StoppTidigare
Full Member level 2
- Joined
- Jul 16, 2002
- Messages
- 148
- Helped
- 1
- Reputation
- 2
- Reaction score
- 0
- Trophy points
- 1,296
- Location
- Northern earth
- Activity points
- 1,703
Hi I've been using the Smith-chart to match a load, with a series stub nearest the load to transform to g=1 circle, and then a shunt stub, the usual way, to cancel out the imaginary part.
Now argue with me somebody, because of I believe this:
This circuit is not reciprocal.
I mean, If you find a solution to matching problem with a series stub transmission line nearest the load to transform the load somewhere on the g=1 circle, and then a shunt stub to cancel out the imaginary part,
ONE CANNOT interchange the stubs.
I mean you cannot put the shunt stub first nearest to the load, and then the series line without modifying their lengths ?
Because:
If I plot my normalized load admittance on the Smith-chart y_load, and then add a shunt stub, I'll be walking up or down a constant conductance circle.(g-circle).If you then add the same series stubas in the old solution, this pont is rotated along a constant Gamma-circle(constant radius with respect to the point (1,0)).
This endpoint will not be the point (1,0).
So stubmatching is not reciprocal.
If you agree, you must then agree on that it is an error in the Gonzales book, first edition page 78. In page 76 he extracts the stub-matching network from the Smith-chart, for the input-matchcircuit, but he has the shunt stub nearest the generator. This is wrong.
Also the outputmatching network is wrong.It should start with a shunt stub, but it doesn't.
Do you agree ?
Kindest regards,
StoppTidigare
P.S. I've been following the instructions to design a stub-tuner network as written in "Field and wave electromagnetics" by David K.Cheng. There the metod start by determining the loadadmittance, therefore I used
y_load=2.8+J*1.9, instead of Gonzales y_s=2.8-j*1.9,for the conjugate match. I arrive at the same results in terms of series and shunt stub lengths.
Now argue with me somebody, because of I believe this:
This circuit is not reciprocal.
I mean, If you find a solution to matching problem with a series stub transmission line nearest the load to transform the load somewhere on the g=1 circle, and then a shunt stub to cancel out the imaginary part,
ONE CANNOT interchange the stubs.
I mean you cannot put the shunt stub first nearest to the load, and then the series line without modifying their lengths ?
Because:
If I plot my normalized load admittance on the Smith-chart y_load, and then add a shunt stub, I'll be walking up or down a constant conductance circle.(g-circle).If you then add the same series stubas in the old solution, this pont is rotated along a constant Gamma-circle(constant radius with respect to the point (1,0)).
This endpoint will not be the point (1,0).
So stubmatching is not reciprocal.
If you agree, you must then agree on that it is an error in the Gonzales book, first edition page 78. In page 76 he extracts the stub-matching network from the Smith-chart, for the input-matchcircuit, but he has the shunt stub nearest the generator. This is wrong.
Also the outputmatching network is wrong.It should start with a shunt stub, but it doesn't.
Do you agree ?
Kindest regards,
StoppTidigare
P.S. I've been following the instructions to design a stub-tuner network as written in "Field and wave electromagnetics" by David K.Cheng. There the metod start by determining the loadadmittance, therefore I used
y_load=2.8+J*1.9, instead of Gonzales y_s=2.8-j*1.9,for the conjugate match. I arrive at the same results in terms of series and shunt stub lengths.