The biggest cause of variation is the test plug on the end of the test lead. From TIA/EIA-568-B.2-10 Draft 8.0 Category 6A we see"
"G.6 Category 6A measurement reproducibility
Measurement reproducibility is provided for test plugs. Measurement reproducibility for mated connecting hardware is dependent on the performance of test plugs. "
It is impossible to terminate RJ45 plugs and expect exactly the same performance each time.
The TIA therefore defines a band of acceptable performance for test plugs. For the critical pair 3,-6-4,5 this band can be drawn as:
To terminate an RJ45 plug and get it within this band is extremely difficult. There is a real skill in creating an RJ45 test plug. This causes a great problem for tester manufacturers making test leads. It gets worse. The only way to verify the performance of the terminated RJ45 plug is to cut it off the end of the patch cord (test lead), rendering it useless.
The only solution to this dilemma is to have a test lead where the test plug can be removed and tested independently.
By using a printed circuit board design, the electrical properties of the test plug are extremely consistent. How consistent?
What does this mean for you? Very consistent test results between DTX CableAnalyzers and their test leads. The primary reason why the leading cabling vendors and professionals prefer the DTX-PLA002 and DTX-1800 CableAnalyzer.
You can download a full copy of the electrical distribution for the DTX-PLA002PRP. Some users prefer Log frequency graphs, others prefer Linear frequency graphs. We are pleased to be able to offer both:
Caution: Extreme science content