Calibration method for carrying out multiport measurements on semiconductor wafers
View Patent ↗The invention relates to a method for calibrating a vectorial network analyzer having n measurement ports and at least 2n measurement locations (n>1) by successive measurement of the reflection and transmission parameters at different two-port calibration standards, which are connected between the measurement ports in any desired order and must all have a transmission path, and three different n-port calibration standards, which are connected between the measurement ports in any desired order and which are not permitted to show transmission and by calculation of error coefficient and scattering matrix [Sx] with the 10-term or 7-term multiport method. An object of the invention is to propagate a method for calibrating these vectorial network analyzer used for multiport measurement which permits a calibration with increased precision and considerable reproducibility of measurement.
1. A method for calibrating a vectorial network analyser having n measurement ports and at least 2n measurement locations (n>1), the method comprising:
successive measurement of the reflection and transmission parameters at k=sum (n−1) for (i=1, 2, . . . , n−1) or n−1 different two-port calibration standards, which are connected between the measurement ports in any desired order and all have a transmission path, and three different n-port calibration standards, which are connected between the measurement ports in any desired order and are permitted to have no transmission;
mathematical determination of error coefficients of the network analyser with the 10-term method in the k-fold application and measured two-port calibration standards; and by mathematical determination of the scattering matrix [S x ], in which the errors are corrected, with the 10-term method,
wherein the method further comprises:
a) carrying out the first k calibration measurements at a two-port, which is realized by means of the direct connection of the measurement ports (through connection, T=Thru) or a short matched line (L=Line) of known length and attenuation, and which is connected between each of the k possible measurement port combinations,
b) carrying out a further calibration measurement at an n-fold one-port (n-one-port), which is realized by means of n-known wave terminations (M=Match), wherein the n-known wave terminations have unrestricted values so that they can be different from each other,
c) carrying out a further calibration measurement at an n-one-port, which is realized by means of n-unknown greatly reflective terminations (R=Reflect), which are similar to short circuits (S=Short),
d) carrying out a further calibration measurement at an n-one-port, which is realized by means of n-unknown greatly reflective terminations (R=Reflect) which are similar to open circuits (O=Open) and
e) mathematically determining the reflection accounts of the n-one-ports that are realized by means of n-unknown greatly reflective terminations which are similar to open circuits or to short circuits.
2. The method for calibrating a vectorial network analyser according to claim 1 , wherein
a) n>2 holds true,
b) the further calibration measurement is carried out at a one-port, which is realized by means of a known wave termination (M=Match), instead of at a n-one-port, which is realized by means of n-known wave terminations.
3. The method for calibrating a vectorial network analyser claim 1 wherein one of the greatly reflective terminations is known.
4. A method for calibrating a vectorial network analyser having n measurement ports and at least 2n measurement locations (n>1), the method comprising:
successive measurement of the reflection and transmission parameters at n−1 different two-port calibration standards, which are connected between the measurement ports in any desired order and all have a transmission path, and three different n-port calibration standards, which are connected between the measurement ports in any desired order and are permitted to have no transmission;
mathematical determination of error coefficients of the network analyser with the 7-term method in the n-1-fold application and measured two-port calibration standards; and mathematical determination of the scattering matrix [S x ], in which the errors are corrected, with the 7-term method,
wherein the method further comprises:
a) carrying out the first n−1 calibration measurements at a two-port, which is realized by means of the direct connection of the measurement ports (through connection, T=Thru) or a short matched line (L=Line) of known length and attenuation, and which is connected between a reference measurement port and the remaining ports (n−1),
b) carrying out a further calibration measurement is carried out at an n-one-port, which is realized by means of n-known wave terminations (M=Match), wherein the n-known wave terminations have unrestricted values so that they can be different from each other,
c) carrying out a further calibration measurement at an n-one-port, which is realized by means of n-unknown greatly reflective terminations (R=Reflect), which are similar to short circuits (S=Short),
d) carrying out a further calibration measurement at an n-one-port, which is realized by means of n-unknown greatly reflective terminations (R=Reflect) which are similar to open circuits (O=Open) and
e) mathematically determining the reflection accounts of the n-one-ports that are realized by means of n-unknown greatly reflective terminations which are similar to open circuits or to short circuits.
5. The method for calibrating a vectorial network analyser according to claim 4 wherein
a) n>2 holds true,
b) the further calibration measurement is carried out at a one-port, which is realized by means of a known wave termination (M=Match), instead of at a n-one-port, which is realized by means of n-known wave terminations.
6. The method for calibrating a vectorial network analyser according to claim 4 wherein one of the greatly reflective terminations is known.
7. The method for calibrating a vectorial network analyser according to claim 1 or 4 , wherein the further calibration measurement is carried out at a (n-i)-one-port, wherein i<n, which is realized by a known wave termination (M=Match), instead of at a n-one-port, which is realized by means of n-known wave terminations.