Method for flattening the impedance of a power delivery network
By selecting the appropriate decoupling capacitor in the power delivery network and setting it in a key port position, adjusting the impedance of the power delivery network, the problem of uneven self-impedance of the power delivery network is solved, and more stable power delivery is achieved, reducing the number of times Voltage disturbance and distortion phenomena.
Patent Information
- Application Number
- CN202110369177.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2021-04-06
- Publication Date
- 2024-11-22
- Estimated Expiration
- 2041-04-06
AI Technical Summary
Uneven self-impedance of the power delivery network leads to voltage disturbance and distortion problems, especially on the printed circuit board of the server, affecting the stability of the equipment.
By capturing the impedance parameters of the power delivery network, defining the target impedance, calculating the intersection frequency, selecting the appropriate decoupling capacitor and setting it at the key port, performing parallel operations to adjust the impedance until the target impedance is flattened.
It effectively reduces the self-impedance of the power delivery network, improves the flatness of the impedance, reduces voltage disturbance and distortion phenomena, and improves the power delivery stability of the server.
Smart Images

Figure CN115175455B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to a method for flattening the impedance of a power delivery network, and more particularly to a method for flattening the impedance of the power delivery network by selecting a suitable decoupling capacitor. Background Art
[0002] The power distribution network (PDN) is a transmission medium for the voltage source to provide DC power to the load, and on a printed circuit board, it is usually the transmission path from the voltage regulator module (VRM) to the load circuit. The power distribution network can be composed of the power layer and the ground layer of the printed circuit board, and the cables, connectors, capacitors and other components electrically connected to the power layer and the ground layer can all be part of the power distribution network.
[0003] The parasitic effect of the power delivery network can cause parallel resonance, resulting in high self-impedance at the load circuit location. At the load circuit location, voltage disturbances may occur due to changes in load current and the high self-impedance of the power delivery network.
[0004] In addition, when the self-impedance versus frequency curve of the power delivery network is not flat, the voltage disturbances overlap to cause distorted waves, which will make the voltage disturbance problem more serious. A server is an electronic device that includes a printed circuit board. The power layer and ground layer of the printed circuit board can constitute a part of the power delivery network. Therefore, in the design of the power delivery network of the printed circuit board, the server needs to consider not only reducing the impedance to avoid noise interference, but also the impedance flattening problem to avoid distorted waves. Summary of the invention
[0005] The present application provides a method for flattening the impedance of a power delivery network, comprising: acquiring a set of impedance parameters of the power delivery network; obtaining an impedance of the power delivery network based on the set of impedance parameters; defining a target impedance; performing a first importance calculation to determine a first port; obtaining a crossover frequency based on the target impedance and the impedance of the power delivery network; selecting a first decoupling capacitor based on the crossover frequency; and setting the first decoupling capacitor at the first port.
[0006] Preferably, it also includes: performing a parallel operation to generate an adjusted impedance of the power delivery network; performing a second importance calculation to determine a second port; obtaining an adjusted intersection frequency based on the target impedance and the adjusted impedance of the power delivery network; selecting a second decoupling capacitor based on the adjusted intersection frequency; and setting the second decoupling capacitor at the second port.
[0007] Preferably, the first decoupling capacitor is selected, comprising: selecting a group of capacitors, wherein a self-resonant frequency of each capacitor of the group of capacitors is between a first value and a second value; and selecting one of the group of capacitors as the first decoupling capacitor; wherein the first value is A times the intersection frequency, the second value is B times the intersection frequency, A and B are positive parameters, and 0 <A<B。
[0008] Preferably, the A is substantially 1.5.
[0009] Preferably, the B is substantially 3.
[0010] Preferably, the selecting of the first decoupling capacitor comprises: selecting k capacitors; calculating the parallel value of each of the k capacitors and a self-impedance of the power delivery network respectively to obtain k parallel impedance values; selecting m parallel impedance values within a specific range from the k parallel impedance values; and selecting one of the m capacitors corresponding to the m parallel impedance values as the first decoupling capacitor; wherein m and k are positive integers, and 0 <m≤k。
[0011] Preferably, the specific range is between α times and β times the target impedance, and 0<α<β.
[0012] Preferably, the first decoupling capacitor is selected, including: selecting k capacitors; calculating the parallel value of each capacitor of the k capacitors and a self-impedance of the power delivery network respectively, to obtain k parallel impedance values; selecting a parallel impedance value between a first specific range among the k parallel impedance values; and when the k parallel impedance values do not fall within the first specific range, selecting m parallel impedance values between a second specific range among the k parallel impedance values; and selecting one of the m capacitors corresponding to the m parallel impedance values as the first decoupling capacitor; wherein m and k are positive integers, 0 <m≤k。
[0013] Preferably, selecting the first decoupling capacitor includes: calculating the parallel value of each capacitance of a plurality of capacitors and a self-impedance of the power delivery network respectively to obtain a plurality of reference impedances; generating a plurality of reference intersection frequencies based on the plurality of reference impedances and the target impedance; and selecting a capacitor corresponding to a highest one of the plurality of reference intersection frequencies as the first decoupling capacitor.
[0014] Preferably, performing the first importance calculation includes: respectively calculating a plurality of inductance characteristic values that are reduced after a plurality of ports of the power delivery network are provided with capacitors; and selecting a port corresponding to the highest of the plurality of inductance characteristic values as the first port. BRIEF DESCRIPTION OF THE DRAWINGS
[0015] Figure 1 FIG. 4 is a flow chart of a method for flattening the impedance of a power delivery network according to an embodiment.
[0016] Figure 2 FIG. 4 is a curve diagram showing the relationship between impedance and frequency in an embodiment.
[0017] Figures 3 to 5 for Figure 1 Flowchart for selecting decoupling capacitors.
[0018] Figure 6 For execution Figure 5 A curve diagram of the process.
[0019] Explanation of symbols:
[0020] 100: Method
[0021] 110 to 199, 310, 320, 410 to 450, 510 to 530: Steps
[0022] Z11, Z11', Z11'a, Z11'b, Z11'c,: impedance
[0023] Fh: Predetermined frequency
[0024] Fx, Fx', Fx1, Fx2, Fx3: intersection frequency DETAILED DESCRIPTION
[0025] To solve the above problems, decoupling capacitors can be placed to reduce the self-impedance of the power delivery network. However, although the impedance of the power delivery network can be reduced to the target impedance, the impedance of the power delivery network is still not flattened and optimized. The method provided in the embodiment can be used to flatten and optimize the impedance of the power delivery network, as described below.
[0026] Figure 1FIG. 1 is a flow chart of a method 100 for flattening the impedance of a power delivery network, according to an embodiment. Figure 2 FIG. 1 is a graph showing the relationship between impedance and frequency in an embodiment. Figure 1 and Figure 2 As shown, method 100 may include the following steps:
[0027] Step 110: Acquire a set of impedance parameters of the power delivery network;
[0028] Step 120: Generate an impedance Z11 of the power delivery network according to the set of impedance parameters;
[0029] Step 130: define the target impedance Ztarget;
[0030] Step 140: Perform importance calculation to determine the port location;
[0031] Step 150: Obtain the crossover frequency Fx according to the target impedance Ztarget and the impedance Z11 of the power delivery network;
[0032] Step 160: Select a decoupling capacitor according to the intersection frequency Fx;
[0033] Step 170: Setting a decoupling capacitor at the port;
[0034] Step 180: Perform a parallel operation to generate an adjusted impedance Z11 ′ of the power delivery network;
[0035] Step 190: Observe whether the adjusted impedance Z11' meets the target impedance Ztarget; if so, proceed to step 199; if not, proceed to step 140;
[0036] Step 199: End.
[0037] The following describes by way of example Figure 1 In step 110, scattering parameters (S-parameters) of the power delivery network may be acquired, and impedance parameters may be calculated based on the scattering parameters, wherein the impedance parameters may be elements of an impedance matrix (Z-matrix) of the power delivery network.
[0038] In step 120, the impedance Z11 of the power delivery network can be obtained according to the set of impedance parameters. The curve of the impedance Z11 is as follows: Figure 2 As shown, the curve of impedance Z11 can present the self-impedance of the power delivery network at each frequency. Figure 2 This is only an example and is not intended to limit the embodiments. Figure 2In this example, 1e+05 can represent 1×105, 1e+06 can represent 1×106, and so on.
[0039] The power delivery network may include multiple ports, the impedance Z11 may be the input impedance of the port coupled to the load device (eg, integrated circuit), and the other ports may be used to set decoupling capacitors to flatten the impedance Z11, as described below.
[0040] In step 130, the target impedance Ztarget may be defined according to actual requirements and product specifications, such as Figure 2 As shown, the curve of the target impedance Ztarget may correspond to the target impedance value of each frequency. For example, in the frequency band below the predetermined frequency Fh, the target impedance Ztarget may be substantially constant. Step 130 may be independent of steps 110 and 120. For example, step 130 may be performed first, and then steps 110 and 120 may be performed.
[0041] According to the embodiment, as described in steps 140 to 170, the most important port can be selected from multiple ports of the power delivery network, and then the most suitable capacitor can be selected from multiple candidate capacitors as a decoupling capacitor to be set at the selected port.
[0042] In step 140, the multiple inductance characteristic values reduced after the multiple ports of the power delivery network are provided with capacitors can be calculated respectively, and the port corresponding to the highest of the multiple inductance characteristic values can be selected as the selected port. In other words, among the multiple ports, the port whose inductance characteristic is reduced the most after the capacitor is provided can be the port with the highest importance.
[0043] For example, if the power delivery network is a 2-port network having a first port (e.g., coupled to a load device such as an integrated circuit) and a second port (e.g., coupled to a decoupling capacitor), the calculations of equations eq-1 and eq-2 may be performed:
[0044]
[0045]
[0046] Wherein, L11 and L22 may be the self-inductance of the first port and the second port, L12 may be the mutual inductance of the first port and the second port, L 11wdecap@2 The second port may be an adjusted element of the inductance matrix after the decoupling capacitor is set, and ΔL 11wGND@2 It can be the inductance characteristic reduced after the second port is set with a decoupling capacitor. Among them, Ld can be the self-inductance value of the decoupling capacitor set, which can be a constant and can be ignored for the convenience of calculation. The above formulas eq-1 and eq-2 are only examples. If in an n-port network, the general formula related to the mutual inductance of the Ath port can be as follows:
[0047]
[0048]
[0049] For example, if the power delivery network is a 10-port network, then according to the principle of equation eq-2, Table 1 can be obtained:
[0050]
[0051] (Table 1)
[0052] In the example of Table 1, it can be seen that when the decoupling capacitor is set at port P2, the inductance characteristic is reduced the most, so port P2 is the most important and can be selected in step 140. The number and value of ports in Table 1 are only examples, and the embodiment is not limited thereto.
[0053] In step 150, if Figure 2 As shown, the intersection frequency Fx can be obtained on the impedance versus frequency graph according to the target impedance Ztarget curve and the self-impedance Z11 curve of the power delivery network at this time.
[0054] In step 160 , a decoupling capacitor may be selected from a decoupling capacitor library, and then in step 170 , the selected decoupling capacitor may be placed at the port determined in step 140 .
[0055] After executing step 170 , each element of the equivalent impedance matrix of the power delivery network may be changed due to the setting of the decoupling capacitor, so the self-impedance of the power delivery network needs to be recalculated.
[0056] For example, if the power delivery network has 100 ports and one port coupled to a load device (such as an integrated circuit), the dimension of the original impedance matrix may be 101×101. After executing step 170 and setting a decoupling capacitor, the dimension of the impedance matrix may be 100×100, and each element of the matrix may change, so the impedance matrix must be regenerated and the self-impedance must be calculated based on it.
[0057] In step 180, a parallel operation can be performed based on the decoupling capacitors set in step 170 to generate a curve of the adjusted impedance Z11' of the power delivery network. The parallel operation can be shown in equation eq-5:
[0058] Z11'=Z11-(Z1i×Zi1) / Z11+Zd...eq-5;
[0059] Among them, Z11 can be the self-impedance of the power delivery network before adjustment, Z11' can be the self-impedance of the adjusted power delivery network after setting the decoupling capacitor, Z1i and Zi1 can be the elements of the impedance matrix. If the matrix dimension is n×n, then i and n are positive integers and 0 < i ≤ n. Zd can be the impedance of the set decoupling capacitor.
[0060] In step 190, it can be observed whether the adjusted impedance Z11' meets the target impedance Ztarget. If so, the process can enter step 199 to end the process; if not, then it can enter step 140 to re-select the port, and select and set a suitable decoupling capacitor at the selected port according to the adjusted intersection frequency Fx', so that the curve of the impedance of the power delivery network is closer to the curve of the target impedance Ztarget and flatter on the impedance-frequency coordinate.
[0061] In step 190, it can be determined whether the adjusted self-impedance Z11' meets the target impedance Ztarget at least according to whether the adjusted impedance Z11' is less than the target impedance Ztarget within a specific frequency band. Also, the flatness can be determined according to the standard deviation and average value of the adjusted impedance Z11' within a specific frequency band. For example, the standard deviation and average value of the adjusted impedance Z11' within a specific frequency band can be divided and expressed as a percentage to obtain the flatness.
[0062] Regarding steps 140 to 190, for example, in step 140, a first importance calculation can be performed to determine the first port, and as Figure 1 shown, steps 150 to 180 are performed to set the first decoupling capacitor at the first port and obtain the adjusted impedance; if in step 190, the adjusted impedance Z11' does not meet the target impedance Ztarget, step 140 can be executed again to perform a second importance calculation to determine the second port, and steps 150 to 180 are executed again to set the second decoupling capacitor at the second port and obtain the adjusted impedance; in this way, the flatness of the adjusted impedance Z11' can be improved in a loop manner. The first importance calculation and the second importance calculation mentioned here are only examples used to describe the steps that can be executed in a loop manner, Figure 1 however, the embodiments are not limited to this. If necessary, Figure 1The steps can be executed in a loop for more times (e.g., more than two times) to perform importance calculations, thereby improving the flatness of the adjusted impedance Z11'. In this example, as described above, the first importance calculation may include calculating the reduced complex inductance characteristic values respectively after capacitors are set for a plurality of ports of the power delivery network, and selecting the port corresponding to the highest one of the plurality of inductance characteristic values as the first port; similarly, the second importance calculation may include calculating the reduced complex inductance characteristic values respectively after capacitors are set for a plurality of ports of the power delivery network, and selecting the port corresponding to the highest one of the plurality of inductance characteristic values as the second port; if there are a third importance calculation, a fourth importance calculation, etc., they can be analogized accordingly.
[0063] Figure 3 For Figure 1 FIG. 158 is a flowchart of selecting a decoupling capacitor in step 160. According to an embodiment, Figure 1 Step 160 of
[0064] Step 310: Select a group of capacitors, and the self-resonant frequency (SRF) of each capacitor in the group is between a first value and a second value;
[0065] Step 320: Select one of the group of capacitors as the decoupling capacitor.
[0066] Figure 3 The first value mentioned can be A times the intersection frequency Fx, the second value can be B times the intersection frequency Fx, A and B are positive parameters, and 0 < A < B.
[0067] For example, if there are 200 capacitors available for selection in the database, and the self-resonant frequencies of 30 capacitors are between A×Fx and B×Fx, then one of the 30 capacitors can be selected as the decoupling capacitor. According to an embodiment, 0.5 < A < B < 5. According to an embodiment, A can be substantially 1.5, and B can be substantially 3.
[0068] Figure 4 For Figure 1 FIG. 172 is a flowchart of selecting a decoupling capacitor in step 160. According to an embodiment, Figure 1 Step 160 of
[0069] Step 410: Select k capacitors;
[0070] Step 420: Calculate the parallel values of the self-impedance of each of the k capacitors and the power delivery network respectively to obtain k parallel impedance values;
[0071] Step 430: Among the k parallel impedance values, is there a parallel impedance value within a specific range? If yes, go to Step 440; if no, go to Step 460;
[0072] Step 440: Select m parallel impedance values within the specific range from the k parallel impedance values;
[0073] Step 450: Select one of the m capacitors corresponding to the m parallel impedance values as the decoupling capacitor; and
[0074] Step 460: Adjust the specific range and go to Step 430.
[0075] Figure 4 where m and k are positive integers, and 0 < m ≤ k. The specific range in Step 430 is between α times and β times the target impedance Ztarget, and 0 < α < β.
[0076] For example, if 100 candidate capacitors are respectively arranged at the port determined by Figure 1 Step 140, 100 parallel impedance values, that is, the adjusted self-impedances Z11' of 100 power delivery networks, can be obtained. Among these 100 adjusted self-impedances Z11', if 40 adjusted self-impedances fall within the range of α × Ztarget to β × Ztarget, one of the 40 candidate capacitors corresponding to these 40 adjusted self-impedances can be selected as the decoupling capacitor, and the remaining 60 candidate capacitors are not selected.
[0077] For example, α can be 0.9, and β can be 1. If there is no parallel impedance value within the specific range in Step 430, α can be decreased to increase the specific range described in Step 430, so that a parallel impedance value within the specific range can be selected in Step 430. For example, α can be gradually adjusted from 0.9 to 0.85, 0.8, 0.75, etc., to gradually increase the specific range. According to another embodiment, α and β can also be adjusted so that the specific range after adjustment in Step 460 is different from the specific range before adjustment.
[0078] Figure 5 is Figure 1 the flowchart for selecting the decoupling capacitor in Step 160 of Figure 1 Step 160 of
[0079] Step 510: Calculate the parallel values of each capacitor of the complex number of capacitors and the self-impedance Z11 of the power delivery network respectively to obtain a complex number of reference impedance curves;
[0080] Step 520: Generate a complex number of reference intersection frequencies according to the complex number of reference impedance curves and the target impedance; and
[0081] Step 530: Select the capacitor corresponding to the highest one of the plurality of reference intersection frequencies as the decoupling capacitor.
[0082] Figure 6 Can be executed Figure 5 A curve diagram of the process. Figure 6 This is only an example and does not limit the embodiment. For example, when three candidate capacitors are respectively set at Figure 1 After the port position is determined in step 140, three reference impedances Z11'a, Z11'b and Z11'c and their reference impedance curves can be obtained. Figure 6 As shown, according to the curve of the target impedance Ztarget and the reference impedance curve, the reference intersection frequencies Fx1, Fx2 and Fx3 can be obtained, wherein the reference intersection frequency Fx3 is the highest of the three reference intersection frequencies. Therefore, among the three candidate capacitors, the capacitor corresponding to the curve of the reference intersection frequency Fx3 and the reference impedance Z11'c can be selected as the decoupling capacitor.
[0083] Figure 3 , Figure 4 and Figure 5 The process can be used to gradually select decoupling capacitors from the capacitor database. For example, if there are 100 candidate capacitors, Figure 3 The process is to select 30 capacitors and then Figure 4 The process selects 10 capacitors from the 30 capacitors, and then Figure 5 The process selects one capacitor from the 10 capacitors as Figure 1 The decoupling capacitor selected in step 160 is used. However, this is only an example. Figure 3 , Figure 4 and Figure 5 The order of the process is not limited to this.
[0084] In summary, the method provided by the embodiment can effectively select the most important port and select a suitable decoupling capacitor to be set at the selected port. Figure 1 The loop method can be used to flatten the self-impedance of the power delivery network by setting appropriate decoupling capacitors at appropriate ports. According to experiments, the curve of the impedance Z11' after adjustment obtained by the method of the embodiment in the impedance-frequency coordinate is closer to the curve of the target impedance Ztarget, that is, it has better flatness. Therefore, it is helpful to solve the problems in this field.
[0085] The method of the present invention can be applied to the design of a printed circuit board of a server, so that the server has a high-quality power delivery network to enhance the stability of the server, so that the server can be applied to artificial intelligence (AI) computing and edge computing, and can also be used as a 5G server, a cloud server, or an Internet of Vehicles server.
[0086] The above description is only a preferred embodiment of the present invention. All equivalent changes and modifications made according to the claims of the present invention should fall within the scope of the present invention.
Claims
1. A method for flattening the impedance of a power delivery network, It is characterized in that Include: Retrieving a set of impedance parameters of the power delivery network; Obtaining an impedance of the power delivery network according to the set of impedance parameters; Define a target impedance; Performing a first importance calculation to determine a first port; Obtaining a crossover frequency according to the target impedance and the impedance of the power delivery network; A first decoupling capacitor is selected according to the intersection frequency, wherein the selecting the first decoupling capacitor comprises: Calculating the parallel value of each capacitor of the plurality of capacitors and a self-impedance of the power delivery network to obtain a plurality of reference impedances; generating a plurality of reference crossing frequencies according to the plurality of reference impedances and the target impedance; and selecting a capacitor corresponding to a highest one of the plurality of reference intersection frequencies as the first decoupling capacitor; and The first decoupling capacitor is disposed at the first port.
2. The method according to claim 1, It is characterized in that Also includes: performing a parallel operation to generate an adjusted impedance of the power delivery network; performing a second importance calculation to determine a second port; Obtaining an adjusted crossover frequency according to the target impedance and the adjusted impedance of the power delivery network; Selecting a second decoupling capacitor according to the adjusted intersection frequency; and The second decoupling capacitor is disposed at the second port.
3. The method according to claim 1, It is characterized in that The step of selecting the first decoupling capacitor comprises: Selecting a group of capacitors, each capacitor of the group of capacitors having a self-resonant frequency between a first value and a second value; and Selecting one of the group of capacitors as the first decoupling capacitor; The first value is A times the intersection frequency, the second value is B times the intersection frequency, A and B are positive parameters, and 0 <A<B。 4. The method according to claim 3, It is characterized in that Wherein A is substantially 1.
5.
5. The method according to claim 3, It is characterized in that Wherein B is substantially 3.
6. The method according to claim 1, It is characterized in that The step of selecting the first decoupling capacitor comprises: Select k capacitors; calculating the parallel value of each capacitance of the k capacitances and a self-impedance of the power delivery network respectively, so as to obtain k parallel impedance values; Selecting m parallel impedance values within a specific range from the k parallel impedance values; and Selecting one of the m capacitors corresponding to the m parallel impedance values as the first decoupling capacitor; Where m and k are positive integers, and 0 <m≤k。 7. The method according to claim 6, It is characterized in that The specific range is between α times and β times the target impedance, and 0<α<β.
8. The method according to claim 1, It is characterized in that The step of performing the first importance calculation comprises: respectively calculating a plurality of inductance characteristic values reduced after a plurality of ports of the power delivery network are provided with capacitors; and A port corresponding to the highest of the plurality of inductance characteristic values is selected as the first port.
9. A method of flattening the impedance of a power delivery network, Characterized in that, Comprising: Extracting a set of impedance parameters of a power delivery network; Obtaining an impedance of the power delivery network according to the set of impedance parameters; Defining a target impedance; Performing a first importance calculation to determine a first port; Obtaining an intersection frequency according to the target impedance and the impedance of the power delivery network; Selecting a first decoupling capacitor according to the intersection frequency, wherein the selection of the first decoupling capacitor comprises: Selecting k capacitors; respectively calculating the parallel values of each of the k capacitors and a self-impedance of the power delivery network to obtain k parallel impedance values; selecting the parallel impedance values within a first specific range among the k parallel impedance values; and when none of the k parallel impedance values fall within the first specific range, selecting m parallel impedance values within a second specific range among the k parallel impedance values; and selecting one of the m capacitors corresponding to the m parallel impedance values as the first decoupling capacitor; wherein m and k are positive integers, 0 < m ≤ k; and Setting the first decoupling capacitor at the first port.
