Installing a “Bender Baffle” Piggybacked With Regulator Relocation Mod - Part Nine

This post will cover mounting a terminal barrier block which will be used as an interface between the existing RV-12 wiring harness and the wiring harness that will run to the Silent Hektik F-4118 regulator mounted inside the lower cowl on the baffle plate.


RV-12 builders planning to relocate their regulator inside the lower cowl's air duct please take note of the following regarding the use of quick disconnect connectors for the wiring:


The wiring connector dilemma: Mounting the regulator on the baffle plate inside the air duct creates a wiring challenge because the wiring will need to be disconnected every time the lower cowl is removed. On a new aircraft like mine, removing the lower cowl will occur more frequently during the initial flight testing and then taper off to a few times a year for oil changes and condition inspections.  This, in and of itself, creates some issues. What’s the big deal you may be asking yourself? Just use a connector that has pins rated for the maximum current the Rotax 912 can produce (approximately 18 amps). Well, that’s easier said than done and here’s why ....


Because the regulator is a high current device, ANY connector utilized needs to have pins rated for high currents. Sticky wicket #1: … sure there are plenty of connector styles available on the market but ones with pins rated 15 – 20 amps are quite limited. Sticky wicket #2 and the main reason I suggest builders DO NOT use any quick disconnects for the regulator wiring is: ... if one looks deeper into the application data sheets for the connectors, unfortunately, you will quickly discover all the connectors that have pins rated for high current have a VERY LIMITED INSERTION LIFE CYCLE, with the majority of connectors being under 10 insertions. This means that after removing the cowling just 10 times (some were as low as 6) the pins in the connector would need to be replaced in order to remain within the max current specification the connector’s pins are rated at. The best connector I could find is a very popular expensive military connector rated at 20 amps per pin and that only has a life span of 25 insertions. Because of the limited insertion ratings of the connectors, I decided it best to forget about using a quick disconnect connector altogether and would suggest others do so as well …  at the currents we are talking about, a little increase in resistance at the connector pin can begin to create a considerable amount of heat at the connection leading to oxidation which increases the resistance generating more heat etc.,  to the point of melting insulation. This application requires a continuous low resistance connection … period.


My solution will be to use a very high quality screw-down terminal block rated for high temperature environments. Suggest others doing this not use the lesser quality plastic terminal blocks … you want the good stuff for this application. I chose a Cinch 6-142 phenolic terminal barrier block which is a 6 circuit barrier block rated at 30 amps per circuit and able to withstand temperatures to 300 degrees. Astute RV-12 builders may be asking themselves why use a 6 circuit barrier block when only 5 wires are run to the Ducati regulator? The answer is both the Ducati and Silent Hektik F-4118 regulators obtain their ground from the regulator's case being bolted to the RV-12's airframe ... so this necessitates running a sixth wire that will be bolted to the regulator's case to supply the necessary ground for the regulator. So a ground wire will need to be run from the ground block on the DOG Aviation RV-12's firewall to the sixth terminal in the Cinch terminal block to provide the necessary ground.

The Cinch 6-142 phenolic terminal barrier block.

Decided a good and easy location to mount the Cinch terminal barrier block would be on the two rudder pedal studs that protrude the nutplates on the RV-12’s firewall shelf. Two 3/4" coupler nuts were screwed onto the protruding studs and blue Loctite was placed on the threads to prevent the nuts from coming loose.

Plenty of threads to protruding the nutplates to install coupler nuts to use as a mounting point for the Cinch 6-142 terminal barrier block.

3/4" long stainless coupler nuts with 10-32 threads were screwed onto the portion of the rudder pedal mounting bolts that protrude the nutplates. Loctite 242 (blue) was placed onto the threads prior to installing the coupler nuts.


Next a small mounting plate was fabricated with an offset to mount the Cinch terminal block. The small offset was used so there would be easy access for a screwdriver to remove the wires, The offset is not really necessary …. however, without the offset the coolant hose blocks easy access to some of the screws and would require the use of a stubby screwdriver. The Cinch 6-142 terminal block has an exposed bottom … meaning if one were to look up from the bottom one can see the metal inside. Although not necessary, I decided to cut a thin strip of Neoprene and use it as a sealing gasket under the connector block.


The Cinch terminal blocks can use various mounting accessories. The DOG Aviation R & D department decided on using fast on connectors the side of the connector block that interfaces with the existing wiring which normally connects to the Ducati regulator. The beauty of this is ... if I decide to leave the Ducati regulator mounted on the firewall shelf as a spare, should the Silent Hektik regulator fail, the wires can easily be unplugged from the Cinch terminal block and moved back to the Ducati regulator. The wiring harness going to the regulator inside the lower cowl will have #8 ring terminals terminating at the Cinch terminal barrier block. This will allow the wires to be easily disconnected whenever the lower cowl needs to be removed. Sure it will be a bit of a hassle removing six screws every time the lower cowl needs to be removed, but the screw down terminal block will provide a very solid and low resistance interface connection between the existing wiring and the extension wiring to the regulator inside the lower cowl's air duct.

Fabricated mounting plate and the Cinch 6-142 terminal barrier block attached to the rudder petal bolt studs via two stainless 3/4" 10-32 coupler nuts. The Ducati regulator can be seen in the background and there is more than enough wire to easily reach the Cinch terminal barrier block.


Looking at the above photo, one can see that there are two 3/8" coupler nuts on the terminal block these will support a cover plate made from high temperature plastic that will cover the wires and connections when the final wiring is completed.

Installing a “Bender Baffle” Piggybacked With Regulator Relocation Mod - Part Eight

Oops …. Hit a big bump in the road when beginning the wiring instillation for the Silent Hektik 4112 regulator. Fortunately, before wiring up the Silent Hektik regulator purchased well over a year ago, I decided to visit the Silent Hektik WEB site for one last look. Fortunately, while perusing the information on the regulator, a major difference was noticed in a couple of the wiring drawings on the Silent Hektik site which show typical instillations. Thankfully, I had the presence of mind to filter the Silent Hektik WEB pages through a German to English translator. As mentioned in a post well over a year ago (when discussing the problematic Ducati regulator) much to my surprise, the Silent Hektik regulator the DOG Aviation procurement department purchased through a motorcycle shop in England came with wires coming out of the regulator, as opposed to spade leads (which is what I was expecting). As can be seen in the photo below my order was placed for a F-4112  Silent Hektik regulator that has spade leads. I asked the vendor to check with the manufacturer and was told that the only 4112 regulator being made now was the version with the wires and that the part number for the version with the spade leads was changed to F-4118. I didn't notice it at the time but the part number on the regulator I received was R-4112 and not F-4112. Apparently, there were variants of the 4112 regulator prior to the time my order was placed. It is now my understanding that all the 4112 Silent Hektik regulators with wires coming out of the regulator are wired differently …. and as such, NOT what we want to use as a direct plugin replacement for the Ducati regulator in the RV-12.  Of course, this prompted another long delay as I decided to just order the new F-4118 regulator from Europe. (Gee, wish I would have caught this issue over the winter months so there would not have been a long delay… oh well). Sure, I could have used the wired regulator by changing the wiring, but should the Silent Hektik regulator fail at some point in the future, I would have to remember one can’t directly plug in the Ducati regulator which I will now have as a spare or use another regulator obtained locally such as a NAPA, John Deere, etc. without making a wiring change as well.


Builders beware!!! There is the BIG BIG difference between the wired version of the R-4112 Silent Hektik regulator and the spade connector version of the F-4112. On the standard Ducati regulator reading the terminals from left to right you have G G R +B L C.  On the Silent Hektik R-4112 WITH WIRES, the terminals are marked as G G R B- L C.  Yet  on the F-4112 version of the Silent Hektik regulator with the spade connectors (like the Ducati unit), the terminals are marked as G G R B+ L C.

This is a photo of the F-4112 regulator with the spade leads … it is the direct replacement for the Ducati regulator.  Note that the fourth terminal from the left is labeled as … B+ Bat.

This is a photo of the R-4112 regulator the DOG Aviation procurement department received from England… however, it is the version that has wires. Note the label shows the fourth wire from the left is labeled as … B- Bat …. Not good.


The big difference between the two regulator versions is how the regulator obtains its ground …. The R-4112 unit WITH WIRES gets it ground from the B- Bat lead (blue wire) which needs to be connected to the battery's ground terminal. However, the RV-12 wiring configuration connects that lead to the battery's positive terminal, not good!!! The version of the F-4112 regulator with the spade leads uses the regulator’s case as the ground and the B+ terminal is connected to the battery's positive terminal … as it is for our RV-12 application and therefore the spade connector version of the F-4112 regulator IS a direct replacement for the Ducati regulator. Of note: Because I’m planning on installing the version of the regulator with the spade leads on the newly constructed baffle plate, a ground wire will need to be attached to the case of the regulator which will be run to the grounding block that was installed on the DOG Aviation RV-12's firewall.


I’m assuming having two very different versions of the regulator sharing the same  base part number must have created quite a bit of confusion. Perhaps it is why Silent Hektik changed the part number to F-4118 for the regulator with the spade terminals … and now the one they recommend on their WEB site as a replacement for the Ducati regulator on Rotax912/914 engines. The new F-4118 regulator IS the one we want to install as the direct plugin replacement for the Ducati regulator. It also has some nice features that will be mentioned below. If you have one of the F-4112 regulators with the spade leads you should be OK,  just check that the sticker above the terminals says B+ Bat and NOT B- Bat.  Just be sure NOT to use one of the regulators with the wires because it IS NOT pin for pin compatible with the RV-12’s wiring for the Ducati regulator.

Close-up of the new Silent Hektik F-4118 regulator.

As can be seen in the previous photo, the spade leads on the F-4118 regulator are marked as G G R [B+ Bat] L C which is what we want as a direct replacement for the Ducati regulator. The F-4118 regulator also has an added built-in feature of a four stepped voltage output ... so it can be used not only on lead acid batteries, but also on AGM and Ion batteries as well … both of which benefit from the voltage stepping. The voltage stepping ranges from 13 to approximately 14.3 volts in four steps spaced at approximately 9 minute intervals.

Photo of a charging graph obtained from the Silent Hektik site showing the charging steps for the F-4118 regulator.

Also of note, it appears to me the F-4118 regulator seems better suited for an aircraft instillation than the R-4112 regulator … in that, the regulator appears to be sealed much better from the elements. The R-4112 regulator has a seam running around the regulator because it is assembled from two machined aluminum halves which appear to have no gasket or sealant to make the seam water resistant … also, there are exposed tiny threaded screw holes between the fins that are used to mount internal components. The seam and the tiny holes would have required an application of a thin bead of high temp silicone to help make the regulator water resistant. However, the construction of the new Silent Hektik F-4118 regulator is very different … in that, there is no mating seam or threaded mounting holes between the heat sink fins. The F-4118 regulator is beautifully machined from one solid billet of aluminum and the circuitry is inserted into the body of the regulator from the bottom and has an aluminum cover plate which appears to be potted in place … the regulator is seemingly sealed very nicely with potting resin. At first I was a little disappointed that the new F-4118 regulator did not have a smooth flat bottom as there is on the R-4112 regulator (which should offer great heat transfer performance). But with the regulator installed in the lower cowl's air duct, there will be an abundance of cooling air flowing past the regulator’s fins at all times ... so excess heat should never become a performance issue for either regulator. The way the new F-4118 regulator case is made from one piece of aluminum and nicely sealed with potting resin made me quickly forget about having a solid flat bottom on the regulator.

The Silent Hektik R-4112 regulator (with wires) on the left and the new F-4118 regulator with spade connections on the right.

The new F-4118 regulator is on the right and the R-4112 regulator is on the left. As one can see in the photo, my finger is pointing to the seam resulting from constructing the regulator from two pieces.

As can be seen, the F-4118 regulator on the right has no seams or holes protruding the aluminum. The same can not be said for the R-4112 regulator on the left which one can see has at least five or six threaded screw holes between the fins for mounting internal circuitry.

View of the bottom of the regulators. The new F-4118 regulator on the right has the control circuitry inserted from the bottom and sealed with a potting resin. The R-4112 regulator on the left has a nice flat bottom which I would prefer … however, the fact there are no screw holes between the fins or seams on the body of F-4118 regulator coupled with a nicely sealed potted base scores more points in my book for an aircraft instillation.


The next post will cover working out some of the wiring details and dilemmas … plus, the changes that needed to be made when it was discovered the wired version of the R-4112 regulator would create issues.

Installing a “Bender Baffle” Piggybacked With Regulator Relocation Mod - Part Seven

Work resumed on the under the panel mounting bracket for the baffle’s control knob by trimming the excess material off the bracket and establishing a curved shape as can be seen in the photo below. Once the final shape was established, I asked Bernie to swing by the hanger so I could sit inside the cockpit and locate the final mounting position for the control cable bracket. With Bernie’s help holding the bracket under the instrument panel base, I verified there was enough clearance for my fingers. Once the general position was established, a piece of masking tape was placed alongside the bracket to mark its position. Next measurements were taken based on the piece of tape to verify the rivet holes that needed to be drilled would not go into the flange on the F-1202H-L canopy rib or into the AV-5000A Control module. Now certain the mounting holes for the bracket would miss the areas of concern, the holes in the bracket were match drilled into the instrument panel base and the bracket was secured with Clecos.  Because the control cable is more or less in alignment with a manufacturing tooling hole already in the firewall that was previously filled with Proseal, rather than drill another hole, decided to just scrape away the Proseal and run the cable through the existing 3/16" tooling hole. During the final instillation, a silicone grommet will be installed in the firewall so the tooling hole will need to be drilled slightly larger to accommodate the grommet … but for the trial fit, the Bowden cable was just slid through the tooling hole after the Proseal was removed.

Final shape for the control cable’s under the panel mounting bracket.

The “Bender baffle’s” control knob’s final mounting position under the instrument panel base with the cable running through a tooling hole in the firewall. The control knob is tucked under the instrument panel base so it is out of the way when not in use … which will be the majority of the time.


With the control cable/bracket assembly Clecoed in place, Bernie gave me a hand installing the lower cowl so the firewall forward positioning of the control cable could be established. Once again, the control arm was temporarily clamped onto the baffle shaft so the overall range of movement could be verified. Not wanting to cut the Bowden cable just yet, decided to just make a loop at the end of the cable and slip it over the cotton swab stick shown in the previous post (Part 6).  It was at this point that I created a problem for myself which ultimately lead to my remaking the control arm. While testing there appeared to be some cable play or slop (the cable was still 6 foot at this point) so I drilled the mounting hole based on the amount of slop so there would be full range of motion. However, as it turned out, after the control cable was cut shorter to its final length, virtually all the slop went away and the full range of motion occurred long before the 3" pull of the knob (think it was 2 3/8") so decided, screw it, I’ll scrap the control arm … but not before drilling a second hole for the B Nut at the 1 7/8" point where previous measurements determined it should be. This worked out great, so I fabricated another control arm and also changed the design as well.

Test fitting of the control cable to the old control arm. Note: the hole for the B Nut was incorrectly drilled based on the slop that occurred when the cable was at its full length. After this photo was taken, another hole was drilled at the 1 7/8" point and all was well with the overall baffle movement. So a new control arm was fabricated.


The silver coupler that can be seen in the above photo will be used to mount an Adel clamp that will be used to secure the control cable. Not sure of the exact position for the coupler yet, but it will likely be installed approximately in the vicinity it is currently standing. The plan is to place another Adel clamp under the coupler which will secure the wiring bundle that will run to the voltage regulator that is going to be installed onto the baffle plate.


Knowing the trial fit using the actual control cable would work, decided to fabricate another control arm. This time the control arm was fabricated from a flat piece of .063 aluminum that was hand bent using seaming pliers to form a 90 degree angle. Next the control arm was shaped to remove excess material and matched drilled to the two #19 holes that were VERY carefully drilled in the baffle shaft. Next the mounting hole for the B Nut was drilled in the control arm at a point 1 7/8" out from the center of the baffle shaft. The lower cowl was set in place and the overall  baffle operation was tested using the new control arm. The control knob in the cockpit was able to easily move the baffle from stop to stop within its 3" range of motion.

Fitting the newly designed control arm onto the baffle shaft. Note the change in design from the last photo in the previous post (Part 6). Originally, the angle was made with a hole in so it could slide over the baffle shaft and be secured with one #10 bolt …  the thought was if the bolt became a little lose, the hole would prevent the control arm from flopping down and possibly getting jammed.  That idea was dropped shortly after the hole for the baffle shaft was drilled. Decided the better way would be to use two #8 screws to secure the control arm onto the baffle shaft.

Final test fit of the remade control arm with the baffle in the full open position … the B Nut is located at a point on the control arm 1 7/8" from the center of the baffle shaft. This allowed for full baffle movement from stop to stop with just under a 3" pull of the cable knob (think it was around 2 7/8"). This is just what I was hoping for because the control knob has detents that end at the 3" point … so although the control knob is being operated near its limit, it is still within a normal range of movement. I'm holding the cable on the coupler that will be used as a cable support post.

Final test fit of the remade control arm with the baffle in the fully closed position.

With the baffle control components now completed, figure may as well go ahead and prime all the baffle parts as well ... especially since everything else on the RV-12 is primed . I also have some other parts for upcoming changes that also need prepped and primed ... so the next couple of days will be spent readying a batch of parts to make a priming session worthwhile.

Installing a “Bender Baffle” Piggybacked With Regulator Relocation Mod - Part Six

This post will cover the majority of the fabrication of the control hardware that will make the baffle plate assembly inside the air duct move by pulling a knob from inside the cockpit. To adjust the baffle position, a Bowden style cable was chosen. Decided to use an A-740 control cable that has built in friction detents … this is the same control cable that was installed for the parking for the parking brake control.

The A-740 control cable that will be installed in the cockpit to control the “Bender baffle”. The control has a 3" range of motion.


Quite a bit of consideration was given to where the control for the baffle should be located within the cockpit, taking into consideration ease of installation and ease of use. Almost immediately dismissed the idea of mounting the baffle control knob in the center console area along with the other controls because of the way the control cable would need to be run on the forward side of the firewall. Not say’ in it can’t or shouldn’t be done, but it would require multiple bends of the cable and figured it easier if a straighter approach was incorporated.


As previously mentioned, when deciding on a mounting location for the baffle control ease of installation and operation were high on the list but I also desired the control to be out of the way. The location finally decided upon to mount the baffle control knob is to the far left and under the instrument panel base. This location creates a couple of challenges but overall is straight forward. The plan is to tuck the control as far left as practical and still be able to get fingers around the knob but also forward enough so the entire control knob is under the instrument panel when the baffle is fully open (which will be the norm 95% of the time). This will eliminate the possibility of snagging it with a shoe or knee during cockpit ingress and egress. Taking a photo while holding the control in position from outside the cockpit proved very difficult, so I have doctored an older photo and added a red arrow in the photo to show the approximate mounting location for the knob.

The red arrow in this photo points to the general location the control knob for the “Bender baffle” will be located. The plan is to tuck the knob a little forward under the instrument panel so it is completely out of the way when not in use.


The above mounting location presents a couple of small challenges that fellow builders considering this location need to be cognizant of when designing the mounting bracket for the control knob. The first issue is not to drill into the edge of the  flange on the F-1202H-L canopy rib (although if one does decide to do so, it would be OK as long as the hole ends up in the center of the flange … I just opted to avoid it altogether). The second is not to drill into the AV-50000A Control Module. Even with the mentioned concerns, there is still plenty of room to rivet a bracket onto the bottom of the instrument panel base and avoid conflicts with the surrounding items. To avoid the flange of the F-1202H canopy rib, decided to make a mounting bracket with an offset which will allow the control knob to be mounted further to the left yet keeps the rivets used to mount the bracket clear of the flange of the flange on the F-1204H-L canopy rib. Below is a photo of the unfinished bracket that shows the offset which will allow the control knob to be further to the left.

Proof of concept bracket for the “Bender baffle” control cable made from .063" aluminum sheet. The trial fit looks promising although do plan on remove some more material from the bracket to trim it down a little more.


Prior to working on the control arm for the baffle, figured it best to establish hard point stops for the baffle so the overall rotational angles can be determined. This was done by making marks denoting the desired range of motion and using some tracing paper to make a template for cutting a piece of aluminum. Decided to make both stops from one piece of aluminum which complicated the fabrication a little. First an Avery hand seaming tool was used to make a measured bend in a scrap piece of aluminum which I discovered required 3/32" of material to make the bend. Knowing this, I made two marks on the aluminum one where I wanted the stop to be placed and another 3/32" back from the desired line which is where the seeming tool jaws were placed. The fit of the resulting bracket was spot on if I must say so myself …love it when a plan comes together the first try.

The hard stops for the “Bender baffle” are made from one piece of aluminum and meet the baffle plate perfectly square at both ends. Decided to go this route as opposed to making two stops and then needing to drill 4 holes. This bracket uses the rivet holes already drilled in the upper doubler plate. The stop is wider than it needs to be so will likely trim some more material away.

Baffle plate in the full open position on the stop.

Baffle plate in the full closed position on the stop.

Now onto making the baffle assembly swing. For the control arm, decided to use a piece of angle which I drilled a 1/2" hole into so it could be slid over the baffle’s shaft. I did play around with trying to twist a piece aluminum bar stock but think it may have been too thick because I was getting stress marks so decided to scrap that idea. Eventually, two screws will secure the control arm onto the baffle shaft but for the mockup the holes were left undrilled and the control arm material was left longer than necessary.

Drilling the 1/2" hole in angle stock that will become the control arm so it can be slipped over the baffle shaft.

Quite a while ago I did the math and figured I would need approximately a 2" control arm attached to the baffle’s pivot shaft to make a 90 degree swing using a 3" pull. However, had to use that data as a reference point because as it turns out, the baffle actually swings a little beyond 90 degrees. So I mocked up a control arm and placed a cotton swab stick at the 2" point from centerline of the baffle shaft and measured how much pull was necessary to go stop to stop. This was done using a piece of waxed cord slipped over the swab stick and using a ruler to measure the pull distance. To do that, the cord was run through the cowl hinge eyelets so the measuring point would not move. The 2" point was close … but finally decided upon using 1 7/8" which seems to nicely move the baffle from stop to stop using slightly under a 3" pull.

The mockup to determine the necessary length for the control arm. Looking closely at the photo, one can see the waxed cord attached to a cotton swab stick (hard to see but it is under the right clamp) placed 2" from center of the baffle shaft. After a little testing, decided to move the stick to the 1 7/8" point which allowed for a full range of baffle motion from stop to stop with a 3" pull on the cord.


At this point, the cowl was placed in position to verify the clearances between the baffle control arm and the Rotax 912’s valve cover, spark plug wire and the three hoses in the vicinity. Looking at the following three photos below one can see the clearances are acceptable … plus the control arm will be shortened a little more because for this trial a fair amount of excess material was left beyond where the hole will be drilled for the control cable attachment point.

Squaring up the baffle control arm and the cotton swab stick for a range of motion test.

Baffle control arm position with the baffle in the fully open position.

Baffle control arm position with the baffle in the fully closed position.

Satisfied the control arm created no interference with nearby objects throughout its full range of motion, the next step was to fabricate a mounting pad that will become the control cable’s mounting point. Once made, the mounting pad was temporarily taped in position to check the overall fit. The “B” nut for the cable end was placed on top of the pad just to verify there were still no clearance issues, even though it is actually higher than what the final position will be when the “B” nut is placed in the mounting hole that will be drilled in the pad.

The mounting pad for the control cable temporarily taped onto the control arm for a test fit. The “B” nut for the end of the control cable was just set on the pad upside down to verify clearances. Later a 3/16" hole will be drilled into the pad at a point 1 7/8" from the center of the baffle shaft so the “B” nut can be inserted into the hole and secured with a cotter pin.


At this point the mounting pad was riveted onto the control arm. Of note: The pad for the “B” nut mounting is still larger than it needs to be. This was done on purpose because the cotter pin or quick release clip used to secure the shaft of the “B” nut will also need to have enough clearance so it does not hit the body of the control arm as it swings through its range of motion.


Return from the future: The mounting pad on the control arm shown below was  miss-drilled  and needed to be remade ... in doing so, the design was changed to a more simplistic version. The new design and reason for the miss-drilled hole are covered in the post  Part 7.

Apologize for the poor lighting it was almost dark when this photo was taken ... would have used the flash, but the camera battery was almost dead. Completed control arm with the mounting pad riveted in place. The mounting pad will be trimmed down more based on the final location of the mounting hole and cotter pin clearances.


The next work session will involve mounting the control knob bracket and securing the Bowden cable.