I reported a successful repair of my 42PF732A on 8-29-2009 in my post
http://www.avsforum.com/avs-vb/showt...0#post17083910
The problem started to re-appear in June or July of this year. At first it was very rare and would just flash for a few seconds. It gradually became worse. Lately, sometimes the bottom would go black, occasionally the top would go black, and even more rarely it would go completely black. Tapping on the right side, as viewed form the front would cause the problem to come and go.
I finally broke down and repeated the re-seating of the 2 driver cards. I used the same procedure I did before.
In my case, a Mexico assembled 42PF7320A dated September 2005, there is no packing material behind the 2 drive cards. There is a connector between them, and each is connected to the main board by 2 connectors, a black one with a lift up latch and a white one with no latch. I did not separate the 2 cards. I carefully worked them out together, and re-seated them. Actually I did that twice.
When I got it set back up, tapping had no effect. The picture was solid. It looks like I have bought myself some more time.
If I can continue to get a couple of years out of a re-seating, I guess I'll keep on doing it. However, I will not spend any money on the set. I just cannot justify that to myself, considering the current cost of a unit of similar size and features.
If these units did not use matched connector parts, I would almost suspect metal migration. I used to work as a test engineer for Western Electric. The Bell Labs metallurgists who supported our Bell Labs designers were always worried about components getting inserted in sockets with different plating. We were required to only put Gold leaded parts in Gold plated sockets and Tin leaded parts in Tin plated sockets. When they got mixed, there was fear of metal migration, when low currents passed through the connections for a long time. Eventually, the quality of the connection would deteriorate. The contamination layer was extremely thin, and re-seating would temporarily solve it, but that was not an acceptable repair. The factory was required to replace both the socket and the component with parts that had the same finish.
I do suspect that the plating on one of the connector sets is of poor quality. Since the connectors seem to be used in multiple versions of the set, that would explain why the problem seems to exist in different versions of the Philips TVs that were probably assembled in different countries. I believe Sam's Club had units from Mexico and Europe when I got mine. Actually, the final assembly point may not be relevant since all of the modules may have been assembled in the same plant and both of the connector types probably came from single suppliers.
FOLLOW-UP
It has been a few days, and all is well.
I am optimistic that I have bought myself a year+ of freedom from the 1/2 black screen. It took about 2 years for the problem to originally show itself on extremely rare occasions. I had the set for almost 4 years when it got bad enough that I did the first re-seat of the drive cards. That got me well over a year without any problems, and 2 years until it got bad enough that I did the second re-seat. I guess I can live with re-seating the drive boards at one year plus intervals until the set actally dies.