What was found
SHORTLY AFTER DEPARTURE THE CAPTAIN NOTICED THE COCKPIT LIGHTING OSCILLATING BETWEEN NORMAL AND SLIGHTLY DIMMER. THE CAPTAIN TURNED THE STROBES OFF AND THEN BACK ON TO SEE IF THIS WAS THE SOURCE. NOTHING CHANGED. HE THEN TURNED THE ROTATING BEACON OFF IN WHICH THERE WAS NO CORRECTION. WHEN THE ROTATING BEACON WAS TURNED BACK ON, ALL POWER WAS LOST TO THE NUMBER 1. DUAL FEED BUS. THE CAPTAIN DID AN IMMEDIATE RETURN TO THE MODESTO AIRPORT WITHOUT ANY INCIDENT. COMMENCED TROUBLESHOOTING BY COMPLETING A TEXTRON DUAL BUS FEEDER - INSPECTION CHECK. ALL CURRENT LIMITERS WERE FUNCTIONAL. RESULTS LED TO A SUSPICION OF A PROBLEM WITH THE DUAL BUS NO. 1 & 2 DIODE PANEL ASSEMBLY. UPON VIEWING THE PANEL, IT WAS IMMEDIATELY APPARENT THAT WIRES WERE BURNT AT THE PANEL ASSEMBLY. THE BURNED WIRES ARE 6 GAUGE IN SIZE. THEY ARE PROTECTED BY 60-AMP CURRENT LIMITERS. AGAIN, THESE CURRENT LIMITERS WERE INTACT YET THERE WAS EXTENSIVE WIRE BURNING. FEEDER WIRE P5A6, WHICH FEEDS POWER TO DIODE CR2 WAS IN HARD CONTACT WITH DIODE CR1. BURN MARKS ON THE WIRE INDICATE HEAVY CURRENT FROM WIRE P5AG TO DIODE CR1 AT THE POINT OF CHAFING. PLASTIC INSULATING HAT SECTIONS AND THIN INSULATORS BELONG UNDERNEATH EACH DIODE TO ELECTRICALLY ISOLATE THE DIODES FROM THE METAL HEAT SINK. BOTH HAT SECTIONS WERE COMPLETELY BURNED AWAY. THIS PROVIDED AN OPPORTUNITY FOR THE DIODES TO SHORT TO THE HEAT SINK. THAT CLEARLY HAD HAPPENED. WIRES P2A6 AND P6A6, NOT READILY VISIBLE DURING NORMAL PHASE INSPECTIONS WERE HIGHLY BURNT. THE WIRES WERE NUMBERED WITH OLDER MASKING TAPE SO WORK WAS COMPLETED ON THEM AT SOME PREVIOUS AND UNDETERMINED TIME. THE LOCK WASHERS FOUND ON THE DIODE ATTACHMENT HARDWARE WERE THE STAR WASHER TYPE AND DID NOT HAVE A PLAIN WASHER BETWEEN THE LOCK WASHER AND THE THIN INSULATING WASHER. THE LOCK WASHERS ON THE DIODE ATTACHMENT HARDWARE SHOULD BE THE SPLIT LOCK WASHER TYPE. A TEXTRON CUSTOMER SERVICES ENGINEER PROVIDED US DRAWINGS TO PROPERLY INSTALL THE DIODES. THOSE DRAWINGS INDICATE A SPLIT LOCK WASHER SHOULD BE USED (PN MS35338-44) AGAINST A THIN PLAIN WASHER THEN THE INSULATING WASHER. WE BELIEVE IT IS LIKELY THE LOCK-WASHER COULD HAVE CUT THROUGH THE INSULATION AND PROVIDED CONTINUITY TO THE HEAT SINK. THE BEECH MANUALS DO NOT PROVIDE A BREAKDOWN OF WHAT PROPER ATTACHMENT HARDWARE IS REQUIRED FOR THE DIODES. DIODE CR1 WAS FOUND TO BE ELECTRICALLY OPEN AND DIODE CR2 WE BELIEVE HAD INTERMITTENT CONTINUITY AS IT WAS FOUND BROKEN ON THE INPUT SIDE THAT THE WIRE P5A6 ATTACHES TO. HOWEVER, BEFORE WE DISTURBED IT, IT PROVIDED SUFFICIENT CONTINUITY DURING THE DUAL BUS FEEDER - INSPECTION CHECK. THE ROTATING BEACON WAS IN A FAILED CONDITION. OUR THOUGHTS ARE THAT THIS WAS EITHER COINCIDENTAL TO THE INCIDENT OR AN ELECTRICAL CURRENT FOR THE BEACON WAS A "STRAW THAT BROKE THE CAMEL'S BACK." THE ACTUAL HEAT SINK OF THE RECTIFIER ASSEMBLY IS NORMALLY NEITHER ELECTRICALLY POWERED OR A GROUNDED PART. NYLON WASHERS UNDER THE HEAT SINK ATTACHMENT SCREWS PROVIDE ELECTRICAL ISOLATION. HOWEVER, WITH THE LIKELY CUT DIODE INSULATORS AND / OR WIRE P5A6 CHAFFING HARD AGAINST DIODE CR1, THE HEAT-SINK BECAME CHARGED WITH BUS POWER. WITH THE HEAT SINK AT 28 VOLTS, WE BELIEVE AN AIRCRAFT WOULD STILL OPERATE NORMALLY, THOUGH IT WOULD LACK THE INTENDED PROTECTION OF THE DUAL BUS FEEDER DESIGN. FOR UN-DETERMINED REASONS, IT APPEARS THAT THE HEAT SINK SHORTED TO THE FRAME (FRAME PN 101-364053-5) THAT IT MOUNTS TO AND PROVIDED A PATH TO GROUND. THE SPACE BETWEEN THE FRAME AND THE HEAT SINK IS JUST SUFFICIENT ENOUGH FOR A DASH 4 WASHER TO SNUGLY FIT AND PROVIDE AN ELECTRICAL PATH FROM THE HEAT SINK TO ITS MOUNTING FRAME. THOUGH NO BURN MARKS ARE EVIDENT BETWEEN THE HEAT SINK AND THE MOUNTING FRAME, THIS TYPE OF HARDWARE WAS FOUND ON THE FUSELAGE FLOOR BELOW THE HEAT SINK ASSEMBLY AND ITS MOUNT. IN ADDITION, THE FORWARD OUTBOARD MOUNTING AREA FOR THE FRAME TO FUSELAGE STRUCTURE HAS SLIGHT CARBON MARKS. THE FLOOR IS NOT READILY VISIBLE AS THE VIEW IS BLOCKED BY AIR-CONDITIONING DUCTS. ONE COULD ALMOST SAY WE WERE LUCKY TO FIND THE LOOSE HARDWARE. SO, THERE IS A PLAUSIBLE ELECTRICAL PATH TO GROUND TO PROVIDE A SHORT BUT IT IS NOT READILY APPARENT WHY A CURRENT LIMITER DID NOT BLOW TO PROTECT THE CIRCUIT. OUR THOUGHTS ARE THAT EACH INDIVIDUAL WIRE, P1A6 AND P5A6, THOUGH PROPERLY PROTECTED INDIVIDUALLY BY THEIR RESPECTIVE 60-AMP CURRENT LIMITER, HAD INADEQUATE PROTECTION FOR THE FOLLOWING REASON. THE SHORT TO EACH CIRCUIT AT THE DIODE PANEL ASSEMBLY RENDERED THE DEFECT TO BE PROTECTED BY BOTH CURRENT LIMITERS AT THE SAME TIME. IF THAT WAS THE CASE, THE WIRING IN THIS SITUATION HAD A 120-AMP LIMIT PROTECTION WITH TWO CURRENT LIMITERS ELECTRICALLY IN PARALLEL WHEN IT REALLY STILL NEEDED A 60-AMP LIMIT PROTECTION. THE WIRE COULD THEN BURN BEFORE THE CURRENT LIMITER OPENED. IT WOULD BE GOOD TO HAVE AN ELECTRICAL ENGINEER VERIFY THIS IDEA. THE FACT THAT -6 GAUGE WIRES BURNED AND NO CURRENT LIMITERS BLEW TO ISOLATE THE DEFECT IS THE REASON WE FEEL IT IS IMPORTANT TO SHARE THIS INFORMATION. WE HAVE THE FOLLOWING IDEAS TO MITIGATE A REOCCURRENCE OF THIS INCIDENT. ADD TO THE PHASE 3 INSPECTION, (MM 5-20-03, TABLE 603) TO PAY PARTICULAR ATTENTION TO WIRES ABOVE AND BELOW THE DUAL BUS FEEDER DIODE HEAT SINKS FOR DISCOLORATION INDICATIONS OF OVERHEATING. CURRENTLY THE PHASE 3 INSPECTION REQUIRES A DUAL BUS FEEDER - INSPECTION CHECK. ACCESS TO COMPLETE THAT CHECK IS FROM BELOW A CENTER ISLE FLOORBOARD. THE ACTUAL HEAT SINKS ARE UNDER THE COCKPIT FLOORS. I CAN'T SAY THAT I HAVE EVER LOOKED AT THE WIRES UNDER THE HEAT SINKS AS THIS WOULD EITHER REQUIRE A MIRROR OR A BORESCOPE DUE TO LIMITED ACCESS. REVISE TEXTRON'S MANUALS TO SHOW THE PROPER WAY TO INSTALL A REPLACEMENT DIODE. THE INFORMATION IS NOT AVAILABLE IN TEXTRON MANUALS. REVISE TEXTRON'S MANUALS TO ILLUSTRATE PROPER ADEQUATE WIRE CLEARANCE AT THE DIODES.REVISE THE HEAT SINK OR ITS MOUNTING PLATE TO MITIGATE THE POSSIBILITY OF HARDWARE OR FOD ENABLING AN ELECTRICAL PATH BETWEEN THE TWO PARTS. CHANGE THE DIODE DESIGN IF TODAY'S TECHNOLOGY HAS A LESS DELICATE DESIGN AVAILABLE.