Saturday, July 22, 2017

Swinging The Backup PAI-700 Vertical Card Compass - A Failure - Airpath C2300 To The Rescue

The last remaining action item left to resolve on the DOG Aviation RV-12 was the Precision Aviation Inc. PAI-700 vertical card compass. The PAI-700 vertical card compass is installed as a backup just in case the Dynon Skyview develops a problem while in flight. I knew prior to going into it that mounting a vertical card compass in an instrument panel can be a tricky proposition, in that, panel mounting subjects the compass to magnetic interference from the surrounding instruments and wiring. From what I’ve read, apparently vertical card compasses are much harder to adjust when panel mounted compared to the typical aviation whiskey compass. As you will read below, many many hours were spent trying to get the PAI-700 compass to work correctly, but it was just not to be. Also, yes, I know there were hangars in the vicinity and the aircraft was not at a compass rose. But that said, I figured there was no point heading out to the compass rose until the backup compass can be made to display readings somewhat close to the Dynon SkyView's magnetometer (which was calibrated at the compass rose) .
The Precision Aviation Inc. PAI-700 vertical card compass.

Magnetic compasses used in aircraft have built-in compensation screws that are adjusted to correct for minor deviations in the magnetic field created by the instruments, wiring and surrounding structures. There is a compensation screw for making North/South deviations and another screw for making East/West deviations. To adjust the compass, the nose of the airplane is positioned towards the four cardinal directions and the appropriate compensation screw is turned to hopefully swing the compass to correct for heading errors. This adjustment or calibration process is called “swinging” the compass. The manufacturer suggests tapping the compass gently while positioning the aircraft which is what we always did for every directional change mentioned below.

Last fall the first attempt to swing the PAI-700 at the airport's compass rose resulted in a dismal failure, in that, the compass deviations were so far off the compensating screws could not correct the error ... the North/South readings were OK but the East/West readings were WAY off. Since this was a backup instrument anyway, swinging the compass was placed on the back burner, so to speak, to be taken up again the next time the forward top fuselage skin was removed. In the meantime the DOG Aviation procurement department purchased a couple of items that sometimes allow the PAI-700 compass to become calibrated or “swung” as it is called. Items purchased were a pair of balancing balls and a product called Mu-metal. The balancing balls are akin to those used on ships to compensate for all the metal in the bridge or wheelhouse surrounding the compass and can typically be seen on either side of the ship’s compass. The balancing balls contain magnets and rotating the balls one at a time helps compensate for the magnetic interference surrounding the compass. Mu-metal on the other hand is a specialized nickel alloy sheet that is made from a blend of 80% nickel and 20 % ferrous material which has been annealed. Mu-metal passes magnetic lines of force very easily so it can be used sometimes to “shield” unwanted magnetic fields from the compass.
The compensating balls for the PAI-700 vertical card compass.
Compensating balls installed on the back of the PAI-700 vertical card compass.
The compensating balls were installed first and although they helped, there was still not enough adjustment available on the internal compensators of the PAI-700 compass to be usable.

During the midst of fussing with the PAI-700 compass DOG Aviation was visited by the three Weiss guys from Michigan and their brother in law Carl from Indiana. Of course, I jumped at the opportunity of having slave labor available and quickly enlisted the help of Karl to help me position the aircraft at various headings using the Dynon SkyView's magnetometer then adjust the PAI-700. We spent quite a bit of time following the adjustment procedure supplied with the compensating balls only to find out after aligning the compensating balls as best as possible there was still not enough control using the compass’s internal compensating screws to make the necessary adjustments to swing the compass.
Left to right - Karl and Carl listen to an explanation of how easily the wings can be removed on the RV-12.
Left to right - Carl and long time friend Kurt looking at how the flaperon connects to the flaperon torque tube
Karl lending a hand calling out the Dynon SkyView compass reading while the aircraft was being positioned so the PAI-700 compass could be adjusted. After cooking Karl for about a half hour or so under the canopy in the hot sun, we discovered it made no difference in the compass readings if the canopy was left open so we just left it open.
Kurt looking on as brother Karl records the compass readings as we attempted to swing the PAI-700 vertical card compass. Unfortunately, the third Weiss guy, John, was manning the camera and didn't make it in any of the images. Thanks for all the photos John.

Once it was determined compensating balls alone were not enough to swing the PAI-700 compass, decided to begin playing with the Mu-metal. I though the majority of the magnetic interference was coming from a couple of nutplates adjacent to the compass but playing with the Mu-metal it became quickly apparent that was not the case … the major change came when the Mu-metal was slid between the compass and the vertical speed indicator directly below the compass, in that, the compass position moved at least 90 degrees when the Mu-metal was slid underneath the compass.. So with Karl’s help, it appeared the Mu-metal would work as the compass began to show hope that it could now be adjusted much closer than previous attempts. We ran out of time for more testing so the next day I glued the Mu-metal down onto the top of the vertical speed indicator and extended it across the top of the glove box (almost in the same position used during testing) using silicone RTV.
The Mu-metal was placed over the top of the vertical speed indicator and glued in position with RTV.  A water bottle, clamp and drill V blocks were used to weigh down the Mu-metal while the RTV cured.
The completed Mu-metal instillation makes a magnetic shield for the backup compass.

The following day, enlisted Bernie to help with the positioning of the aircraft and recording of the readings. Here again, had to start from scratch and first adjusted the compensating balls as best as possible then moved on to the internal compensating screws on the compass. At one point, it looked great for the cardinal headings but the in-between headings were still way off. Bernie and I fussed with it for a very long time and ran out of time before getting the issue fully resolved. But we did notice some flaky behavior, in that, once a heading was established it was not always repeatable …. As an example, once North was spot on turning to South may be say 7 degrees off ... the instructions say to use the adjuster and take out half the error. OK once that adjustment was made turning to north would yield a reading of say 2 or 3 degrees off … rotating back around to south could result in a reading that was 15 or 20 degrees off! By the way, as the airplane was being rotated around the compass was being lightly tapped as suggested to mitigate the internal drag from all the gears. (We also tapped prior to and after making an adjustment). I felt we were close and wanted to try tweaking the compensating balls a little more but that would have to wait until the next work session.
Bernie lending assistance to help with the swinging of the PAI-700 compass.

For the next work session enlisted the help of Jan to help with the positioning and recording the compass readings after readjusting the compensating balls and then the internal compensators. The readings were still flaky, so I decided to rip out the PAI-700 and just replace it with an Airpath whiskey compass shown below.
The Airpath C2300 whiskey compass.

With assistance from Jan, was able to swing the Airpath compass after just a few minutes of adjusting and it is within a couple of degrees for most all the headings. In order to swing the compass, the East/West compensation screw is turned full counterclockwise to its stop. That got me thinking, perhaps I don’t need the Mu-metal … so I removed the Mu-metal out of curiosity just to see if the compass could be swung and discovered the East/West readings were so far off they were not adjustable using the internal compensators … so I glued the Mu-metal back in place and now all is well again.
The back of the Airpath C2300 whiskey compass that will be replacing the seemingly unusable PAI-700 vertical card compass. Note the sheet of Mu-metal isolating the vertical speed indicator from the compass seems a little crinkled now that it was removed then reinstalled … it was necessary to keep the Mu-metal to allow the Airpath compass to be successfully swung.

Panel view of the new backup Airpath C2300 compass. Was able to be successfully swing the Airpath with no muss or fuss ... but it did require a Mu-metal shield. Still need to change out the mounting screws to match those used on the other instruments.

Of course, the Airpath C2300 compass was swung without the upper forward fuselage skin in place … so there may be a little more adjustment necessary once that is screwed in place. Have been using 8-18 stainless screws so they should have very little effect on the compass readings. But won’t know for sure until later in the week after returning from attending AirVenture in Oshkosh. Out of curiosity at some point in the future I may try to reinstall the PAI-700 again and see if it works better while actually flying … but I’m not very hopeful about that.


Sunday, July 2, 2017

Installing a “Bender Baffle” Piggybacked With Regulator Relocation Mod - Part Ten The Completion

Finally the “Bender Baffle modification and voltage regulator relocation onto the “Bender Baffle” is finally completed! Seems as though I have been turning this project into a career. Have not started the engine yet due to weather and performing some maintenance items such as changing the oil and filter, cleaning the gascolator and draining the fuel tank to get rid of all of the winter blended fuel so it can be replaced with the summer blend. If there are any issues with the regulator, I’ll return from the future and edit this post accordingly.


Fellow builders take note: Not to get off onto a tangent, but there is a very important safety point to be made here. The regulatory Gods have deemed it necessary to mandate auto fuels used in most areas of the USA have two formulations … a summer blend and a winter blend. The aspect of the two blends that affects RV-12 pilots the most is … winter blend auto fuel is formulated to vaporize easier in cold weather than the summer blend. The issue that affects us pilots using auto fuel in the Rotax 912 is that of vapor lock … in that, using a winter blend of auto fuel on warmer spring and summer days can and often times WILL lead to a situation where there is vapor lock. This can happen during takeoff and even while flying with increases in altitude because the reduction of air pressure allows the warmed winter fuel blend fuel to get to its vapor point. The best advice …. if you know you have a winter blend fuel left in the tank and are flying on days with summer temperatures, just drain the fuel and replace it with a summer blend if available. If the summer blend has not come out yet, it has been reported using at least a 30% blend of aviation 100LL will raise the vapor point of the auto fuel enough as to mitigate the chances of having a vapor lock issue.


I digress … this post will cover mounting the voltage regulator on the baffle plate and associated wiring along with the instillation of the control cable. As mentioned in a previous post, the Silent Hektik I originally received came with very flexible wires exiting the regulator as opposed to the spade lead connectors I was expecting …. so when drilling the mounting holes for the regulator, the regulator was purposely positioned quite high on the baffle plate so there would be plenty of wire exiting the air duct to play with. Unfortunately, now that I’ve discovered that the wired version of the 4112 Silent Hektik regulator is NOT a direct replacement for the Rotax regulator, it will be necessary to make my own wiring harness from 12 gauge Tefzel wire which is much stiffer than the wires that came on the wired regulator. This created a small issue due to the extra stiffness of the Tefzel wires and the fact the regulator mounting holes are high on the baffle plate … now there will be much less wire flex between the regulator and wire grommet as the baffle plate moves than originally planned for the wired version of the Silent Hektik regulator.


To iron out the issues created by using stiffer wire, a mockup harness was made from 12 gauge stranded house wiring (which is even stiffer than the Tefzel wires) so the range of motion and wire flex could be evaluated. The areas where the baffle swings were defined and marked in red and the open area available for mounting a wiring grommet was marked in green. A couple of hole locations were played with but it quickly became apparent having a hole close to the baffle shaft bushing was the best location. During the testing, because of the high mounting location for the regulator, it was also discovered the wiring will for sure require a loop to establish enough free wire movement to accommodate the baffles range of motion yet keep the wire movement at the grommet to a minimum. The plan is to move forward and use the already drilled mounting holes for the regulator and but keep a close eye on the wires. If doing it again from scratch with the F-4118 regulator or the Ducati regulator in mind, the regulator would have been mounted much lower on the baffle plate which may also eliminate the need for a loop in the stiffer Tefzel wires. Regardless, my advice to fellow builders considering this modification would be to mount the regulator lower on the baffle plate so there is more flex room above the regulator for the wires.
A trial fit of the regulator on the cardboard mockup to see how the wires act as the baffle plate is rotated so the placement for a wiring grommet could be located. The pie shaped wedge marked in green is the area where a wire grommet can be located. The vacant hole closest to the baffle shaft proved to be the better hole location for the wire grommet and generated the least amount wire movement in and out of the grommet.
Based on the cardboard mock-up, the wire grommet will need to be installed in the general area where it is sitting in this photo.


After the hole for the wire grommet was drilled and wire grommet installed, work began on installing the regulator onto the baffle plate. The regulator was mounted so the terminals face downward ... the thought being if there is a point in time when the regulator gets wet, water won’t sit in the connector. A connector housing and associated spade connectors (Silent Hektik part number F-4116) were ordered from Silent Hektik along with the F-4118 regulator. The wire used for wiring the spade connector pins was 12 gauge Tefzel except for the wire on the “C” terminal of the regulator … which received an 18 gauge wire which is way overkill size wise, but wanted the extra strength the 18 gauge offers since there will be some wire flex necessary when the baffle nears full closed.. The wires were placed in blue Taylor 2586 thermal protective sleeving which is rated to 1200 degrees so when the wire harness touches the warm regulator during a fully closed baffle condition, the wires should be nicely protected.
The Silent Hektik R-4118 regulator wired and ready to be installed into the RV-12 lower cowl’s air duct. One can see there are plastic tabs on the connector that was ordered with the Silent Hektik regulator. The tabs on the connector hook onto metal tabs on the regulator that were bent for a snug fit … the tabs make a positive lock so the connector should not back out on its own. Also note in addition to the wires going to the connector on the regulator, there is an additional black wire attached to the mounting bolt for the regulator… this wire supplies the necessary ground to the regulator’s case.


While experimenting with the cardboard mockup, it was discovered making a half loop where the wire harness exits the wire grommet in the cowl aided in giving the harness more flex making the baffle much easier to move throughout its range of motion. An Adel clamp will be installed adjacent to the wire grommet to keep the wire harness in the proper position. A second Adel clamp is installed along the aft edge of the frame for the radiator to secure and support the wire harness as it drops down to the firewall shelf to the Cinch terminal barrier strip.
The Silent Hektik R-4118 regulator/baffle plate assembly mounted in the lower cowl’s air duct.
Wire harness exiting the wire grommet with a half loop and the two Adel clamps used to secure the wire harness.

The wiring of the Cinch 6-142 terminal barrier block is straight forward. Metal spade lead adapters were used so the existing wiring on the RV-12 could be used without cutting off the spade connectors. My thought is should the Silent Hektik regulator develop a problem in the future, it should be an easy task to just move the wires from the Cinch terminal block back to the Ducati regulator. I was going to use 45 degree tabs for the spade connectors but discovered they would interfere with the cover plate I want to install over the top of the Cinch terminal block … so opted to install the flat version. Once the mounting tabs were installed on the Cinch terminal block, the wires were moved from the Ducati regulator to the terminal block. In addition, a 12 gauge ground wire was run from the firewall mounted ground buss to the terminal block … this will supply the necessary ground (mentioned above) to the case of the Silent Hektik regulator. The Silent Hektik WEB site recommends and sells a 33,000 MFD capacitor that should be used with the regulator … however, the RV-12 already has a 22,000 MFD capacitor so the DOG Aviation procurement department purchased a 10,000 MFD capacitor that is automotive rated to 275 degrees. The capacitor is installed between ground and the regulator’s R lead … however, on the RV-12, the R lead and B+ Bat are tied together in the wire harness so the capacitor’s + lead can be tied to either of those two wires. For those not familiar with electrolytic capacitors, the leads are polarity sensitive …. so extra care must be used when wiring so make sure the – lead runs to ground and the + lead on the capacitor is run to a positive voltage. Often times the capacitor will be marked with a black stripe or marking to denote which lead is negative. Sometimes one lead will be cut shorter, this typically denotes the negative lead …. That said it is best to refer to the manufactures application data sheet just to be sure because capacitors can pop like a firecracker if not wired correctly. Wires were soldered onto the leads of the additional 10,000 MFD capacitor to extend them so the capacitor could be securely mounted up and away from the Cinch terminal block. After soldering the extension leads onto the capacitor, the solder joints and body of the capacitor were covered with heat shrink tubing and the capacitor was wire tied onto the wire harness coming from the Rotax engine.


To finalize the wiring for the regulator’s wiring harness, all of the wires coming from the Silent Hektik regulator mounted on the baffle plate were terminated with ring terminals and attached to the Cinch terminal block. I placed yellow heat shrink on the two wires that go to the input to the regulator to correspond with the color of the wires coming from the Rotax engine’s stator coil assembly (which is very similar to that of a typical motorcycle). Red heat shrink was used to denote the R & B+ Bat leads from the Van’s wiring harness and I used blue heat shrink to identify the remaining C wire.
Completed wiring of the Cinch 6-142 terminal barrier block. One can see in this photo the two extra ring terminals that were placed on top of the ground and B + Bat lead on the terminal strip … these run to the additional 10,000 MFD capacitor.
The additional 10,000 MFD filter capacitor mounded onto the wire harness from the Rotax engine.

The two coupler nuts that can be seen in the previous photo will be used to secure a protective cover for the Cinch terminal block. The DOG Aviation procurement department was able to locate a high temperature plastic at McMaster-Carr that is not rigid like a hard plastic. It is Noryl PPO and has all the necessary traits for this application ... high heat resistance, good electrical properities and it is a slightly flexible material that should be well suited for this application.
Sheet of high temperature plastic that will be cut down to make a protective cover for the wiring attached to the Cinch terminal block.


After the wiring was completed, the control cable bracket was riveted onto the instrument panel base and the control cable was fed through a high temp silicone grommet on the firewall. The control cable was secured onto the lower cowl by using a high temp Adel clamp that was attached to the 10-32 coupler nut. Oversized washers were used above and below the fiberglass for added strength.
Finalized instillation of the baffle control cable under the instrument panel.
Baffle control cable firewall forward to the baffle control arm.
Close-up of the baffle control arm in the full open position.