PPS fix determination vs. cgps?

[Steve Bourland]

I have a very strange case running with one of my Garmin 18 LVC units. 
Once it is locked in, both cgps and xgps that it has a 3D fix, and all 
seems just fine, but won't feed the PPS signal to ntpd with the following 
in the debug message:

gpsd:INFO: PPS:/dev/ttyS0 Assert hooks called clock:  1611094941.139482499 
real:  1611094941.000000000: no fix

I have tried to sift through the source code to figure out how the PPS 
determines the fix status vs. how cgps and/or xgps determine the fix 
status, to better understand why I am getting conflicting results.  The 
device is sending NMEA sentences GPRMC, GPGGA, GPGSA, and GPGSV.

Is it easy to point me to which files in the source I should be looking 
at to get this figured out?
                                        Thanks,
                                          Steve