Hi aa_talaat,
Your smith plot, using linear sample points, is similar to expected results using a Simsmith model. First straight line caused by very wide sample points.
SimSmith model is shown in:
Second SimSmith model in above link is performed with Log sample points to fill in low frequency range.
Ferrite material loss for this device, represented as parallel resistance, Rp, is 600 to 700 Ohms at high frequencies. Near constant parallel loss results with near constant S21 loss at high frequencies, as shown in your measured S21 trace.
John KN5L
