Method for efficiently improving impedance of packaged power distribution network
Through simulation calculation and circuit model adjustment, the time-consuming problem of decoupling capacitor selection in the improvement of impedance of the package power distribution network is solved, efficiency and economy are improved, and the target impedance is achieved quickly.
Patent Information
- Application Number
- CN202411114607.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-14
- Publication Date
- 2024-11-19
AI Technical Summary
The prior art uses a long time to select decoupling capacitors when improving the distribution network impedance of the package power supply, and requires frequent modification of the capacitance model and name on the package, which is inefficient.
By determining the target impedance and frequency range, the capacitance value and number of decoupling capacitors are simulated, and the cascade circuit model simulation is used to simulate the cascade circuit model, adjust the capacitor parameters to achieve the target impedance, and reduce actual packaging modifications.
It improves the effectiveness and efficiency of the selection of decoupling capacitors, shortens the simulation time, and realizes the use of the minimum number of decoupling capacitors to achieve the target impedance, improving economic benefits.
Smart Images

Figure CN118983891A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of circuit design, and in particular to a method for efficiently improving the impedance of a package power distribution network. Background Art
[0002] The power distribution network (PDN) is a circuit path that delivers power from the source to the load. The current flows from the power supply to the load through the power plane of the PDN, and then flows from the load back to the power supply through the ground plane of the PDN. The function of the power distribution network is to provide a stable voltage for the load, respond quickly to changes in load current, and reduce switching noise.
[0003] The power distribution network (PDN) in the package needs to keep the supply voltage between the chip pads constant and keep it within a very small tolerance range. The main reason for power noise is the voltage drop caused by the sudden change of PDN impedance. In order to improve the noise problem of package PDN, the PDN impedance needs to be controlled below the target impedance. The impedance of PDN can be changed by adding decoupling capacitors to the package PDN plane. The core of this process is to determine the type and number of decoupling capacitors. The existing method is to add decoupling capacitors to the package PDN and then perform PDN impedance simulation. If the PDN impedance does not meet the requirements, it is necessary to re-add and change the decoupling capacitors on the package, and then perform PDN simulation until the obtained PDN results meet the requirements. In this method of the prior art, when the PDN impedance does not meet the requirements and the decoupling capacitor model or capacitance value needs to be changed, the decoupling capacitor model and capacitor name need to be modified on the package. If the number of decoupling capacitors is large, this process will be very time-consuming. Summary of the invention
[0004] In view of the above-mentioned deficiencies in the prior art, the technical problem to be solved by the present invention is: how to provide a method for effectively improving the effectiveness of decoupling capacitor selection and reducing time consumption to efficiently improve the impedance of the package power distribution network.
[0005] In order to solve the above technical problems, the present invention adopts the following technical solutions:
[0006] A method for efficiently improving the impedance of a package power distribution network comprises the following steps:
[0007] Step 1) determining the target impedance, the highest cutoff frequency and the lowest cutoff frequency of the power distribution network;
[0008] Step 2) simulating and obtaining the planar capacitance of the package power distribution network, and calculating the required minimum decoupling capacitor capacitance based on the planar capacitance;
[0009] Step 3) simulating and obtaining the actual impedance-operating frequency curve of the package power distribution network, and determining the required maximum decoupling capacitor capacitance value according to the actual impedance-operating frequency curve, and combining the required minimum decoupling capacitor capacitance value obtained in step 2) to obtain the required decoupling capacitor capacitance value range;
[0010] Step 4) calculating the minimum number of decoupling capacitors required to meet the target impedance;
[0011] Step 5) according to the capacitance range of the decoupling capacitor determined in step 3) and the minimum number of the decoupling capacitor determined in step 4), the decoupling capacitor and the package power distribution network are simulated until the peak value of the actual impedance curve obtained is lower than the target impedance;
[0012] Step 6) Obtain the capacitance and quantity of decoupling capacitors that need to be added to the package power distribution network.
[0013] Preferably, the calculation formula of the target impedance in step 1) is:
[0014]
[0015] Where: Z target is the target impedance, V DD is the supply voltage of a specific rail of the power distribution network, ripple% is the allowable ripple, I max The maximum peak current for a specific rail in a power distribution network.
[0016] Preferably, in step 2), one third of the capacitance of the planar capacitor is taken as a reference capacitance to obtain a capacitance of a capacitor closest to the reference capacitance, and the capacitance of the capacitor is used as the capacitance of the required minimum decoupling capacitor.
[0017] Preferably, in step 3), the actual impedance-operating frequency curve obtained by simulation is compared with the target impedance, and the lowest frequency at which the actual impedance peak exceeds the target impedance is obtained as the first frequency. According to the impedance-frequency curve of the capacitor, the frequency range in which the capacitor behaves as capacitance is obtained, and the capacitance of the largest capacitor that behaves as capacitance at the first frequency is selected as the capacitance of the required maximum decoupling capacitor.
[0018] Preferably, in step 3), the capacitance range of all available decoupling capacitors is determined according to the capacitance of the maximum decoupling capacitor and the capacitance of the minimum decoupling capacitor; and the capacitance of each decoupling capacitor is determined according to the geometric mean of the capacitances of its two adjacent decoupling capacitors.
[0019] Preferably, in step 4), the minimum number of required decoupling capacitors is calculated according to the following formula:
[0020]
[0021] Where: n is the minimum number of decoupling capacitors required, F max is the highest cutoff frequency of the power distribution network, ESL is the equivalent series inductance of the decoupling capacitor, Z target is the target impedance.
[0022] Preferably, in step 5), an S-parameter file containing the characteristics of the package power distribution network is generated through simulation, and a cascade circuit model is formed in SPICE with a sub-circuit file containing the RLC parameters of the decoupling capacitor to equivalently add the power distribution network after the decoupling capacitor is added, thereby realizing the simulation of the decoupling capacitor and the power distribution network.
[0023] Preferably, the step cascades the decoupling capacitors within the capacitance range in order from large to small capacitance with the s parameter file containing the characteristics of the package power distribution network. The number of decoupling capacitors with different capacitance values added is determined by the impedance peak change of the impedance curve after adding the decoupling capacitor with the capacitance value is less than the set value. When the impedance peak of the impedance curve after adding a decoupling capacitor with a certain capacitance value is less than the set value, decoupling capacitors with adjacent capacitance values are continuously added for cascading until the actual impedance curve peak value obtained is less than the target impedance.
[0024] Compared with the prior art, the present invention has the following advantages:
[0025] 1. The present invention combines the extracted package PDN and decoupling capacitors into a circuit in simulation software (such as spice) to simulate and obtain impedance. The impedance at this time is the PDN impedance after adding the decoupling capacitors. At this time, the sub-circuit model and number of the decoupling capacitors are directly changed in the simulation software to obtain the best PDN curve below the target impedance, without modifying the decoupling capacitor model and capacitor name on the package, thereby reducing the time for selecting the decoupling capacitors.
[0026] 2. The present invention predetermines the capacitance range of the required decoupling capacitor and the minimum number of decoupling capacitors required, greatly improving the effectiveness and pertinence of the selection of the capacitance and number of decoupling capacitors. Then, the spice simulator is used to simulate the impedance of the PDN after adding the decoupling capacitor. During the simulation process, the decoupling capacitor parameters can be adjusted by changing the decoupling capacitor sub-circuit model in the spice circuit. The simulation speed of each iteration is very fast, which improves the simulation efficiency of the power distribution network PDN impedance. At the same time, a capacitor adding method that uses the minimum number of decoupling capacitors to achieve the target PDN impedance can be obtained, effectively improving the economic benefits. BRIEF DESCRIPTION OF THE DRAWINGS
[0027] Figure 1 A flow chart of a method for efficiently improving the impedance of a package power distribution network according to the present invention;
[0028] Figure 2 It is the relationship curve between the actual impedance-operating frequency curve and the target impedance of the packaged PDN;
[0029] Figure 3 The actual impedance-operating frequency curve and target impedance relationship curve of the package PDN after adding decoupling capacitors. DETAILED DESCRIPTION
[0030] In order to make the purpose, technical solution and advantages of the embodiment of the present invention clearer, the technical solution of the embodiment of the present invention will be clearly and completely described below in conjunction with the drawings of the embodiment of the present invention. Obviously, the described embodiment is a part of the embodiment of the present invention, not all of the embodiments. Based on the described embodiment 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. Unless otherwise defined, the technical terms or scientific terms used herein should be the common meaning understood by people with general skills in the field to which the present invention belongs.
[0031] The words "first", "second" and similar words used in the patent application specification and claims of the present invention do not indicate any order, quantity or importance, but are only used to distinguish different components. Similarly, unless the context clearly indicates otherwise, the singular forms of "a", "an" or "the" and other similar words do not indicate a quantitative limitation, but indicate the existence of at least one. Words such as "include" or "comprise" mean that the elements or objects appearing before "include" or "comprise" cover the features, wholes, steps, operations, elements and / or components listed after "include" or "comprise", and do not exclude the existence or addition of one or more other features, wholes, steps, operations, elements, components and / or their collections. "Up", "down", "left", "right" and the like are only used to indicate relative positional relationships. When the absolute position of the described object changes, the relative positional relationship may also change accordingly.
[0032] As attached Figure 1 As shown, the present solution provides a method for efficiently improving the impedance of a package power distribution network, comprising the following steps:
[0033] Step 1) Determine the target impedance, the highest cutoff frequency and the lowest cutoff frequency at which the power distribution network takes effect. The highest cutoff frequency and the lowest cutoff frequency can be directly obtained from the corresponding technical requirements specification, and the target impedance can be calculated using the following formula.
[0034] The target impedance is calculated as:
[0035]
[0036] Where: Z target is the target impedance, V DD is the supply voltage of a specific rail of the power distribution network, ripple% is the allowable ripple, I max The maximum peak current for a specific rail in the power distribution network. V DD , ripple%,I max All of them can be obtained from the corresponding technical requirements specification. The corresponding values in the technical requirements specification are substituted into the formula to calculate the target impedance.
[0037] Step 2) The planar capacitance of the package power distribution network is simulated, and the capacitance of the minimum decoupling capacitor required is calculated based on the planar capacitance. One third of the capacitance of the planar capacitance is taken as the reference capacitance, and the capacitance of the capacitor closest to the reference capacitance is obtained, and the capacitance of the capacitor is used as the capacitance of the minimum decoupling capacitor required.
[0038] Specifically, a simulation tool is used to simulate the planar capacitance of the package power distribution network (the tools and methods for simulating the planar capacitance of the power distribution network are all prior art, so the simulation tools and simulation methods used in this scheme are not limited), and the capacitance of the required minimum decoupling capacitor is calculated through this planar capacitance value. This scheme selects 1 / 3 of the planar capacitance value as the reference capacitance value. The capacitance value of the minimum decoupling capacitor needs to be selected from the capacitance values of existing capacitors. Not all capacitors have capacitance values. Common capacitors have capacitance values of 1.0, 1.5, 2.2, 3.3, 4.7, 6.8 and integer multiples of 10. According to the calculated reference capacitance value, the capacitance value of the common capacitor closest to it is selected as the capacitance value of the required minimum decoupling capacitor. For example, if the calculated reference capacitance value is 2.0, 2.2 is selected as the capacitance value of the required minimum decoupling capacitor.
[0039] Step 3) simulates the actual impedance-operating frequency curve of the package power distribution network, and determines the required maximum decoupling capacitor capacitance based on the actual impedance-operating frequency curve, combined with the required minimum decoupling capacitor capacitance obtained in step 2) to obtain the required decoupling capacitor capacitance range.
[0040] Specifically, the actual impedance-operating frequency curve obtained by simulation is compared with the target impedance, as shown in the attached figure. Figure 2As shown, the lowest frequency at which the actual impedance peak exceeds the target impedance is obtained as the first frequency. According to the impedance-frequency curve of the capacitor, the frequency range in which the capacitor behaves as a capacitor is obtained, and the capacitance of the largest capacitor that behaves as a capacitor at the first frequency is selected as the capacitance of the required maximum decoupling capacitor. In actual operation, according to the information provided by the capacitor manufacturer, the impedance-frequency curve of each capacitance capacitor can be obtained, thereby obtaining the frequency range in which the capacitor can behave as a capacitor. This frequency range is compared with the frequency range in which the actual impedance peak exceeds the target impedance, and the frequency range in which the capacitor can behave as a capacitor is selected to be slightly larger than the lowest frequency at which the actual impedance peak exceeds the target impedance. Thus, the capacitance of the largest decoupling capacitor that can be selected can be obtained. At this time, it is combined with the capacitance of the required minimum decoupling capacitor obtained in the previous step to determine the capacitance range of the decoupling capacitor. Specifically, when the actual impedance peak is higher than the target impedance at low frequency, it is necessary to start from the capacitor with a larger capacitance and select a lower capacitance. When the actual impedance peak is higher than the target impedance at high frequency, it is necessary to start from the capacitor with a smaller capacitance and select a lower capacitance.
[0041] Furthermore, the capacitance range determined by the capacitance of the maximum decoupling capacitor and the capacitance of the minimum decoupling capacitor is used to obtain the capacitance of all available decoupling capacitors; and the capacitance of each decoupling capacitor is determined according to the geometric mean of the capacitances of its two adjacent decoupling capacitors. In this way, in order to obtain the minimum parallel resonant impedance peak, it is necessary to select a group of decoupling capacitors with small and similar equivalent series inductance (the equivalent series inductance of the decoupling capacitors can be obtained by simulating the existing capacitors separately, and the acquisition method is not limited), and the capacitance of each decoupling capacitor is the geometric mean of the capacitances of the decoupling capacitors on both sides, and the specific capacitance of the decoupling capacitors available for addition can be determined at this time.
[0042] Step 4) Calculate the minimum number of decoupling capacitors required to meet the target impedance. The number of decoupling capacitors will affect the packaging manufacturing cost. Although adding too many decoupling capacitors can ensure that the PDN impedance is below the target impedance, it will make the manufacturing cost too high. Therefore, it is necessary to calculate the minimum number of decoupling capacitors.
[0043] Specifically, the minimum number of required decoupling capacitors is calculated according to the following formula:
[0044]
[0045] Where: n is the minimum number of decoupling capacitors required, F max is the highest cutoff frequency of the power distribution network, ESL is the equivalent series inductance of the decoupling capacitor, Z target is the target impedance.
[0046] Step 5) According to the capacitance range of the decoupling capacitor determined in step 3) and the minimum number of decoupling capacitors determined in step 4), the decoupling capacitor and the package power distribution network are simulated until the peak value of the actual impedance curve obtained is lower than the target impedance.
[0047] An S parameter file containing the characteristics of the package power distribution network is generated through simulation (the method of obtaining the S parameter file is not limited and belongs to the prior art), and a cascade circuit model is formed in SPICE with a sub-circuit file containing the RLC parameters of the decoupling capacitor, which is used to add the equivalent power distribution network after the decoupling capacitor is added, so as to realize the simulation of the decoupling capacitor and the power distribution network. During the specific simulation, the decoupling capacitor sub-circuit is replaced and iterated at the circuit connection node to optimize the PDN impedance. When selecting the decoupling capacitor, the decoupling capacitors within the capacitance range are sequentially cascaded with the S parameter file containing the characteristics of the package power distribution network in order from large to small capacitance. The number of decoupling capacitors with different capacitance values to be added is determined by the impedance peak change of the impedance curve after adding the decoupling capacitor with the capacitance value being less than the set value. When the impedance peak of the impedance curve after adding a decoupling capacitor with a certain capacitance value is less than the set value, the decoupling capacitors with adjacent capacitance values are continuously added for cascading until the actual impedance curve peak value obtained is less than the target impedance, as shown in the attached figure. Figure 3 shown.
[0048] In the specific implementation, starting from low frequency, select the decoupling capacitor with the maximum value in the decoupling capacitor capacitance range and cascade it with the s-parameter file containing the package power network characteristics, add a sufficient number of decoupling capacitors to make the impedance peak lower than the target impedance, and then add a sufficient number of decoupling capacitors of the next capacitance value until the target impedance is met. During the simulation process, decoupling capacitors of a certain capacitance value can also be skipped, especially at the low frequency end, and low target impedance can be achieved without them. Generally speaking, decoupling capacitors with large capacitance values mainly act at low frequencies to reduce the PDN impedance at low frequencies, and decoupling capacitors with small capacitance values mainly act at high frequencies to reduce the PDN impedance at high frequencies. When adding decoupling capacitors, the total number needs to be limited to the minimum number of decoupling capacitors required to meet the target impedance in theory. If the impedance requirement is still not met after exceeding the minimum number of decoupling capacitors required to meet the target impedance in theory, the number of large-capacitance decoupling capacitors can be appropriately adjusted. Because small-capacitance decoupling capacitors not only have a major effect on high-frequency PDN impedance, they also affect low-frequency PDN impedance. Therefore, after adding a sufficient number of small-capacitance decoupling capacitors, the number of large-capacitance decoupling capacitors can be appropriately reduced. At this time, the actual low-frequency impedance can still meet the target impedance requirements. According to this method, a sufficient number of decoupling capacitors of various capacitances are added until the peak of the actual impedance curve is lower than the target impedance. By iterating the above method multiple times, a capacitor adding method that uses the minimum number of decoupling capacitors to achieve the target PDN impedance can be obtained. This method is mainly simulated through spice, which can greatly improve the simulation efficiency.
[0049] Step 6) Obtain the capacitance and quantity of decoupling capacitors that need to be added to the package power distribution network.
[0050] Compared with the prior art, the present invention combines the extracted package PDN and decoupling capacitors into a circuit in simulation software (such as spice) for simulation to obtain impedance. The impedance at this time is the PDN impedance after adding the decoupling capacitor. At this time, the sub-circuit model and quantity of the decoupling capacitor are directly changed in the simulation software to obtain the best PDN curve below the target impedance, without modifying the decoupling capacitor model and the capacitor name on the package, thereby reducing the time for selecting the decoupling capacitor. The present invention predetermines the capacitance range of the required decoupling capacitor and the minimum number of decoupling capacitors required, greatly improving the effectiveness and pertinence of the selection of the capacitance and number of decoupling capacitors, and then uses the spice simulator to simulate the impedance of the PDN after adding the decoupling capacitor. During the simulation process, the decoupling capacitor parameters can be adjusted by changing the decoupling capacitor sub-circuit model in the spice circuit. The simulation speed of each iteration is very fast, which improves the simulation efficiency of the power distribution network PDN impedance. At the same time, a capacitor adding method that uses the minimum number of decoupling capacitors to achieve the target PDN impedance can be obtained, effectively improving the economic benefits. Therefore, the present invention does not need to iterate in the full layout simulation, and does not need to modify the capacitor model and capacitor name on the package. The speed of full layout simulation and modification of capacitors on the package is slow and inefficient. Depending on the scale of the layout, one iteration may take several hours or more. The effect of the present invention is higher.
[0051] Below, the method of the present invention is described with a specific example:
[0052] According to the technical requirements, the operating frequency band of the packaged PDN is 100MHz-2GHz. The target impedance is calculated to be 0.08Ω, and the impedance threshold curve is obtained.
[0053] Open RLC parameter extraction software such as XtractIM, import the package file, set the relevant parameters, and simulate the planar capacitance of the VDDQ power distribution network to be 1.2nf. The calculation shows that the capacitance of the minimum decoupling capacitor required is about 0.4nf. The closest common capacitor capacitance is 0.47nf.
[0054] The total capacitance range available may be between 1μf and 0.47nf. The capacitance of each capacitor is the geometric mean of the capacitances of the capacitors on both sides. There are 10 different values to choose from: 1000nf, 470nf, 220nf, 100nf, 47nf, 22nf, 10nf, 4.7nf, 2.2nf, 1nf, 0.47nf. The impedance of the PDN is simulated by spice to obtain the actual impedance-operating frequency curve, such as Figure 2The actual impedance of the PDN in the working frequency band will fluctuate to a certain extent, and part of it will exceed the impedance threshold curve. It can be seen that the PDN impedance exceeds the impedance threshold curve after 1Ghz. By looking at the impedance-frequency curve of the capacitor, it can be seen that the 1000nf-22nf capacitor is inductive after 1Ghz, and the effect of reducing the PDN impedance is not obvious at this time. Therefore, when selecting the capacitance of the capacitor, start from the 10nf capacitor to the lower capacitance. It can be concluded that the capacitance of the decoupling capacitor available in this embodiment is: 10nf, 4.7nf, 2.2nf, 1nf, 0.47nf.
[0055] If the equivalent series inductance of each decoupling capacitor in a further group of decoupling capacitors is similar, a smaller parallel resonant impedance peak can be obtained. In practical applications, the equivalent series inductance of the decoupling capacitor can be determined based on the information provided by the capacitor manufacturer, or the impedance curve of the capacitor can be obtained through simulation testing, thereby calculating the equivalent series inductance of the decoupling capacitor. Substituting the equivalent series inductance into The minimum number of capacitors required to meet the target impedance is calculated. Round to n=24.
[0056] In this embodiment, the maximum capacitance of the decoupling capacitor is selected starting from 10nf. Through S parameter simulation software such as PowerSI software, the S parameter file of the VDDQ power distribution network without adding decoupling capacitors is simulated, and this file is used to establish an S parameter model containing VDDQ power distribution network information in SPICE. According to the capacitor sub-circuit model file containing capacitor parameters provided by the capacitor manufacturer, the 10nf capacitor sub-circuit model and the S parameter model of the VDDQ power distribution network are cascaded into a power network in SPICE, and simulation is performed to obtain the impedance curve of the VDDQ power distribution network after adding the 10nf decoupling capacitor. This impedance curve is compared with the impedance curve without adding the 10nf decoupling capacitor. If the impedance peak of the impedance curve after adding the 10nf decoupling capacitor is significantly improved compared with the impedance peak of the impedance curve before adding, it indicates that adding a decoupling capacitor with the current capacitance value is effective. At this time, a 10nf decoupling capacitor can be cascaded in SPICE, and the impedance curve is simulated and compared. When adding a decoupling capacitor with the current capacitance value does not significantly improve the impedance curve, a decoupling capacitor with a capacitance of 4.7nf is added and simulation comparison is performed. Some decoupling capacitors with capacitance values that do not improve impedance after adding can be skipped directly, and the next decoupling capacitor with capacitance value can be added. When adding decoupling capacitors, the total number needs to be limited to the minimum number of capacitors required to meet the target impedance in theory. If the impedance requirement is still not met after exceeding the minimum number of decoupling capacitors required to meet the target impedance in theory, the number of large-capacitance decoupling capacitors can be appropriately adjusted. Because small-capacitance decoupling capacitors not only have a major effect on high-frequency pdn impedance, they will also affect low-frequency pdn impedance. Therefore, after adding a sufficient number of small-capacitance decoupling capacitors, the number of large-capacitance decoupling capacitors can be appropriately reduced. At this time, the actual low-frequency impedance can still meet the target impedance requirements. According to this method, a sufficient number of capacitors of various capacitance values are added until the peak of the actual impedance curve is lower than the target impedance, such as Figure 3 shown.
[0057] By iterating the above method multiple times, we can get a capacitor adding method that uses the minimum number of capacitors to achieve the target PDN impedance. This method is mainly simulated through SPICE, which can greatly improve the simulation efficiency.
[0058] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit the technical solution. Those skilled in the art should understand that those modifications or equivalent substitutions of the technical solution of the present invention that do not depart from the purpose and scope of the technical solution should be included in the scope of the claims of the present invention.
Claims
1. A method for efficiently improving the impedance of a package power distribution network, characterized in that: The following steps are involved: Step 1) determining the target impedance, the highest cutoff frequency and the lowest cutoff frequency of the power distribution network; Step 2) simulating and obtaining the planar capacitance of the package power distribution network, and calculating the capacitance of the required minimum decoupling capacitor based on the planar capacitance; Step 3) simulating and obtaining the actual impedance-operating frequency curve of the package power distribution network, and determining the required maximum decoupling capacitor capacitance value according to the actual impedance-operating frequency curve, and combining the required minimum decoupling capacitor capacitance value obtained in step 2) to obtain the required decoupling capacitor capacitance value range; Step 4) calculating the minimum number of decoupling capacitors required to meet the target impedance; Step 5) according to the capacitance range of the decoupling capacitor determined in step 3) and the minimum number of the decoupling capacitor determined in step 4), the decoupling capacitor and the package power distribution network are simulated until the peak value of the actual impedance curve obtained is lower than the target impedance; Step 6) Obtain the capacitance and quantity of decoupling capacitors that need to be added to the package power distribution network.
2. The method for efficiently improving the impedance of a package power distribution network according to claim 1, characterized in that: The calculation formula of the target impedance in step 1) is: Where: Z target is the target impedance, V DD is the supply voltage of a specific rail of the power distribution network, ripple% is the allowable ripple, I max The maximum peak current for a specific rail in a power distribution network.
3. The method for efficiently improving the impedance of a package power distribution network according to claim 1, characterized in that: In step 2), one third of the capacitance of the planar capacitor is taken as a reference capacitance to obtain the capacitance of a capacitor closest to the reference capacitance, and the capacitance of the capacitor is used as the capacitance of the required minimum decoupling capacitor.
4. The method for efficiently improving the impedance of a package power distribution network according to claim 1, characterized in that: In step 3), the actual impedance-operating frequency curve obtained by simulation is compared with the target impedance, and the lowest frequency at which the actual impedance peak exceeds the target impedance is obtained as the first frequency. According to the impedance-frequency curve of the capacitor, the frequency range in which the capacitor behaves as a capacitor is obtained, and the capacitance of the largest capacitor that behaves as a capacitor at the first frequency is selected as the capacitance of the required maximum decoupling capacitor.
5. The method for efficiently improving the impedance of a package power distribution network according to claim 4, characterized in that: In step 3), the capacitance range of all available decoupling capacitors is determined according to the capacitance of the maximum decoupling capacitor and the capacitance of the minimum decoupling capacitor; and the capacitance of each decoupling capacitor is determined according to the geometric mean of the capacitances of its two adjacent decoupling capacitors.
6. The method for efficiently improving the impedance of a package power distribution network according to claim 1, characterized in that: In step 4), the minimum number of decoupling capacitors required is calculated using the following formula: Where: n is the minimum number of decoupling capacitors required, F max is the highest cutoff frequency of the power distribution network, ESL is the equivalent series inductance of the decoupling capacitor, Z target is the target impedance.
7. The method for efficiently improving the impedance of a package power distribution network according to claim 1, characterized in that: In step 5), an S-parameter file containing the characteristics of the package power distribution network is generated through simulation, and a cascade circuit model is formed in SPICE with a sub-circuit file containing the RLC parameters of the decoupling capacitor to equivalently add the power distribution network after the decoupling capacitor is added, thereby realizing the simulation of the decoupling capacitor and the power distribution network.
8. The method for efficiently improving the impedance of a package power distribution network according to claim 7, characterized in that: Decoupling capacitors within the capacitance range are cascaded in descending order with the S-parameter file containing the characteristics of the package power distribution network. The number of decoupling capacitors with different capacitance values to be added is determined by the impedance peak change of the impedance curve after adding the decoupling capacitor of the capacitance value is less than the set value. When the impedance peak of the impedance curve after adding a decoupling capacitor of a certain capacitance value is less than the set value, decoupling capacitors with adjacent capacitance values are continued to be added for cascading until the actual impedance curve peak value is less than the target impedance.
