Millimeter wave / terahertz frequency band high-consistency rapid de-embedding method

By obtaining S parameters at one time and using off-chip deembedding modules with multiple calibration algorithms, the measurement error and measurement cumbersome problems caused by different substrates in traditional methods are solved, and a high consistency and efficient deembedding process is achieved.

CN119001574APending Publication Date: 2024-11-22UNIV OF ELECTRONICS SCI & TECH OF CHINA
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411185153.8
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-27
Publication Date
2024-11-22

AI Technical Summary

Technical Problem

At high frequencies, the traditional deembedding method results in measurement errors due to different substrates, and multiple measurements are cumbersome and costly, and the measurement results are poorly consistent.

Method used

A high consistency fast deembedding method for millimeter wave/terahertz band is adopted to obtain S parameters through one measurement, and an off-chip deembedding module containing multiple calibration algorithms is used to realize the processing of different calibration algorithms, avoiding multiple measurements and The cumbersomeness of on-chip calibration.

Benefits of technology

While ensuring accuracy, it is possible to avoid poor consistency due to different positions of GSG probes, simplify the measurement process and reduce costs.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119001574A_ABST
    Figure CN119001574A_ABST
Patent Text Reader

Abstract

The invention discloses a millimeter wave / terahertz frequency band high-consistency rapid de-embedding method, and belongs to the technical field of semiconductor device modeling and test calibration, and the method specifically comprises the steps: employing an impedance standard substrate to calibrate a vector network analyzer; measuring S parameters of the to-be-measured device and the calibration device by using a vector network analyzer; converting the S parameters of the device to be tested and the calibration device into corresponding S2P data; constructing an off-chip de-embedding module, wherein the off-chip de-embedding module comprises a plurality of calibration algorithms and calibration device structure types required by the calibration algorithms; and inputting the S2P data of the device to be tested and the calibration device into the off-chip de-embedding module, selecting the S2P data of the calibration device required to be processed by each calibration algorithm, and after the S2P data is processed by the calibration algorithm, outputting a true value of the device to be tested, thereby realizing off-chip de-embedding of the device to be tested based on different calibration algorithms. According to the invention, use of various calibration algorithms can be realized by only one test, and the method has the advantages of high accuracy, good consistency and the like.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention belongs to the technical field of semiconductor device modeling and test calibration, and in particular relates to a high-consistency fast de-embedding method in a millimeter wave / terahertz frequency band. Background Art

[0002] The vector network analyzer is the most widely used test equipment in the field of RF engineering. When testing, the vector network analyzer must first be calibrated to remove the influence of system errors. However, when measuring a device, the test data is often affected by redundant structures such as ground-signal-ground (GSG) PAD and interconnection transmission lines. Therefore, in order to obtain the true response of the required device, it is necessary to move the test plane from the GSG PAD to both ends of the device and accurately remove the influence of the GSG PAD, interconnection transmission lines, etc., that is, de-embedding.

[0003] There are two traditional de-embedding methods. One is to first use a commercial impedance standard substrate for calibration, transfer the test plane to the probe tip, and then measure the device under test and the de-embedding device, and finally process the measured data (generally using open-short, L-2L (Line-to-Line) and other de-embedding algorithms) to complete the de-embedding. There is a problem with this method. When the operating frequency of the device under test is high, the electromagnetic environment is different due to the difference between the commercial impedance standard substrate (generally a quartz substrate) and the device under test and the de-embedding device (generally Si, InP, GaAs, GaN and other substrates). The de-embedding algorithm alone cannot eliminate the error introduced by the different substrates, which may lead to measurement errors.

[0004] In order to solve the problems existing in the first de-embedding method, a second de-embedding method has been proposed. This method first uses a commercial impedance standard substrate for calibration and transfers the reference plane to the probe tip. Unlike the first de-embedding method, this method performs a "second calibration" after completing the first calibration, that is, uses the vector network built-in calibration algorithm (such as SOLT (Short-Open-Load-Thru), TRL (Thru-Reflect-Line) and other algorithms) to calibrate the device under test and transfer the reference plane from the probe tip to both ends of the device under test. In this way, the true value of the device under test can be directly measured on the chip, thereby avoiding measurement errors caused by different substrates.

[0005] However, the second method still has the following problems: when performing the "second calibration", if a different calibration algorithm is required, it is necessary to measure the same device with the same structure multiple times, or measure different devices with the same structure. The process of repeated measurements is cumbersome and costly, and during multiple measurements, the measurement results of the device will be inconsistent due to the different locations of the GSG probe pressure points.

[0006] Therefore, it is necessary to develop a high consistency and fast de-embedding method in the millimeter wave / terahertz frequency band to solve the above problems. Summary of the invention

[0007] In view of the cumbersome on-chip calibration / de-embedding process of different calibration methods and the poor consistency caused by multiple tests, the present invention provides a high-consistency fast de-embedding method in the millimeter wave / terahertz frequency band. It can realize the use of multiple calibration algorithms with only one test and has the advantages of high accuracy and good consistency.

[0008] The technical solution adopted by the present invention is as follows:

[0009] A high consistency and fast de-embedding method in the millimeter wave / terahertz frequency band comprises the following steps:

[0010] Step 1: Use an impedance standard substrate to calibrate the vector network analyzer and transfer the reference plane to the probe tip;

[0011] Step 2: using a vector network analyzer to measure S parameters of the device under test and the calibration device, wherein the calibration device includes a basic structure and a combination of any two basic structures, wherein the basic structure includes an open circuit structure, a short circuit structure, a load structure, a semi-through structure, a through structure, and a transmission line structure;

[0012] Step 3: Convert the S parameters of the device under test and the calibration device into S2P files to obtain corresponding S2P data;

[0013] Step 4: construct an off-chip de-embedding module, which includes multiple calibration algorithms and the calibration device structure type required by each calibration algorithm;

[0014] Step 5. Input the S2P data corresponding to the device under test and the calibration device into the off-chip de-embedding module. The off-chip de-embedding module selects the S2P data corresponding to the calibration device to be processed according to the calibration device structure type corresponding to each calibration algorithm, and uses the calibration algorithm to process the S2P data corresponding to the selected calibration device and the S2P data corresponding to the device under test, and outputs the true value of the device under test after off-chip de-embedding by the calibration algorithm, thereby realizing off-chip de-embedding of the device under test based on different calibration algorithms.

[0015] Furthermore, the calibration algorithm includes calibration algorithms based on 8-item error models, 12-item error models, and 16-item error models, specifically including SOLT calibration algorithm, TRL calibration algorithm, m-TRL (Thru-Reflect-Line1-Line2...) calibration algorithm, LRRM (Line-Reflect1-Reflect2-Match) calibration algorithm, LRM (Line-Reflect-Match) calibration algorithm, LMR16 (Thru-'Match-match'-'Reflect-reflect'-'Reflect-match'-'Match-reflect') calibration algorithm, and SixteenTerm calibration algorithm.

[0016] Furthermore, the calibration device structure types required by the SOLT calibration algorithm include an open circuit structure, a short circuit structure, a load structure and a through structure.

[0017] Furthermore, the calibration device structure types required by the TRL calibration algorithm include a through structure, a transmission line structure and a short circuit structure, or include a through structure, a transmission line structure and an open circuit structure.

[0018] Furthermore, the m-TRL calibration algorithm includes a NIST-TRL calibration algorithm and a TUG-TRL calibration algorithm, and the required calibration device structure types include a straight-through structure, a transmission line structure, and a short-circuit structure, or include a straight-through structure, a transmission line structure, and an open-circuit structure.

[0019] Furthermore, the calibration device structure types required by the LRRM calibration algorithm include an open circuit structure, a short circuit structure, a load structure and a through structure.

[0020] Furthermore, the calibration device structure types required by the LMR16 calibration algorithm and the SixteenTerm calibration algorithm include calibration devices formed by a combination of an open circuit structure, a short circuit structure, a load structure, and a through structure.

[0021] Furthermore, the device under test and the calibration device are prepared by using a HEMT (High Electron Mobility Transistor) process, and the substrate is Si, InP, GaAs or GaN.

[0022] The beneficial effects of the present invention are:

[0023] For devices under test operating in the millimeter wave / terahertz frequency band, the present invention proposes a high-consistency fast de-embedding method in the millimeter wave / terahertz frequency band, which only requires one S parameter measurement for the device under test and the calibration device. By utilizing an off-chip de-embedding module including a plurality of calibration algorithms, off-chip de-embedding of the device under test based on different calibration algorithms can be achieved. There is no need to perform multiple measurements on the same device with the same structure or to measure different devices with the same structure. While ensuring the same de-embedding accuracy as the traditional second de-embedding method, it effectively avoids the problems of poor consistency caused by multiple tests of the GSG probe (the locations of the GSG probe pressure points are different) and cumbersome on-chip calibration / de-embedding. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 This is a flow chart of the millimeter wave / terahertz frequency band high consistency fast de-embedding method provided in Example 1 of the present invention;

[0025] Figure 2 This is a schematic diagram of the structure of a calibration device used in Example 1 of the present invention;

[0026] Figure 3 Schematic diagram of the structure of the device under test in embodiment 1 of the present invention;

[0027] Figure 4 The actual capacitance value of the device under test in Example 1 of the present invention is compared with the S parameter after on-chip TRL calibration, off-chip LRRM calibration, off-chip TRL calibration and off-chip NIST-TRL calibration respectively;

[0028] Figure 5 The Smith chart comparison results of the actual capacitance value of the device under test in Example 1 of the present invention and the on-chip TRL calibration, off-chip LRRM calibration, off-chip TRL and off-chip NIST-TRL calibration are shown;

[0029] The reference numerals are as follows:

[0030] 201, open circuit structure; 202, short circuit structure; 203, load structure; 204, semi-through structure; 205, through structure; 206, first transmission line structure; 207, second transmission line structure; 301, capacitor structure. DETAILED DESCRIPTION

[0031] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Generally, the components of the embodiments of the present invention described and shown in the drawings here can be arranged and designed in various different configurations.

[0032] Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the invention claimed for protection, but merely represents selected embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0033] Example 1

[0034] This embodiment provides a high consistency and fast de-embedding method in the millimeter wave / terahertz frequency band, wherein the device under test and the calibration device involved are prepared based on the 35nm InP HEMT process. Specifically, the device under test is as follows: Figure 3 The capacitor structure 301 shown in FIG. Figure 2 As shown, it includes an open circuit structure 201, a short circuit structure 202, a load structure 203, a semi-straight-through structure 204, a straight-through structure 205, a first transmission line structure 206 and a second transmission line structure 207; the substrates of the device under test and the calibration device are both 50μm thick InP substrates; the signal line structures of the device under test and the calibration device are both grounded coplanar waveguides, the line width of the signal line is 10μm, and the spacing between the signal line and the ground planes on both sides is also 10μm.

[0035] Based on the device under test and the calibration device, the process of a high consistency fast de-embedding method in the millimeter wave / terahertz frequency band provided in this embodiment is as follows: Figure 1 As shown, the specific steps include:

[0036] Step 1: Use a commercial impedance standard substrate (specifically a quartz substrate) to calibrate the vector network analyzer, transfer the reference plane to the probe tip, and remove the influence of errors such as switching items;

[0037] Step 2: Use a vector network analyzer to measure the S parameters of the device under test and the calibration device. The S parameters of the device under test are recorded as S cap The S parameters of the open circuit structure 201, the short circuit structure 202, the load structure 203, the semi-through structure 204, the through structure 205, the first transmission line structure 206 and the second transmission line structure 207 in the calibration device are S open , S short , S load , S halfthru , S thru , S line1 and S line2 ;

[0038] Step 3, converting the S parameters of the device under test and the calibration device into S2P files to obtain corresponding S2P data; wherein, the S2P data corresponding to the device under test is cap.s2p; the S2P data corresponding to the open circuit structure 201, the short circuit structure 202, the load structure 203, the half-through structure 204, the through structure 205, the first transmission line structure 206 and the second transmission line structure 207 in the calibration device are open.s2p, short.s2p, load.s2p, halfthru.s2p, thru.s2p, line1.s2p and line2.s2p respectively;

[0039] Step 4: construct an off-chip de-embedding module, which includes multiple calibration algorithms and the calibration device structure type required by each calibration algorithm;

[0040] Specifically, the calibration algorithms include LRRM calibration algorithm, TRL calibration algorithm and NIST-TRL calibration algorithm, all of which are embedded in the off-chip de-embedding module based on the corresponding algorithm functions in the python library;

[0041] The calibration device structure types required by the LRRM calibration algorithm include an open circuit structure 201, a short circuit structure 202, a load structure 203, and a through structure 205; the calibration device structure types required by the TRL calibration algorithm include a through structure 205, a short circuit structure 202, and a first transmission line structure 206; the calibration device structure types required by the NIST-TRL calibration algorithm include a through structure 205, a short circuit structure 202, a first transmission line structure 206, and a second transmission line structure 207;

[0042] Step 5. Input the S2P data (including cap.s2p, open.s2p, short.s2p, load.s2p, halfthru.s2p, thru.s2p, line1.s2p and line2.s2p) corresponding to the device under test and the calibration device into the off-chip de-embedding module. The off-chip de-embedding module selects the S2P data corresponding to the calibration device to be processed according to the calibration device structure type corresponding to each calibration algorithm, and uses the calibration algorithm to process the S2P data corresponding to the selected calibration device and the S2P data corresponding to the device under test, and outputs the capacitance value of the device under test after off-chip de-embedding by the LRRM calibration algorithm (referred to as off-chip LRRM calibration), the capacitance value of the device under test after off-chip de-embedding by the TRL calibration algorithm (referred to as off-chip TRL calibration), and the capacitance value of the device under test after off-chip de-embedding by the NIST-TRL calibration algorithm (referred to as off-chip NIST-TRL calibration), thereby realizing off-chip de-embedding of the device under test based on different calibration algorithms.

[0043] In order to verify the de-embedding accuracy of the off-chip de-embedding of the device under test based on different calibration algorithms implemented in this embodiment, it is compared with the traditional second de-embedding method. Specifically, the TRL calibration algorithm built into the vector network analyzer is used in the "second calibration" to obtain the capacitance value of the device under test after on-chip de-embedding using the traditional second de-embedding method (referred to as on-chip TRL calibration).

[0044] Figure 4 : is the comparison result between the actual capacitance value of the device under test in this embodiment and the S parameter after on-chip TRL calibration, off-chip LRRM calibration, off-chip TRL calibration and off-chip NIST-TRL calibration, respectively. Figure 5 The comparison result of the Smith chart shows that the capacitance value of the device under test obtained after the off-chip LRRM calibration, off-chip TRL calibration and off-chip NIST-TRL calibration in this embodiment is basically consistent with the de-embedding accuracy of the capacitance value of the device under test obtained after the traditional on-chip TRL calibration.

[0045] To summarize, the high-consistency fast de-embedding method in the millimeter-wave / terahertz frequency band proposed in this embodiment only requires one S parameter measurement of the device to be tested and the calibration device. It can effectively avoid the poor consistency caused by multiple tests of the GSG probe (the location of the GSG probe pressure point is different) and cumbersome on-chip calibration / de-embedding problems while ensuring the same de-embedding accuracy as the traditional on-chip TRL calibration.

[0046] The above embodiments only illustrate the principles and advantages of the present invention, and are not intended to limit the present invention. They are only intended to help understand the principles of the present invention. The protection scope of the present invention is not limited to the above configurations and embodiments. Those skilled in the art can make various other specific modifications and combinations based on the disclosed technology without departing from the essence of the present invention, but they are still within the protection scope of the present invention.

Claims

1. A high consistency and fast de-embedding method in the millimeter wave / terahertz frequency band, It is characterized in that The following steps are involved: Step 1: Use an impedance standard substrate to calibrate the vector network analyzer; Step 2: using a vector network analyzer to measure S parameters of the device under test and the calibration device, wherein the calibration device includes a basic structure and a combination of any two basic structures, wherein the basic structure includes an open circuit structure, a short circuit structure, a load structure, a semi-through structure, a through structure, and a transmission line structure; Step 3: Convert the S parameters of the device under test and the calibration device into corresponding S2P data; Step 4: construct an off-chip de-embedding module, which includes multiple calibration algorithms and the calibration device structure type required by each calibration algorithm; Step 5. Input the S2P data corresponding to the device under test and the calibration device into the off-chip de-embedding module. The off-chip de-embedding module selects the S2P data corresponding to the calibration device to be processed according to the calibration device structure type corresponding to each calibration algorithm, and uses the calibration algorithm to process the S2P data corresponding to the selected calibration device and the S2P data corresponding to the device under test, and outputs the true value of the device under test after off-chip de-embedding by the calibration algorithm, thereby realizing off-chip de-embedding of the device under test based on different calibration algorithms.

2. According to the millimeter wave / terahertz frequency band high consistency fast de-embedding method of claim 1, It is characterized in that The calibration algorithms include multiple ones of a SOLT calibration algorithm, a TRL calibration algorithm, an m-TRL calibration algorithm, a LRRM calibration algorithm, a LRM calibration algorithm, a LMR16 calibration algorithm, and a SixteenTerm calibration algorithm.

3. According to the millimeter wave / terahertz frequency band high consistency fast de-embedding method according to claim 2, It is characterized in that The calibration device structure types required by the SOLT calibration algorithm include an open circuit structure, a short circuit structure, a load structure and a through structure.

4. According to the millimeter wave / terahertz frequency band high consistency fast de-embedding method according to claim 2, It is characterized in that The calibration device structure types required by the TRL calibration algorithm include a through structure, a transmission line structure and a short circuit structure, or include a through structure, a transmission line structure and an open circuit structure.

5. According to the millimeter wave / terahertz frequency band high consistency fast de-embedding method according to claim 2, It is characterized in that The m-TRL calibration algorithm includes the NIST-TRL calibration algorithm and the TUG-TRL calibration algorithm, and the required calibration device structure types include a straight-through structure, a transmission line structure and a short-circuit structure, or include a straight-through structure, a transmission line structure and an open-circuit structure.

6. According to the millimeter wave / terahertz frequency band high consistency fast de-embedding method as claimed in claim 2, It is characterized in that The calibration device structure types required by the LRRM calibration algorithm include an open circuit structure, a short circuit structure, a load structure and a through structure.

7. According to the millimeter wave / terahertz frequency band high consistency fast de-embedding method as claimed in claim 2, It is characterized in that The calibration device structure types required by the LMR16 calibration algorithm and the SixteenTerm calibration algorithm include calibration devices formed by a combination of an open circuit structure, a short circuit structure, a load structure, and a through structure.

8. According to the millimeter wave / terahertz frequency band high consistency fast de-embedding method of claim 1, It is characterized in that The device to be tested and the calibration device are prepared by using HEMT technology, and the substrate is Si, InP, GaAs or GaN.