Crosstalk Modelling Of Coupled Microstrip Lines
This project present two improved variants of time-domain method for predicting crosstalk on parallel-coupled microstrip lines. The first method derives simple time-domain near-and far-end crosstalk expressions which are applicable to lossless case with significant harmonic frequency <1GHz. The e...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Published: |
2002
|
| Subjects: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | This project present two improved variants of time-domain method for predicting crosstalk on parallel-coupled microstrip lines. The first method derives simple time-domain near-and far-end crosstalk expressions which are applicable to lossless case with significant harmonic frequency <1GHz. The expressions are in polynomial form with geometrical dimensions of the structure and stimulus information as the only required entry parameters. They are simpler as compared to other methods because the difficult -to-determine distributed RLCG electrical parameters of the coupled lines are not needed. A look-up table for the polynomial coefficients is generated for easy application of this technique. In the second method, the FDTD method has been formulated to simulate crosstalk on parallel-coupled microstrip lines. Being a time-domain method, it is applicable to general mismatched linear/nonlinear terminations. An expedition feature by means of scaling the cross-sectional geometry of the structure has been incorporated into the FDTD simulation. This feature significantly reduces the simulation time of the numerical method. The espedition method is applicable up to about 1GHz where the propagation mode can be considered as quasi-TEM with negligible conductor loss and dielectric loss. For significant harmonic frequency > 1GHz, the dielectric loss can be incorporated into the FDTD simulation by using an average conductivity value for the substrate. Comparisons of the results with a number of published data and those generated from commercial software show good agreement. Crosstalk measurements using a sampling oscilloscope are also conducted to further validate the proposed methods. Again the comparison of results demonstrates good agreement between the predicted values and the measurements. The results show that the proposed methods have great potential for computer-aided-design (CAD) purposes. Comparisons of the results with a number of published data and those generated from commercial software show good agreement.
|
|---|