Sonnet User's Guide - Sonnet Software

Sonnet User’s Guide
Generating files matchnet.lib and matchnet_predict.snp.
--Model Log for matchnet--
Data set has 1246 points and 2 ports
--Model Options--
Error threshold(%) = 0.5 Error Threshold Curve Fit
Output predicted file: C:\Program Files\sonnet\project\matchnet_predict.snp
Data File
Max order target: 200
Warning Messages
--Model Results--
S11 Order = 203 Error(%) = 1.387911 WARNING: Error threshold
of 0.5 (%) not achieved
S12 Order = 208 Error(%) = 2.984112 WARNING: Error threshold
of 0.5 (%) not achieved
WARNING: Poor figure of merit on S12 parameter. Visual inspection
of predicted S12 recommended.
S22 Order = 214 Error(%) = 1.833756 WARNING: Error threshold
of 0.5 (%) not achieved
WARNING: Model prediction is not passive at 112 frequency points.
Error threshold may need to be decreased or input data may be non
passive.
--Model Summary for matchnet--
Maximum error was for S12, Error(%) = 2.98411
Total model time: 25 minutes 43 seconds
Model matchnet finished with no errors, 5 warnings
Indicates that a “fit” was found
for all S-parameters but for some
the error exceeded the error
Improving the Accuracy of the Broadband Spice Model
If you need to increase the accuracy of your Broadband Spice model, there are
several strategies you may use.
• If the Broadband model met your error threshold criteria but is still not
acceptable, you may decrease the error threshold to increase the
accuracy of the model. Be aware, however, that the processing time may
be significantly increased by lowering the error threshold. Typically,
values below 0.1% result in unacceptably long analysis times.
• If there are more than 200 frequency points in your response data, try
decreasing the number of frequencies in your response data. To do so,
use the Analysis ⇒ Clean Data command in the project editor to
remove the response data, then run another Adaptive sweep (ABS)
326