Today I conducted a fine timestamp test. The purpose of the test was to determine the relative fine timestamp error rate between two secure concentrators. The experimental setup consisted of two secure concentrators placed 30 feet apart. Both concentrator cards had GPS antennas attached and had a good GPS lock. I had a CubeCell programmed to send packets about once every second on various frequencies and spreading factors.
The results look really good. Here is an image with both log files side by side. Note the
gps_time fields have very good agreement. Here is the raw data log from the two concentrator cards.