Voltage Regulator Failure

H Roger,

My last Rotax 912ULS powered aircraft (ATEC Zephyr), first flew in 1999, ran with a fairly consistent engine bay temperature of 10C above ambient.

From faulty memory, I acquired the above aircraft at about 300 (Hobbs) hrs. The VR, mounted in the engine compartment, was likly original (no replacement mentioned in the Log Book) - it failed at around 700 hrs. The aircraft passed to the next owner at 950 hrs with the replacement Ducati/Rotax VR still going strong. I guess a 700 hrs service life may be considered acceptable by some.

My current Rotax 912 ULS aircraft is a Sonex Legacy. Mindful of the effects of temperature on the VR,  l decided to located it on the the cockpit side of the firewall (footwell), with a dedicated blast/cooling tube, to further mitigate high temperatures.(I have not recorded the temperature in this area) As stated, it would seem to have failed at 153 hrs.

It does seems that RV12's have had a service life shortening effect on their VR's. In recognition of this, owners have been advised to relocate the VR to a cooler location (in the cockpit) and add a dedicated blast/cooling tube. It seem that this a helped BUT premature failures continue.

This Forum has pages & pages of conversations about VR failures There is an inherent problem unrelated to aircraft model/type. Rotax/Ducati have tried to address this within the original "shell"  - this may have helped BUT premature failures are still occurring.

Component failure can be mitigated by viewing the part as a service item ie replacement at regular intervals (determined by the engine manufacture) BUT then should be low cost OR the component made sufficiently robust so that failures are rare. Rotax seem to have done neither.

I have very little understanding of electronics, however it does seem that there is an aftermarket change to VR technology, with the advent of MOSFET,  which promises cooler running & may be higher vibration resistance, in a smaller package (lighter weight?)  and less cost  (in most cases) to a Rotax OM replacement - I live in hope.😈

Roger other than anecdotal observations, did you do any failure analysis to determine what exact component failed and how it failed?  Yes vans has had a lot of regulators fail, but there are a lot of RV12’s out there. I’ve done a failure analysis on many Rotax regulators and all fall into two unique failure modes based on the regulator version.

The newer regulators with black potting compound ALL fail from broken solder joints on the PCB. Please explain how it’s Van’s fault when a sealed potted device has internal lead separation. 

Older pre 2010 regulators with greenish potting compound fail when the diode’s unsolder themselves. The “button” type diode Is intended to be installed between two bus bars that hold the diode in compression, the bus bars also act as the heatsink. The Ducati design of soldering leads to a button case diode and burying it in potting compound is mind bogglingly stupid. 

You said van’s “has one of the highest failure rates”. This image is from a flight school that operated CT’s. It’s pretty obvious this is a Rotax problem not a Van’s problem. It’s also gone on way too long, over 25 years and counting. 

I used temperature strips on the VR, ignition modules and other electronics in the engine. Poor cowling airflow and VR placement caused temps to be above 170F. In my Flight Design I did that same test to check the temps and they stayed down around 140F 145F. VR location and cooling air flow is crucial. Van's RV12's used to mount the VR's right behind the #4 exhaust pipe within inches and on a hot metal shelf and they had put an air flow tube on them, but the way they mount it caused air flow failure. After I made a few simple alterations no one had any failures. Even inside an instrument panel it should have some airflow for cooling. The other thing that could cause damage is amperage draw.

Roger, nothing you’ve done identifies the failure point. I’ll ask again;

Please explain how it’s Van’s fault when a sealed potted device has internal lead separation?

Too much external heat.