Thursday, January 14, 2016

The Ducati Voltage Regulator - A Dilemma Ensues

With the weather being nasty, figured it was a good time to take my vehicle in for maintenance. As a result, I’ve been stranded at the house so decided the time was right to discuss the Ducati voltage regulator. This post does not apply to builders building E-LSA RV-12's, as you are obligated to follow construction plans during the build. The post is long and perhaps somewhat controversial … so it may be time to bust out the popcorn.

The charging system in the Rotax 912 is unique in that, instead of being a typical generator, it is a dynamo … similar to that used on many motorcycle charging systems. As such, the charging system requires a different type of voltage regulator than typically found on aircraft or cars … one that is single phase and has a built in rectifier accompanying the regulator. Rotax has chosen the Ducati voltage regulator for the job. Unfortunately, this voltage regulator has a history and it is not all good. There’s been an abundance of failures in the RV-12 … some at very low hours of usage. Also of note, the failures are not just isolated to the RV-12 … other LSA aircraft using the Rotax 912 engine have experienced regulator failures as well. The Ducati voltage regulator supplied by Rotax has gone through at least two manufacturing changes (that I’m aware of) over the years … noted by a change in part numbers.
The Ducati 965-347 voltage regulator currently supplied by Rotax for the 912 ULS engine.

Van’s originally mounted the Ducati voltage regulator on the RV-12’s firewall shelf. Admittedly, the firewall shelf is not an ideal mounting location because the regulator is near the #4 exhaust pipe and is also in the direct path of the heated air exiting the coolant radiator.
A photo from another builder’s site showing the original mounting position for the Ducati voltage regulator on the RV-12’s firewall shelf.

The first change came about when Van’s began offering a lighting kit for the RV-12, a design change was incorporated to cope with the extra heat generated by the increased electrical load. Van’s added a cover over the top of the regulator allowing a cool air blast tube to be ducted from the engine’s FF-1207 cooling shroud in the hopes it would improve the regulator’s reliability by keeping the regulator cooler.
Photo taken by a builder in Utah shows the addition of the cooling shroud over the Ducati regulator. Cool air is ducted via a blast tube coming from the FF-1207 cooling shroud.

Even with the additional cooling, the Ducati voltage regulators have continued to fail … the regulators either get flaky, begin generating noise heard in the headsets, become intermittent, stop working all together or, in a few cases, have literally gone up in smoke. Some of the issues have been spade connector related, but the majority of them are not. Feeling the root of the problem is heat related, the most recent change incorporated by Van’s involves moving the Ducati regulator inside the cockpit and mounting it under the instrument panel near one of the fan ports.
The new home selected by Van’s for the Ducati voltage regulator … it is currently to be mounted behind the firewall under the instrument panel base near a fan inlet.

Unfortunately, there’s already been at least one reported regulator failure that I’m aware of within 60 hours on a Ducati regulator that has been located exclusively at the new mounting location inside the cockpit under the instrument panel base. This suggests to me there is much more to the failures than ambient heat alone … perhaps vibration, but I suspect the real issue is Ducati design/manufacturing related. Inquisitive builders have opened up several of the failed Ducati regulators and posted shocking photos of numerous bad solder joints discovered on the circuit board inside the Ducati regulator.
A collage of photos taken of the failed solder joints inside multiple Ducati regulators.
Inside the circle on left is a lead that was lifted from the location circled on the right … where it should have been solidly soldered down on the circuit board.

Whether these failures were caused by excessive internal heat melting the solder, vibration, poor design/manufacturing or a combination thereof, is still unclear. Some have reported that they have experienced the regulator failing and then oddly, it begins working again then after being flaky for a while fails altogether. This may lend credence to the theory the solders are melting and when the regulator cools, the solder hardens creating an electrical connection again.

A study of cooling: In the meantime, a savvy builder from North Carolina took the initiative to made real time measurements on the air flow available from the FF-1207 air shroud. Testing revealed that because the air shroud receives its air from near the root of the propeller, the airflow actually delivered through the blast tube to the Ducati regulator is VERY minimal during ground operations. An actual test was made by placing the air duct tubing into a glass of water with the engine running at a RPM typical for ground operations … the test reveled there was not enough air pressure coming from the FF-1207 cooling shroud to create air bubbles in the water glass (even if the tube was barely in the water).  A similar test was made by ducting air from the air duct supplying air to the coolant radiator ... which is part of the lower cowling. The air duct in the lower cowl receives its air from behind a larger blade area of the propeller and created enough air pressure to make bubbles in the glass of water … indicating there is much greater air pressure available to cool the regulator during ground operations from this duct as opposed to the air available from the FF-1207 cooling shroud.

A solution the builder in North Carolina came up with for cooling his replacement Ducati regulator is mounting it inside the air duct ahead of the radiator … there is plenty of cool air at this location and no hot air blasting the regulator from either a nearby #4 exhaust pipe or the outflow of air from the coolant radiator. This area also remains much cooler after engine shutdown.
Photo of an alternative mounting location a builder in North Carolina is experimenting with for the Ducati voltage regulator … it is located inside the air duct in the lower cowl just in front of the coolant radiator. Reportedly, the regulator temperature has remained under 70 c after over 60 hours of flying based on a temperature strip attached to the regulator.

In addition to looking into better ways of cooling the Ducati regulator, some builders have also experimented with alternatives … such as John Deer or NAPA regulators with early success but the verdict is still out regarding long term usage.  Another regulator that looks VERY promising for use in the RV-12 based on the specifications and actual usage in aircraft is the German built Silent Hektik F4112. It is being used as a replacement for the Ducati regulator on quite a few aircraft flying with Rotax 912 engines in Europe … however, unfortunately, it is only sold in Europe.  The Silent-Hektik controller (they call it a controller and not a voltage regulator) claims to be rated for 37 amps continuous and peaks of 47 amps, which is twice the rated continuous output of the Ducati regulator and Rotax’s dynamo. Also, the voltage output of the Silent-Hektik is 14.2 volts which is higher than the Ducati’s 13.8 volts and more inline with, arguably, the proper voltage the RV-12’s Odyssey 12 volt AGM battery should be charged with in the first place.

The DOG Aviation take on all this: Not at all fond of Van’s decision to install a high current device known to have fault issues inside the cockpit … should the regulator fail and go up in smoke (and they have), the toxic smoke will fill the inside of the cockpit. Also not thrilled with the apparent “under designed” Ducati regulator supplied by Rotax.

Return from the future: Fortunately before wiring up the Silent Hektik  I noticed a small difference in a couple of the drawings at the Silent Hektik site. Fortunately, I had the presence of mind to filter the Silent Hektik WEB pages through a German to English translator. As mentioned when purchasing the regulator through a motorcycle shop in England, I was surprised to receive a version with wires coming out of the regulator even though the unit was the same part number as a unit that has spade leads ( of which I had a photo with the exact same part number). Apparently, there were variants at the time the order was placed. To my understanding, all the units with wires coming out of the regulator are wired differently …. and NOT what we want as a replacement for the Ducati regulator.

Builders beware!!! Here is the BIG BIG difference. On the standard Ducati regulator reading the terminals from left to right you have G G R +B L C.  On the Silent Hektik F-4112 WITH WIRES, the terminals are marked as  G G R –B L C.  Yet the Silent Hektik unit that has the same F-4112 part number but with the spade connectors like the Ducati unit the terminals are marked as G G R +B L C.

The big difference is where the regulator receives it ground from …. The F-4112 unit WITH WIRES gets it ground from the –B lead (blue) which needs to be connected to the battery's ground terminal … however, the RV-12 wiring configuration connects this lead to the battery's positive terminal, not good!!! Yet, the regulator version with the spade leads uses the regulator case as the ground and the +B terminal is connected to the battery's positive terminal, as it is in our application.

I’m assuming this must have created quite a bit of confusion which is likely 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 the replacement for the Ducati regulator on  Rotax912/914 engines. … the F-1418 regulator IS the one we want to install as the direct plugin replacement for the Ducati regulator. The leads are marked as G G R +B L C which is what we want. The F-4118 regulator also has an added feature of a four stepped voltage output ... so it can be used not only on lead acid batteries, but also on ion batteries as well. 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 and NOT -B. Just be sure not to use one of the regulators with the wires because it is not pin for pin compatible with the Ducati regulator.

DOG Aviation will attempt a preemptive strike … the procurement department has taken delivery of a Silent Hektik F4112 controller ordered from Europe for installation in the DOG Aviation RV-12. In addition, an attempt will be made to devise a better way to keep the controller cooled during ground operations. I did not know it at the time of ordering, but it appears the Silent Hektik F4112 controller WAS available in two styles … one with spade connectors coming out of the controller and one with wires (yet the part number is the same for both … go figure). All the photos of the F4112 Silent Hektik controller I had viewed had spade connectors on the controller, so I was quite surprised to discover the one I custom ordered from a motorcycle shop in England had wires exiting the controller … however, this may be beneficial because I may move the controller to a different mounting location. When questioning the vender about receiving a unit with wires as opposed to the spade terminals, the manufacturer was contacted and I was told the reply was the Silent Hektik controller was not available with the spade connectors. Not sure if that means they were between production runs or it will no longer be offered with spade connectors.
The Silent Hektik F4112 controller ordered from Germany through a motorcycle shop in England … once received in England, the controller was shipped to DOG Aviation. Prior to instillation, it will be necessary to change the motorcycle style bullet connectors on the ends of the wires.

One can see by comparing the first photo in this post of the Ducati regulator to the photo below that the built in heat sink fins on the Silent Hektik F-4112 controller are of a much better quality than the anemic cooling fins on the Ducati regulator. The Silent Hektik controller is designed to dissipate heat … as evidenced by two more cooling fins and the depth of each cooling fin being almost twice that of the Ducati regulator.
Side view of the Silent Hektik controller shows the number and depth of the heat sink fins.

Plus, the base of the Silent Hektik is thick solid machined aluminum … so coupled with some heat sink compound, the mounting will be capable of transferring much more heat away from the controller compared to the Ducati’s potted resin base … which, frankly offers very little in the way of heat transferring ability.
Base of the Silent Hektik unit is machined aluminum which is good for transferring heat away from the controller when used in conjunction with a thin film of heat sink compound.

The DOG Aviation R&D department has not finalized the mounting location for the regulator (controller), but it certainly WILL NOT be inside the cockpit. If the mounting location remains on the firewall shelf, the cooling scat tube will come from the cool air supply for the coolant radiator and NOT the FF-1207 cooling shroud. The reason for this is during ground operations the air duct for the radiator catches far more air from the propeller blades and thus has more air pressure when compared to the root of the propeller where the air is taken for the FF-1207 cooling shroud. Another idea being kicked around involves mounting the regulator inside the air duct just ahead of the coolant radiator … as previously mentioned above (with photo), this has already been pioneered by one builder with very good results and thus far his regulator has not exceed 70 c after over 60 hours of flying based on a temperature strip attached to the regulator.

There are small downsides to mounting the regulator inside the air duct as it involves creating a jumper cable to extend the lengths of the wires to the regulator (controller) and, of course, with the regulator mounted inside the cooling duct, the electrical connector will need to be disconnected every time the lower cowling needs to be removed … somewhat of a pain but not a huge incontinence. Another downside and perhaps my real concern with mounting the unit at this location is the probability of the Silent Hektik regulator getting very wet if flying through rain. The unit does not appear to be waterproof because there are mounting screw holes that come through the top of the case at a couple of locations between the cooling fins, plus there is a seam all the way around the unit that does not appear to have a gasket. However, all that may be easily solved by sealing the seam with a thin bead of high temp silicone along with a few dollops on the screw hole locations between the cooling fins. Another concern is there is not a metal surface in the air duct to help draw the heat away from the regulator as there is on the firewall shelf … fiberglass is a very poor conductor of heat. However, with the regulator in plentiful airflow, this may not be a legitimate concern.

At this point, I’m inclined to mount the Silent Hektik controller at the original mounting location but improve the cooling by running the blast tube from the coolant radiator’s duct in the lower cowling and try creating some sort of a shield to help protect the regulator from the heat of the #4 exhaust pipe and outflow from the radiator. However, mounting the unit directly in the air duct is appealing on many levels so that thought is still being kicked around and evaluated.

Regardless of where the final mounting location ends up being, the barrel connectors on the ends of the wires coming from the Silent Hektik controller will need to be removed and changed to either spade connectors or ring connectors.

So the dilemma of the final mounting location for the regulator (controller) continues ... but one decision is firm, it WILL NOT be located in the cockpit area.