane. My only minor complaint stems from the rudder pedal configuration, in that, I find it difficult to make large rudder inputs with my feet on the rudder pedals without also pressing on the brakes unless I keep the balls of my feet pulled back in an unnatural position. Of course, for most ground operations or while in the air, this is not an issue … but on short final or during takeoff roll it requires a cognizant effort to pull the tips of the toes aft to insure large rudder inputs are not accompanied with unwanted brake inputs.

Sliding the feet higher on the rudder pedals would help but then the tips of my shoes (size 12) run into interference with the firewall shelf. Admittedly, it is not a big deal … just a few moments of uncomfortableness while pulling the toes aft to keep them off the brakes during takeoff and landing.

That got me thinking ... Gee, if the F-1290 pedal blocks were a little thicker, it might put my feet in more of a natural position on the rudder pedals and  help keep them off the brakes. With all the buzz about 3D printing, I decided to see if I couldn’t make a set of pedal blocks that are exact copies of the Van’s F-1290 pedal block … just a little thicker so the feet are in more of a natural position when on the rudder pedals. But that posed a problem … I have never used any type of 3D modeling software, so where do I begin?

EAA (Experimental Aircraft Association) to the rescue. I was talking to some friends at the airport about wanting to 3D print a set of thicker rudder pedal blocks, but didn’t have access to 3D software. I was informed that as an EAA member it was possible to obtain a free full featured (minus stress analysis) version of Solidworks Educational Premium that college student’s use. The EAA has formed a partnership with Solidworks allowing EAA members to download Solidworks Educational Premium for free … along with a 1 year license agreement which can be renewed as long as EAA membership is valid.

Knowing nothing about Solidworks (or any other drafting program), it seemed like learning to draw the thicker pedal blocks would be a daunting task. When first launching Solidworks, my eyes immediately glazed over … the sheer number of menus, options and submenu items available is truly daunting for a beginner. Fortunately, within Solidworks (under the help menu) there is a very good built in tutorial that starts out simple and builds on previous lessons.

After playing with the tutorials for a few evenings, I was able to acquire enough knowledge about Solidworks to successfully draft up a thicker rudder pedal block by carefully measuring the F-1290 pedal block and entering those dimensions into Solidworks then adding an extra ½" to the overall height of the petal block. I did not want to change the mounting bolt hardware so a recess was made for the mounting bolt to drop into. With a little bit of tutorial practice and some patience, making an accurate drawing for a thicker pedal block was not that hard. Below are a few screen shots of the finalized design drawing.

Top view of the thicker rudder pedal blocks drawn using Solidworks.

Bottom view of the thicker rudder pedal blocks.

Side view of the thicker rudder pedal blocks.

 End view of the thicker rudder pedal blocks.

A couple of years ago Tom, a friend who is building a RV-10, mentioned he bought a 3D printer and made the offer should I ever want to have something printed, to let him know. I told Tom about my using Solidworks to create a drawing for thicker rudder pedal blocks and he said he had plenty of blue HDPE filament … so Tom printed two blocks for me that are hollow inside with a crosshatch pattern for strength to save on material, since they are just for proof of concept. Thanks Tom!!!

The blue RV-12 rudder pedal block on the left was printed by Tom using the Solidworks file I made.

Side by side one can see the blue pedal block is the same as the black Van’s pedal block on the right … just a ½" thicker.

Bottom view of both rudder pedal blocks. Have to say, as a first time effort at making a 3D drawing and then having a part printed from the drawing, the thicker rudder pedal block looks great.


The big question now is how well will the thicker rudder pedal blocks actually work out? Only one way to find that out … install them.

The thicker proof of concept rudder pedal blocks installed on the pilot side of the DOG Aviation RV-12.

For comparison, these are the original Van’s F-1290 rudder pedal blocks on the passenger side.

So you are probably wondering … was it worth the effort … absolutely! My feet are now in a much more natural position when placed on the rudder pedals. Although the blocks could even be a tad thicker, the positioning of my feet on the rudder pedals is much more comfortable while moving the rudder pedals and staying off the brakes ... so I’m calling this effort a success.

Now that I know the Solidworks drawing I’ve made will produce a functional part, the plan is to make up four rudder pedal blocks from a tougher black material such as NylonX or CarbonX. Unfortunately, Tom’s 3D printer does not have the proper type of printer nozzle required for those filaments, so I will need to have the final pedal blocks printed elsewhere.


Return from the future:


As previously mentioned, the printer nozzle on Tom’s 3D printer is not suitable for printing stronger filament plastics like CarbonX or NylonX … so a few months back, I contacted Steve at rvplasticparts.com who is a fellow RV owner on the Van’s Aircraft Forums (VAF) that has a cottage business of 3D printing accessories for owners of various models of Van’s aircraft. I told Steve I've made a drawing of a taller rudder pedal block and had already printed a prototype which fit nicely. However, I wanted to have a set printed using a stronger material such as CarbonX or NylonX … Steve decided NylonX would be a good choice. I told Steve he could have my drawing to use if he wanted to add the thicker rudder pedal blocks to his product line and he decided to take on the project.


Steve tweaked my drawing a tad and will be offering the taller rudder pedal blocks in an assortment of thicknesses. The prototype Steve sent me is 3/16" taller than my original prototype … which should be perfect since my prototype blocks felt good … but, as mentioned earlier in this post, could stand to be a little thicker. Another change Steve made was increasing the thickness of material under the bolt head from ½" to ¾" feeling the end result will be much stronger. The original rudder block from Van’s was ½" so I made my drawing using the same thickness which created a surprise when I tried to test fit the new taller NylonX rudder pedal blocks Steve sent me …. the original AN4-15A bolts were not long enough, so I needed to order longer AN4-17A bolts. Below is a photo showing the new NylonX rudder pedal block on the right, my proof of concept prototype in the center and Van’s stock rudder pedal block on the left.

As can be seen in this photo, the NylonX rudder pedal block on the right is 3/16" taller than my original blue prototype in the center which is 1/2"taller than the stock Van's rudder pedal block on the left.


As previously mentioned, my blue prototype shown above felt good and is a vast improvement over the stock Van’s rudder pedal block … but I felt the block could stand to be even a tad taller. When the longer AN4-17A bolts arrived today, I headed to the hangar to test the fit of the new NylonX rudder pedal blocks Steve printed for me. At first, I only replaced one block so I could compare the feel of my prototype with Steve’s 3/16" taller rudder pedal blocks. The overall feel was about the same … but I noticed Steve’s NylonX rudder blocks allowed larger rudder inputs before the foot actually began pressing on the brake pedal. So overall, I’m very happy with the thicker NylonX rudder pedal blocks Steve printed and will order another set for the right side of the airplane.

The new taller NylonX rudder pedal blocks installed. Note to fellow RV-12 owners, changing to these rudder pedal blocks will require longer AN4-17A bolts to replace the AN4-15A used to install the stock Van’s rudder pedal blocks.


A big thanks to Steve at rvplasticparts.com  for taking on my project and sending me a prototype to test. I think this is one modification other RV-12 owners will want to look into … especially if keeping your foot off the brake during rudder inputs feels slightly uncomfortable for your foot.


___________________________________________________

The photos below are just stored here so I can post them on the forums. Basically, kicking around an idea here. Background: The owner of a RV-12 has a grade in front of his hangar to traverse and is trying to figure out a good way to get his airplane into the hangar without holding onto the tow bar so he can hold a winch controller or pull a rope connected to pulleys.

I suggested reversing the tow bar and pull it with a winch or rope/pulley setup connected to the reversed tow bar. However, it was pointed that without locking the nose wheel somehow, should the tail begin to track off center a little, the castering nose wheel will make matters worse and quickly get the airplane off track … true enough.

So that’s had me thinking for a few days and I think I’ve come up with an idea that may solve the guy’s problem. During the last unexpected warm day we had here I slipped up to the airport to take a few photos I will post on the forums. The idea is to use the RV-12’s steps to support a board traversing the bottom of the fuselage that the tow bar can rest on and be secured between blocks. The blocks will prevent the tow bar from moving side to side thus keeping the nose wheel locked straight while the tow bar is being hauled aft into the hangar by a winch or rope/pulley arrangement. Below are proof of concept mockup photos using a yard stick so the RV-12 owner can see my vision in a picture form.

The tow bar is placed on the nose wheel reversed as shown here and will sit on horizontal piece of wood or metal traversing the belly of the airplane. The areas where the blue tape is placed around the tow bar will need blocks attached to the horizontal piece to capture the tow bar. The blocks will capture the tow bar preventing side to side movement thus keeping the nose wheel held straight as the airplane is pulled backwards into the hangar.

The horizontal piece needs to be connected to the RV-12’s steps using a vertical piece that has wooden pins which will insert snugly into the step’s tubing as shown in this photo … or the vertical piece could slide over the step’s tubing, builders choice. The vertical piece will likely need some gussets where it meets the horizontal piece to stiffen the assembly, so the fixture has no side to side play. The steps will lock the whole structure preventing ant side to side movement and also support the tow bar as the airplane is pulled aft into the hangar.

As can be seen in this photo, the Van’s tow bar ends well forward of the com antenna.

Looking aft, As can almost be seen in this photo (bad focus), the com antenna is biased to the pilot side of the RV-12’s bottom fuselage … so a cable or rope connected to the center of the tow bar handle will easily pass by the com antenna.


I  believe my idea will likely work well, however not having an incline to traverse to get inside my hangar, won’t be constructing a working version to prove the idea actually works.