The signal still went through, but effectively (time difference measured on scope) added 16ns onto the delay of T3.
Using the script I wrote for t3_delay... setting the "delay" to 0ns had given total delay of 58.4ns. After re-seating the patch cable, the total delay is 40ns.
I guess this only makes a difference for coincidence mode... and only
after the cable was not seated correctly.
A copy of this log entry has been emailed to: brads, kalyan