This is an intermittent problem that I have experienced with a a couple of my SIDSTAR files. I have just updated my CYLW (Kelowna, British Columbia, Canada) file to reflect the new runway orientation.On arrival to CYLW approximately 10nm prior to T/D, I get an FMC warning "APPROACH SPEED NOT SET" (or something to that effect). I do not receive a warning to "RESET MCP ALT". The aircraft passes T/D still level at the cruise altitude. A variation to this problem is a timely descent at T/D, but the aircraft slows to 210kts and does not follow the vertical descent path indicator.I have attached the file for your review. If you feel like trying my scenario, depart CYXX for CYLW. Set up for the ILS 16 (YDC transition). You will need Dave Hazell's BC navaids file to fly the ILS.Thanks,Roger P.

Roger,When you begin your descent, select Index from the bottom of the FMC, select the Approach Page and it will give you a few options of which approach speed to select based on weights and flap settings. This certainly works for me.

I tried the flight yet again, this time entering my approach speed as you suggested. I am still encountering some wild VNAV behavior at T/D. On the cruise page, the "distance to T/D" freezes at 11nm. The FMC does not ask me to "RESET MCP ALT". Approximately 10nm prior to T/D, I go ahead and reset the MCP altitude. The aircraft promptly commences a descent, well before T/D. (This is all shown on the attached .gif image.You may wonder why there are two magenta lines: the second line is the missed approach procedure, which brings the aircraft back to the YDC VOR.I've had this happen on other files of mine, and do not recall how I fixed them before. Possibly some bad code, or perhaps having a missed approach procedure that overlaps the route?The code is at the top of this thread. I'd like anyone's ideas. Roger P.

Hi Roger.1. Your FMC warning most likely says "APPROACH VREF NOT SELECTED". Dave pointed out the correct procedure.2. The second one for your descend is more complicated. I tried the flight many.many times with different settings. I also wrote new procedures with diff alt settings, deleted hard constraints, inserted soft constraints, no constraints at all, deleted the missed approach etc, but every time I get the same behaviour as you. I ran my tests without any speed restrictions.However I tried the procedure from another airport, this time KPSC, and then it works OK. Here the T/D point will be before YDC. Also the YKA transition works OK. I also tried to delete YDC and go direct to TOSUS, and then everything is OK. From your picture it seems that you use the -600/700 base pack and don't have -800/900 installed. I have the -800/900, but my test was done with a -600 VC, 167 NM trip, tried both FL 270 and FL 300. Fuel: 4800 lbs in wing tanks, 1000 lbs in center. Start weight 107824 lbs. Take-off from 07 at CYXX, also tried a flight with intsc YARRO - CILLI before entering the procedure at YDC. No weather add-on, but downloaded fs9 weather for yesterday.Here's my procedure setup: (Note that if you don't have the -800/900 installed, soft constraints will be ignored)add to FIXES section:FIX D14.9 LATLON N 50 12.72 W 119 24.51FIX D8.7 LATLON N 50 06.54 W 119 23.76FIX D5.2 LATLON N 50 03.05 W 119 23.34FIX B141 LATLON N 50 14.58 W 119 30.08add to APPROACHES section:APPROACH ILS.DME.16 FIX D14.9 AT OR ABOVE 5600 FIX D8.7 4400 FIX D5.2 3400 RNW 16 NAVAID EX KEEP TRK INTERCEPT RADIAL 218 TO FIX YWB FIX YWB TRK 211 UNTIL 9000 NAVAID YDC 9000 TRANSITION YDC NAVAID YDC AT OR ABOVE 9000 FIX TOSUS AT OR ABOVE 9000 FIX B141 AT OR ABOVE 7200 FIX WTMAN AT OR ABOVE 6300 The strange behaviour remain with this procedure too.This seems to be triggered by the T/D point just after a turn as here at YDC. There is an issue reported that the FMC under certain criterias may stop commanding pitch correctly, take look at this thread:http://forums.avsim.net/dcboard.phpaz=show_topic&forum=136&topic_id=64686&mode=full. This may apply to your scenario. Maybe Anthony Merton could run this scenario in the debugger for a test and see if this is the same.

Thank you so much for your reply, and thank you for testing it out. Now I know I'm not crazy. I will continue to experiment with my file.Roger

After several test flights I am able to define my problem a little more clearly. The VNAV behavior is abnormal when there is no mach transition during the descent (i.e. using a cruise altitude of anything less than about FL290.) As long as the aircraft gets to transition from a mach speed to an indicated speed, it's happy. In my hurry to test all of my approaches, I was flying very short trips to the destination -- trips with a cruise altitude of around FL250. I did not encounter any odd behavior when descending from altitudes above the mach transition.Roger