Omar,
I have had no luck again getting any sync errors to occur on Brad's machine, even by loading up the processor.  I even put in a 33 MHz Radisys (vs. 100 MHz) and it did not crash.  Hugh Sparks thinks there is something on the VME bus that is not within spec on data transmission.  It could even be the GRES-II or ADDA boards.  Since digital i/o can ascerbate the problem on-site, it could be the GRES board.

Both Hugh and I believe that until we put a bus analyzer on the NTT VME we won't know exactly what's happening.

NTT had the sync errors occur again during press day, so the gain will probably go up on this problem.

I am going to try to find some 497's to hook up to Brad's 498; we only need the station controllers (the module type is unimportant) to answer to 2 chassis numbers.  So far, there are not many 497's around in-plant at the moment.  The only console w/ 2 497s is the B600 and that project is hot now cuz of pending customer visit.

I think we have to crash this problem somehow and it might take drastic measures.  The 2 systems that have this problem are the NIED S/V controller and NTT, no other.  It almost points to hardware somehow, because I have found that CPU loading is just as high on HUD and MOC.  Also, I don't think that you or your jobs should carry the full burden of this problem cuz I do not believe it is application related.

Let me know what you think.  I will continue to pursue a way to hook up 497s but I am not holding my breath.
Rod Larson
