Hello, What is the proper way to simulate stub sytem as shown in the photo in LTSPICE with T-line?
Thanks.

 

Not sure if this is the proper way, but below is the sim for an open stub with a 1/4 cycle 3GHz stub-length delay (83.33ps), using the ideal transmission line model:
R2 and R3 were added to eliminate a floating-node error in LTspice.

If you change the input frequency, the output will increase, showing the cancellation is maximum at 3GHZ for the given stub-length delay.

The output spike is apparently due to the startup transient from suddenly applying the input signal.

 

Hello crutshow Ihave another implementation shown below ,why do we plug gnd on both sides of T1?

Thanks.

 

why do we plug gnd on both sides of T1?

It makes it appear as an unbalanced transmission line, such as a coax cable.

 

The trace would show a lot more detail if the interval shown was shorter. Like show possibly ten or 15 ns instead of 100.

 

Hello crutshow Ihave another implementation shown below ,why do we plug gnd on both sides of T1?

S-parameters do not change if both sides are grounded or not.