Voltage Regulator Failure

B&C specialties.

BandC.com

AVC1 voltage regulator.

plug-in replacement for the Ducati.

Over voltage protection.

Adjustable voltage.

State of the art replacement.

Hi Robert,

Thanks for your recommendation.

How does the B&C Advanced Voltage controller (MOSFET) compare with the CARMO CARR5115 MOSFET?

What is the advantage of B&C voltage control/adjust?😈

This doesn’t answer your question but may be an option. 

It’s been my experience the 965349 regulator fail from vibration. The potting compound is not rigid, think of the PCB as floating in jello (not sure if jello is a thing in Australia,) the PCB is held to the case by the eight wire leads of the rectifier and SCRs. The typical failure mode is a low voltage alarm at low and moderate RPMs (half wave operation) or no output at all. Typically it starts as an intermittent problem that worsens over time.

Usually these regulators are easily repairable. The potting compound can be removed cleanly with simple tools and minimal effort. The picture of potting compound removal sequence spanned less that 10 minutes from start to finish, but I’ve had a little practice. Inspect the eight locations in the second photo for broken connections and re-solder as needed. The SCR leads may look ok, inspect them carefully, often they can be lifted off the board with your finger nail. The rectifier leads should have a cone of solder up from the board to the cut lead, what you usually find is a crater of solder with the disconnected lead coming through the center. If a lead falls off as you solder it, that’s an indication the failure point was at the body of the component and it won’t be repairable without replacing that component, I’ve only seen this happen once. If you we’re successful and removed the potting compound in one one piece, you can resecure it with RTV,  after confirming it’s fixed. 

These pictures are 10-15 years old, since Rotax hasn’t updated the part number it’s unlikely any changes have been made or issues resolved. 

Can't comment on other regulators than the AVC1 that I bought to replace my Ducati that died at 300 hours. I have used B&C items in other aircraft, never any problems.

The AVC1 is a plug-in replacement, same mounting holes even. The adjustable voltage is for different battery types. I use EarthX lithium iron phosphate batteries that need 13.8 to 14.5 volts to best charge. I set my voltage at 14.2 and very satisfactory.

Do be mindful of really seating the connector into the regulator, need a solid click at both ends.

Sean, I can't directly answer your first question (how the B&C and Carmo compare), except that: a) both are shunt type regulators, and b) B&C have an excellent reputation here in the U.S. and I wouldn't hesitate to use one of their products.  That said, I have no idea how difficult or costly it would be to get their stuff shipped to Australia.

Carmo's tech info shows that a 20-min test at 20A caused a case temperature of 110°C.  That suggests MOSFET die temperatures at near their datasheet limits, which doesn't speak well for longevity.  Unfortunately, B&C don't publish similar test data, but the heatsinking on the AVC1 looks a lot more effective than on the CARR5115.

Re your second question, the voltage adjustment allows the installer to cater for different battery chemistries and achieve full battery charge without excessive float voltage.  In a lead-acid battery this minimizes out-gassing, and in a lithium battery, it reduces heat dissipation in the BMS cell-balancing circuit.  Basically, you can set whatever bus voltage is best for your specific battery.

----------

The 2013 post by "chanik" in your CT Flier forum link, above, is spot on: this application cries out for synchronous rectification and voltage control via a buck converter instead of a shunt regulator.  Efficiency would increase dramatically (reducing heat in the regulator components) and load on the stator windings would be roughly equal to demand instead of constantly running at its limit.

Take a look at the thermal camera images on pages 1 and 8 of this datasheet for a synchronous rectifier controller IC.  The difference in heat production versus a diode rectifier is dramatic (and this is just the rectification section; you still have to regulate output voltage).  Unfortunately, this rectifier solution is about seven times costlier than four diodes (~$12 vs. ~$1.70 at a build quantity of 100).  Add the higher parts count for a buck regulator (also not cheap at this current level) and you've got a significantly more expensive final product.  This is a hard nut to crack.