Narrative:

Our flight, a widebody transport non-stop jfk to sfo, was cleared 'to sfo via the kennedy 3 departure, bridge climb, gayel intersection as filed, departed runway 31L.' it was the copilot's leg and he set the ADF to 414 KHZ, the bridge NDB. I asked if he desired bridge to be depicted on the RNAV as a backup, but he declined saying, 'no, I'll just home on bridge.' as a backup he programmed RNAV fixes and radials depicting the cri 039 degree right and the jfk 253 degree right as aids to determine turn restrictions and the altitude crossing restriction as depicted on the bridge climb. During my cockpit preflight I had checked the ADF utilizing a local broadcast station and it seemed to perform properly. When I had attempted to check it on a jfk NDB frequency, it had been weak and intermittent. This did ot concern me since adfs are often victimized by interference on low frequency navigation ranges due to the proximity of ground vehs, terminals, etc. Since it had functioned fine on the broadcast station, I wasn't concerned about it. After takeoff the first officer climbed straight ahead to about 500-1000', and then began to home on the #2 needle on the RMI. While it didn't look exactly right to me at first glance, it was pointing left and he had begun a turn to the left as I expected. For the next 30 seconds I was absorbed in contacting departure control, setting climb mode, retracting flaps and completing the after takeoff checklist. When I once again directed my attention to the HSI and RMI I could see that, although the PF was still in a shallow bank toward the double needle, the needle itself was still in the same relative position as when he had initiated the turn. I also noted that he was approaching the cri 039 degree right and would almost certainly cross it prior to completing a turn southwest-ward. I simultaneously began an attempt to identify bridge and also called departure and asked if bridge was showing on the air. His reply was approximately, 'turn left 090 degrees, this'll be a vector to gayel intersection.' we did cross the 039 degree right during the completion of the turn. Departure then queried jfk tower about the status of bridge and informed us that it checked normal. I completed the time consuming task of returning the NDB and found I was unable to get a good identifier, although there was a carrier, and the pointer warning flags were out of view. I further investigated the ADF and learned that although it worked fine on commercial broadcast stations, it was entirely inoperative on lower hertz range stations. Compounding the problem was the fact that the warning flag would retract any time there was a discernable carrier, notwithstanding the fact that the needle was not homing on any particular station. Thus, on our departure, the needle was apparently pointing at a random bearing about our 10 O'clock position and simply stayed there throughout the turn. This led the PF to believe that he was making reasonable progress toward the fix, and he never initiated a large enough rate of turn to avoid crossing the cri radial. In addition, once all these actions are accomplished and you identify the station (assuming it and the radio are functioning) the aircraft lacks a loop function which allows the needle to be swung off its position. The only positive way to ensure that the needle is actively seeking the station is to loop it off its existing position (which may be random) and then return to the range position and watch the needle return. In all other aircraft I've flown this was a normal part of the ADF tuning procedure. On the widebody transport this test is not possible. Departure procedure shouldn't utilize a low frequency NDB such as bridge. The company should make any facility which is part of a complex procedure part of the data base for ease of entry. We didn't put bridge into the RNAV largely because it wasn't in the data base and neither of us were instantly familiar with the proper format for a latitude/long entry since neither of us had ever accomplished such an entry.

Google
 

Original NASA ASRS Text

Title: ACR WDB TRACK HEADING DEVIATION DURING SID FROM JFK.

Narrative: OUR FLT, A WDB NON-STOP JFK TO SFO, WAS CLRED 'TO SFO VIA THE KENNEDY 3 DEP, BRIDGE CLB, GAYEL INTXN AS FILED, DEPARTED RWY 31L.' IT WAS THE COPLT'S LEG AND HE SET THE ADF TO 414 KHZ, THE BRIDGE NDB. I ASKED IF HE DESIRED BRIDGE TO BE DEPICTED ON THE RNAV AS A BACKUP, BUT HE DECLINED SAYING, 'NO, I'LL JUST HOME ON BRIDGE.' AS A BACKUP HE PROGRAMMED RNAV FIXES AND RADIALS DEPICTING THE CRI 039 DEG R AND THE JFK 253 DEG R AS AIDS TO DETERMINE TURN RESTRICTIONS AND THE ALT XING RESTRICTION AS DEPICTED ON THE BRIDGE CLB. DURING MY COCKPIT PREFLT I HAD CHKED THE ADF UTILIZING A LCL BROADCAST STATION AND IT SEEMED TO PERFORM PROPERLY. WHEN I HAD ATTEMPTED TO CHK IT ON A JFK NDB FREQ, IT HAD BEEN WEAK AND INTERMITTENT. THIS DID OT CONCERN ME SINCE ADFS ARE OFTEN VICTIMIZED BY INTERFERENCE ON LOW FREQ NAV RANGES DUE TO THE PROX OF GND VEHS, TERMINALS, ETC. SINCE IT HAD FUNCTIONED FINE ON THE BROADCAST STATION, I WASN'T CONCERNED ABOUT IT. AFTER TKOF THE F/O CLBED STRAIGHT AHEAD TO ABOUT 500-1000', AND THEN BEGAN TO HOME ON THE #2 NEEDLE ON THE RMI. WHILE IT DIDN'T LOOK EXACTLY RIGHT TO ME AT FIRST GLANCE, IT WAS POINTING LEFT AND HE HAD BEGUN A TURN TO THE LEFT AS I EXPECTED. FOR THE NEXT 30 SECS I WAS ABSORBED IN CONTACTING DEP CTL, SETTING CLB MODE, RETRACTING FLAPS AND COMPLETING THE AFTER TKOF CHKLIST. WHEN I ONCE AGAIN DIRECTED MY ATTN TO THE HSI AND RMI I COULD SEE THAT, ALTHOUGH THE PF WAS STILL IN A SHALLOW BANK TOWARD THE DOUBLE NEEDLE, THE NEEDLE ITSELF WAS STILL IN THE SAME RELATIVE POS AS WHEN HE HAD INITIATED THE TURN. I ALSO NOTED THAT HE WAS APCHING THE CRI 039 DEG R AND WOULD ALMOST CERTAINLY CROSS IT PRIOR TO COMPLETING A TURN SW-WARD. I SIMULTANEOUSLY BEGAN AN ATTEMPT TO IDENT BRIDGE AND ALSO CALLED DEP AND ASKED IF BRIDGE WAS SHOWING ON THE AIR. HIS REPLY WAS APPROX, 'TURN LEFT 090 DEGS, THIS'LL BE A VECTOR TO GAYEL INTXN.' WE DID CROSS THE 039 DEG R DURING THE COMPLETION OF THE TURN. DEP THEN QUERIED JFK TWR ABOUT THE STATUS OF BRIDGE AND INFORMED US THAT IT CHKED NORMAL. I COMPLETED THE TIME CONSUMING TASK OF RETURNING THE NDB AND FOUND I WAS UNABLE TO GET A GOOD IDENTIFIER, ALTHOUGH THERE WAS A CARRIER, AND THE POINTER WARNING FLAGS WERE OUT OF VIEW. I FURTHER INVESTIGATED THE ADF AND LEARNED THAT ALTHOUGH IT WORKED FINE ON COMMERCIAL BROADCAST STATIONS, IT WAS ENTIRELY INOP ON LOWER HERTZ RANGE STATIONS. COMPOUNDING THE PROB WAS THE FACT THAT THE WARNING FLAG WOULD RETRACT ANY TIME THERE WAS A DISCERNABLE CARRIER, NOTWITHSTANDING THE FACT THAT THE NEEDLE WAS NOT HOMING ON ANY PARTICULAR STATION. THUS, ON OUR DEP, THE NEEDLE WAS APPARENTLY POINTING AT A RANDOM BEARING ABOUT OUR 10 O'CLOCK POS AND SIMPLY STAYED THERE THROUGHOUT THE TURN. THIS LED THE PF TO BELIEVE THAT HE WAS MAKING REASONABLE PROGRESS TOWARD THE FIX, AND HE NEVER INITIATED A LARGE ENOUGH RATE OF TURN TO AVOID XING THE CRI RADIAL. IN ADDITION, ONCE ALL THESE ACTIONS ARE ACCOMPLISHED AND YOU IDENT THE STATION (ASSUMING IT AND THE RADIO ARE FUNCTIONING) THE ACFT LACKS A LOOP FUNCTION WHICH ALLOWS THE NEEDLE TO BE SWUNG OFF ITS POS. THE ONLY POSITIVE WAY TO ENSURE THAT THE NEEDLE IS ACTIVELY SEEKING THE STATION IS TO LOOP IT OFF ITS EXISTING POS (WHICH MAY BE RANDOM) AND THEN RETURN TO THE RANGE POS AND WATCH THE NEEDLE RETURN. IN ALL OTHER ACFT I'VE FLOWN THIS WAS A NORMAL PART OF THE ADF TUNING PROC. ON THE WDB THIS TEST IS NOT POSSIBLE. DEP PROC SHOULDN'T UTILIZE A LOW FREQ NDB SUCH AS BRIDGE. THE COMPANY SHOULD MAKE ANY FAC WHICH IS PART OF A COMPLEX PROC PART OF THE DATA BASE FOR EASE OF ENTRY. WE DIDN'T PUT BRIDGE INTO THE RNAV LARGELY BECAUSE IT WASN'T IN THE DATA BASE AND NEITHER OF US WERE INSTANTLY FAMILIAR WITH THE PROPER FORMAT FOR A LAT/LONG ENTRY SINCE NEITHER OF US HAD EVER ACCOMPLISHED SUCH AN ENTRY.

Data retrieved from NASA's ASRS site as of July 2007 and automatically converted to unabbreviated mixed upper/lowercase text. This report is for informational purposes with no guarantee of accuracy. See NASA's ASRS site for official report.