Follow along with the video below to see how to install our site as a web app on your home screen.
Note: This feature may not be available in some browsers.
The two artificial lines don't need to be 50ohms. The goal of these artificial lines is to provide a way for Sonnet to de-embed them, so your reference planes do not include the artificial lines. If you have metal leading up to where the artificial lines start, it will have a certain width. You should then make your artificial lines the same width as the lines leading up to them.
If you make them different widths, you introduce a discontinuity, and Sonnet may have a harder time de-embedding to give you the right answer.
Why does Z0 of a microstrip line depends on W but not L ?
Someone told me the above with regards to the two extra rectangles nearest to the ports.
Sonnet can remove the effect of the feedline at the port if reference plane shift is used. Sonnet then calculates the line properties and removes them from the total result.
That reference shift is then visible in the Sonnet editor, by arrows from the port up to the reference plane.
What do you exactly mean by reference plane shift ?
FTYP SONPROJ 13 ! Sonnet Project File
VER 15.53
DIM
ANG DEG
CAP PF
CON /OH
FREQ GHZ
IND NH
LNG MM
RES OH
END DIM
FREQ
ABS_ENTRY 0.001 6.0
END FREQ
CONTROL
STD
OPTIONS -bdj
SPEED 0
CACHE_ABS 1
TARG_ABS 300
Q_ACC N
END CONTROL
GEO
DRP1 LEFT LINK
POLY 13 1
0
DRP1 RIGHT LINK
POLY 15 1
0
TMET "Lossless" 0 SUP 0 0 0 0
BMET "Lossless" 0 SUP 0 0 0 0
MET "Copper" 1 NOR 58000000 1 0.01
BOX 1 36.1 12 722 240 20 0
20 1 1 0 0 0 0 "Air"
1.6 4.2 1 0.02 0 0 0 "FR-4"
LORGN 0 12 U
POR1 STD
POLY 13 1
2
1 50 0 0 0 0 8.78975
POR1 STD
POLY 15 1
2
2 50 0 0 0 36.1 8.790025
NUM 5
0 5 -1 N 11 1 1 100 100 0 0 0 Y
5.010005 8.689745
16.050005 8.689745
16.050005 8.889745
5.010005 8.889745
5.010005 8.689745
END
0 5 -1 N 12 1 1 100 100 0 0 0 Y
20.049995 8.690025
31.089995 8.690025
31.089995 8.890025
20.049995 8.890025
20.049995 8.690025
END
0 5 -1 N 13 1 1 100 100 0 0 0 Y
5.01 9.38975
5.01 8.18975
0 8.18975
0 9.38975
5.01 9.38975
END
0 5 -1 N 15 1 1 100 100 0 0 0 Y
31.09 8.19003
31.09 9.39002
36.1 9.39002
36.1 8.19003
31.09 8.19003
END
0 5 -1 N 17 1 1 100 100 0 0 0 Y
16.050005 3.109975
20.050005 3.109975
20.050005 8.889975
16.050005 8.889975
16.050005 3.109975
END
END GEO
OPT
MAX 100
VARS
END
END OPT
VARSWP
END VARSWP
QSG
IMPORT NO
EXTRA_METAL NO
UNITS YES
ALIGN NO
REF NO
VIEW_RES YES
METALS YES
USED NO
END QSG
TRANSLATOR
DXFEXPORT
UseTLs false
SepObj true
SepMat true
DivideMulti false
Circles false
CircleType inscribed
CircleSize 0
KeepMetals true
KeepVias true
KeepViaPads true
KeepBricks true
KeepEdgeVias true
KeepParent true
ConvertParent false
AllEdgeViasAsVia false
END
END TRANSLATOR
May I know how to generate a DXF layout file (without the extra feedline structure) for actual manufacturing purpose ?
is FR-4 with height 1.6mm and Erel of 4.2 common for chinese PCB manufacturers ?
For your PCB order, you need to include that backside ground metal in your files. One possibility: You can draw a polygon on the Sonnet ground layer. This has no effect for Sonnet, but will be included in export.
FTYP SONPROJ 13 ! Sonnet Project File
VER 15.53
DIM
ANG DEG
CAP PF
CON /OH
FREQ GHZ
IND NH
LNG MM
RES OH
END DIM
FREQ
ABS_ENTRY 0.001 6.0
END FREQ
CONTROL
STD
OPTIONS -bdj
SPEED 0
CACHE_ABS 1
TARG_ABS 300
Q_ACC N
END CONTROL
GEO
DRP1 LEFT LINK
POLY 13 1
0
DRP1 RIGHT LINK
POLY 15 1
0
TMET "Lossless" 0 SUP 0 0 0 0
BMET "Lossless" 0 SUP 0 0 0 0
MET "Copper" 1 NOR 58000000 1 0.01
BOX 1 36.1 12 722 240 20 0
20 1 1 0 0 0 0 "Air"
1.6 4.2 1 0.02 0 0 0 "FR-4"
LORGN 0 12 U
POR1 STD
POLY 13 1
2
1 50 0 0 0 0 8.28975
POR1 STD
POLY 15 1
2
2 50 0 0 0 36.1 8.290025
NUM 6
0 5 -1 N 11 1 1 100 100 0 0 0 Y
5.010005 8.189745
16.050005 8.189745
16.050005 8.389745
5.010005 8.389745
5.010005 8.189745
END
0 5 -1 N 12 1 1 100 100 0 0 0 Y
20.049995 8.190025
31.089995 8.190025
31.089995 8.390025
20.049995 8.390025
20.049995 8.190025
END
0 5 -1 N 13 1 1 100 100 0 0 0 Y
5.01 9.88977
5.01 6.68973
0 6.68973
0 9.88977
5.01 9.88977
END
0 5 -1 N 15 1 1 100 100 0 0 0 Y
31.09 6.69002
31.09 9.89003
36.1 9.89003
36.1 6.69002
31.09 6.69002
END
0 5 -1 N 17 1 1 100 100 0 0 0 Y
16.050005 2.609975
20.050005 2.609975
20.050005 8.389975
16.050005 8.389975
16.050005 2.609975
END
1 5 -1 N 46 1 1 100 100 0 0 0 Y
0 0
36.1 0
36.1 12
0 12
0 0
END
END GEO
OPT
MAX 100
VARS
END
END OPT
VARSWP
END VARSWP
QSG
IMPORT NO
EXTRA_METAL NO
UNITS YES
ALIGN NO
REF NO
VIEW_RES YES
METALS YES
USED NO
END QSG
TRANSLATOR
DXFEXPORT
UseTLs false
SepObj true
SepMat true
DivideMulti false
Circles false
CircleType inscribed
CircleSize 0
KeepMetals true
KeepVias true
KeepViaPads true
KeepBricks true
KeepEdgeVias true
KeepParent true
ConvertParent false
AllEdgeViasAsVia false
END
END TRANSLATOR
I could not see the ground plane using 3D viewer of Kicad PCBNew software.
The open-circuited stub microstrip realization method is a modification of the stepped-impedance microstrip realization method. It approximates the series inductance as a high impedance transmission line while the shunt capacitance effect is simulated by an open-circuited stub using Richard’s transformation.
I am trying to understand why open-circuited stub has a better S21 response compared to stepped impedance according to
It seems that the "open circuited stub" layout used a length of lambda/4 at 5.6 GHz
Is it possible to prove this mathematically using some microstrip expressions ?
The wavelength is multiplied by 1/sqrt(dielectric constant); this does not hold perfectly true in microstrip because of the differing even-/odd- mode propagation velocities due to the substrate below the strip and the air (or free space) above having different dielectric constants.
Someone else told me the above.
Doing the math with the limited information that you supplied the S21 phase that I get is more like ~120° using ADS Linecalc,
Could you post the ADS linecalc window screenshot which shows 120° ?