What are the possible causes for a reflection coefficient greater than 1 (positive S11 in dB)?

Altair Forum User

What are the possible causes for a reflection coefficient greater than 1 (positive S11 in dB)?

JIF

In some cases S-parameters greater than 1 are physically correct, for example when testing a corporate feed on an antenna array with bad matching it could happen that one or more elements can have net power flowing back into the element(s) instead of being radiated.

In other cases the small electrical size of a model could cause an extremely bad match and numerical precision could cause the S11 to be slightly greater than 1.

We list some remedies to improve the results.

1. Electrically small models may require finer meshing.
2. Electrically small models may require the use of double precision.
3. Electrically small models could benefit from low frequency stabilisation.
4. Where very thin substrates are used, or where strong coupling exists between closely spaced surfaces, a mesh size of the order of the spacing between the surfaces (or the thickness of the substrate) may be required.
5. If possible, symmetry should be used.
6. If Surface Equivalence Principle (SEP) (surface meshing) is used for dielectrics, FEM/MoM (volume meshing) with first order basis functions (set under Solution Settings in CADFEKO) may be used, or Volume Equivalence Principle (VEP)