CN102890736B - Measuring method for resistance of three-terminal resistors used for integrated circuits - Google Patents

Abstract

The invention discloses a measuring method for the resistance of three-terminal resistors used for integrated circuits, comprising the following steps that: (10) a macro-model of the three-terminal resistors used for integrated circuits is established; (20) a test structure of the three-terminal resistors used for integrated circuits is constructed; (30) a diode test data file and a three-terminal resistor test data file are established; (40) a model of the alternating current characteristic of the three-terminal resistors used for integrated circuits is established; (50) a model of the direct current characteristic of the three-terminal resistors used for integrated circuits is established; (60) a model file of the three-terminal resistors is established; and (70) the resistance characteristic of the three-terminal resistors in integrated circuits is measured. The measuring method is simple and effective and can solve the problems that the simulation method which is based on a three-terminal resistor physical model R3cmc of the existing integrated circuit engineering is excessively complex and the degree of fitting to size is poor.

Description

The measuring method of three end resistances for a kind of integrated circuit

Technical field

The invention belongs to microelectronic component emulation field, specifically, relate to the measuring method of a kind of integrated circuit with three end resistances.

Background technology

Resistance is a kind of important semiconductor devices in integrated circuit, is widely used in integrated circuit fields.Conventionally the resistance emulation of using in IC project is all the two ends Resistance model for prediction using, and this models fitting parameter is little, carries mold process simple, but it does not consider the impact of substrate bias for resistance.Although but use the simple out of true very of the resistance emulation mode of two ends Resistance model for prediction, therefore the resistance emulation mode of two ends Resistance model for prediction can not well instruct integrated circuit (IC) design and emulation.

In addition, as far back as 2007, the international compact models council (compact model council) has just issued the physical model R3_cmc of a kind of integrated circuit with three end resistance, the physical characteristics of this model based on three end resistance, has considered that the self-heating effect of resistance, speed are saturated, the impact of statistics variations etc.R3_cmc is a kind of nonlinear compact Resistance model for prediction, and its resistance all presents nonlinear characteristic with the variation of resistance three end institute making alives, physical dimension.But the resistance emulation mode based on R3_cmc model is not widely used by industry, its reason is: 1, R3_cmc has more than 90 parameter, makes the simulation process of resistance too complicated loaded down with trivial details; 2, well three end resistance of the good different size of matching simultaneously of R3_cmc model.Therefore,, although existing integrated circuit based on R3_cmc model is more accurate too complicated by the emulation mode of three end resistance, can not effectively in engineering, apply.

Summary of the invention

Technical matters: technical matters to be solved by this invention is: the measuring method of a kind of integrated circuit with three end resistances is provided, this measuring method is simply effective, can solve too complexity and the problem poor to size degree of fitting of the emulation mode based on three end resistance physical model R3_cmc in existing IC project.

Technical scheme: in order to solve the problems of the technologies described above, the present invention adopts following technical scheme:

A measuring method for three end resistances for integrated circuit, this measuring method comprises the following steps:

Step 10) set up the macro model of integrated circuit with three end resistance: this macro model is a circuit that contains source, drain terminal and three ports of substrate terminal, wherein, between source and drain terminal, be connected a resistance, between source and substrate terminal and between drain terminal and substrate terminal, be connected respectively a parasitic diode, this macro model forms three end resistance;

Step 20) build the integrated circuit test structure of three end resistance: according to the macro model of the integrated circuit three end resistance of step 10) foundation, build the test structure of integrated circuit three end resistance, this test structure comprises three end resistance of different size, and different size, with three end resistance with under technique, complete and there is the diode of same structure with the parasitic diode of three end resistance;

Step 30) set up Test Diode data file and three end resistance test data files: according to step 20) in the test structure of integrated circuit three end resistance that builds, be made into chip, then test diode in this chip and the electrology characteristic of three end resistance, form Test Diode data file and three end resistance test data files;

Step 40) set up the model of integrated circuit by the AC characteristic of three end resistance: carrying in mould software, first preference pattern type is diode model, carry mould software and automatically generate diode simulation curve according to diode model, then be written into step 30) the Test Diode data file that obtains, carry mould software and form Test Diode curve according to Test Diode data file; Then adjust the model parameter in diode model, change diode simulation curve, until the root-mean-square error that diode simulation curve and Test Diode curve obtain is less than diode matching setting value, finally preserve the model parameter of diode model, the model parameter of this diode model is the model of integrated circuit by the AC characteristic of three end resistance;

Step 50) set up the model of integrated circuit by the DC characteristic of three end resistance:

Step 501) three end resistances are carried out to the correction of temperature, size, substrate bias, set up the model file of three end resistance DC characteristic;

Tempc=temper-25 (formula 1)

Teff=1+tc1 × tempc+tc2 × tempc ²(formula 2)

In formula, tc1 represents the single order fitting parameter of temperature, tc2 represents the second-order fit parameter of temperature, temper represents the probe temperature of three end resistance, tempc represents the difference of three end resistance test temperature and room temperature, the unit of temper and tempc is degree Celsius, and teff represents the influence coefficient of temperature to three end resistances; Wherein, tc1 and tc2 are the model parameters in the model of three end resistance DC characteristic;

$\mathrm{reff}=\mathrm{rsh}\×\frac{l-\mathrm{dl}}{w-\mathrm{dw}}\×\mathrm{teff}\×(1+\frac{\mathrm{pvc}1}{(l-\mathrm{dl})\×{10}^6}\×|V_{\mathrm{SD}}|+\frac{\mathrm{pvc}2}{{(l-\mathrm{dl})}^2\×{10}^{12}}\×{V_{\mathrm{SD}}}^2)$

(formula 3)

$\×(1+(\mathrm{pv}1+\frac{\mathrm{pv}1l}{l-\mathrm{dl}}+\frac{\mathrm{pv}1w}{w-\mathrm{dw}})\×|V_{\mathrm{SS}}\left|\right)$

In formula: reff represents the DC resistance of three end resistance, unit is ohm; Rsh represents square resistance, and unit is ohms/square; L represents the length of three end resistance, and w represents the width of three end resistance, and dl represents the length offset amount of three end resistance, and dw represents the width offset of three end resistance, and the unit of l, w, dl and dw is micron; L-dl represents the effective length of three end resistance, and w-dw represents the effective width of three end resistance; V _sDrepresent voltage between three end resistance sources and drain terminal, unit is volt; Pvc1 represents V _sDfor the first order influence coefficient of three end resistances, and pvcl is relevant to the effective length of three end resistance; Pvc2 represents V _sDfor the second order influence coefficient of three end resistances, and pvc2 is relevant to the effective length of three end resistance; V _sSrepresent the voltage between three end resistance sources and substrate terminal, unit is volt; Pv1l represents V _sSfor the first order influence coefficient of three end resistances, and pv1l is relevant to the effective length of three end resistance; Pv1w represents V _sSfor the first order influence coefficient of three end resistances, and pv1w is relevant to the effective width of three end resistance; Pv1 represents V _sSto the first order influence coefficient of three end resistances; Wherein, rsh, dl, dw, pvc1, pvc2, pv1l, pv1w and pv1 are the model parameters in the model of three end resistance DC characteristic;

The model parameter of the model of the DC characteristic to three end resistance is given initial value, then be based upon the model file of putting forward the three end resistance DC characteristic that mould software can identify, this model file comprises initial value, formula (1), formula (2) and the formula (3) of model parameter;

Step 502) create and carry the three end Resistance model for prediction files that mould software can be identified, this model file comprises macro model and the step 501 that step 10) is set up) in the model of three end resistance DC characteristic in model parameter, and by this model file and step 501) model file of the three end resistance DC characteristic that obtain is placed under same file folder;

Step 503) carrying in mould software, first preference pattern type is three end Resistance model for prediction, be written into step 30) the three end resistance test data files that obtain, carry mould software and form three end resistance test curves according to three end resistance test data files, then be written into step 502) the middle three end Resistance model for prediction files that create, carry mould software and automatically generate three end resistance simulation curves according to this model file, then set-up procedure 502) in model parameter in the three end Resistance model for prediction files that create, change three end resistance simulation curves, until the root-mean-square error that three end resistance simulation curves and three end resistance test curves obtain is less than three end Resistance Fitting setting values, finally preserve step 502) in three end Resistance model for prediction files in model parameter, this model parameter is the model of integrated circuit by the DC characteristic of three end resistance,

Step 60) set up three end Resistance model for prediction files: by step 40) model and the step 50 of the integrated circuit AC characteristic of three end resistance that obtains) the integrated circuit model of the DC characteristic of three end resistance that obtains integrates, make the three end Resistance model for prediction files that comprise three end resistance DC and AC characteristics for integrated circuit simulating software, in this three ends Resistance model for prediction file, the girth of two parasitic diodes is w+l, and area is w × l/2;

Step 70) measure the resistance characteristic of three end resistance in integrated circuit: in integrated circuit simulating software, for three end resistance in integrated circuit simulating, invocation step 60) the three end Resistance model for prediction files that obtain, the resistance characteristic of three end resistance under alternating current-direct current signal function in integrated circuit measured.

Beneficial effect: compared with prior art, the present invention has following beneficial effect:

1. the parameter using in measuring process is few, makes measuring process simple to operation.The existing resistance measurement method based on R3cmc model has more than 90 parameter, makes the measuring process of resistance too complicated loaded down with trivial details.And only have 10 parameters in the measuring method that the present invention adopts, be respectively tc1, tc2, rsh, dl, dw, pvc1, pvc2, pv1l, pv1w and pv1.Like this, in the process that integrated circuit is measured with three end resistances, the parameter of use is few, and measuring process is just more simple and efficient.The clear of 10 parameters that meanwhile, use in the present invention.

2. the degree of accuracy that pair integrated circuit is measured with three end resistances is high.The existing resistance measurement method based on R3_cmc model is three end resistance of the different size of matching simultaneously well.In the measuring method of integrated circuit of the present invention with three end resistances, considered temperature, size especially substrate bias for the impact of resistance, three end resistance under the different temperatures of emulation simultaneously, different size, different bias voltage, make the emulation of resistance more approach the actual working characteristics of resistance in integrated circuit, effectively raise the accuracy of resistance emulation.In addition, measuring method of the present invention has been considered the impact of three end resistance two ends parasitic diodes, makes the DC characteristic that the method not only can emulation three end resistance, and AC characteristic that can well emulation three end resistance, has further improved resistance.

3. be suitable for applying in engineering.The measuring method of three end resistances for integrated circuit of the present invention, parameter is few, measures three end resistance degree of accuracy high, is suitable for the enterprising line integrated circuit design of engineering and emulation.

Brief description of the drawings

Fig. 1 is process flow diagram of the present invention.

Fig. 2 is the structural drawing of macro model in step 10) of the present invention.

Fig. 3 is in the embodiment of the present invention 1, the test curve of diode " capacitance-voltage " characteristic and the fitted figure of simulation curve.

Fig. 4 is in the embodiment of the present invention 2, the test curve of diode " current-voltage " characteristic and the fitted figure of simulation curve.

Fig. 5 is in the embodiment of the present invention 3, three end resistance " resistance-V _sD" test curve of characteristic and the fitted figure of simulation curve.

Fig. 6 is in the embodiment of the present invention 4, three end resistance " resistance-V _sD" test curve of characteristic and the fitted figure of simulation curve.

Fig. 7 is in the embodiment of the present invention 5, three end resistance " resistance-V _sD" test curve of characteristic and the fitted figure of simulation curve.

Fig. 8 is in comparative example 1 of the present invention, three end resistance " resistance-V _sD" test curve of characteristic and the fitted figure of simulation curve.

Fig. 9 is in comparative example 2 of the present invention, three end resistance " resistance-V _sD" test curve of characteristic and the fitted figure of simulation curve.

Figure 10 is in comparative example 3 of the present invention, three end resistance " resistance-V _sD" test curve of characteristic and the fitted figure of simulation curve.

Embodiment

Further illustrate technical scheme of the present invention below in conjunction with accompanying drawing.

As shown in Figure 1, the measuring method of three end resistances for a kind of integrated circuit of the present invention, comprises the following steps:

Step 10) set up the macro model of integrated circuit with three end resistance: this macro model is a circuit that contains source, drain terminal and three ports of substrate terminal, wherein, between source and drain terminal, be connected a resistance, between source and substrate terminal and between drain terminal and substrate terminal, be connected respectively a parasitic diode, this macro model forms three end resistance.

As shown in Figure 2, macro model has three ports, is respectively S end, D end and Sub end, and S end is source, and D end is drain terminal, and Sub end is substrate terminal, and D1 and D2 represent respectively parasitic diode.

Step 20) build the integrated circuit test structure of three end resistance: according to the macro model of the integrated circuit three end resistance of step 10) foundation, build the test structure of integrated circuit three end resistance, this test structure comprises three end resistance of different size, and different size, with three end resistance with under technique, complete and there is the diode of same structure with the parasitic diode of three end resistance.

Step 30) set up Test Diode data file and three end resistance test data files: according to step 20) in the test structure of integrated circuit three end resistance that builds, be made into chip, then test diode in this chip and the electrology characteristic of three end resistance, form Test Diode data file and three end resistance test data files.

In step 30) in, test the electrology characteristic of the diode in this chip, comprise " current-voltage " family curve and " electric capacity-voltage " family curve that test different size diode obtains under different probe temperatures, form Test Diode data file.It is the voltage V that 0-5V, step-length are 0.1V that the electrology characteristic of testing three end resistance in this chip is included in loading range between source and drain terminal _sD, between substrate terminal and source, loading range is that 0--5V, step-length are-the voltage V of 1V _sS, the test different size three " resistance-Vs of end resistance under different probe temperatures _sD" family curve, form three end resistance test data files.Different probe temperatures, for example, choose 125 DEG C, 80 DEG C, 25 DEG C ,-40 DEG C ,-85 DEG C.

According to proposing the requirement of mould software to three end resistance datas, if desired, the form of Update Table file.For example, test data is written into the English full name of MBP(: Model Builder program, a kind of instrument of proposing mould and modeling for integrated circuit (IC)-components level of being developed by Accelicon company) when software, need change the DataType{n} in test data file into DataType{r}.

Step 40) set up the model of integrated circuit by the AC characteristic of three end resistance: carrying in mould software, first preference pattern type is diode model, carries mould software and automatically generates diode simulation curve according to diode model; Then be written into step 30) the Test Diode data file that obtains, carry mould software and form Test Diode curve according to Test Diode data file; Then adjust the model parameter in diode model, change diode simulation curve, until the root-mean-square error that diode simulation curve and Test Diode curve obtain is less than diode matching setting value, finally preserve the model parameter of diode model, the model parameter of this diode model is the model of integrated circuit by the AC characteristic of three end resistance.

Diode matching setting value can, according to the design needs of integrated circuit, be predetermined a numerical value.When the root-mean-square error obtaining when diode simulation curve and Test Diode curve is less than this setting value, just stops adjusting the model parameter in diode model, and preserve the model parameter of diode model now.

The expression formula of root-mean-square error RMS is $\mathrm{RMS}=\sqrt{\frac{1}{N}\underset{i=1}{\overset{N}{\mathrm{\Σ}}}{\left(\frac{\mathrm{mes}\left(i\right)-\mathrm{sim}\left(i\right)}{\max (\mathrm{mes}\left(i\right),\mathrm{sim}\left(i\right))}\right)}^2},$ Wherein, N represents counting of test data, and mes (i) represents the test data of i test point, and sim (i) represents the emulated data of i test point.

Step 50) set up the model of integrated circuit by the DC characteristic of three end resistance:

Step 501) three end resistances are carried out to the correction of temperature, size, substrate bias, set up the model file of three end resistance DC characteristic;

Tempc=temper-25 (formula 1)

Teff=1+tc1 × tempc+tc2 × tempc ²(formula 2)

In formula, tc1 represents the single order fitting parameter of temperature, tc2 represents the second-order fit parameter of temperature, temper represents the probe temperature of three end resistance, tempc represents the difference of three end resistance test temperature and room temperature, the unit of temper and tempc is degree Celsius, and teff represents the influence coefficient of temperature to three end resistances; Wherein, tc1 and tc2 are the model parameters in the model of three end resistance DC characteristic;

$\mathrm{reff}=\mathrm{rsh}\×\frac{l-\mathrm{dl}}{w-\mathrm{dw}}\×\mathrm{teff}\×(1+\frac{\mathrm{pvc}1}{(l-\mathrm{dl})\×{10}^6}\×|V_{\mathrm{SD}}|+\frac{\mathrm{pvc}2}{{(l-\mathrm{dl})}^2\×{10}^{12}}\×{V_{\mathrm{SD}}}^2)$

(formula 3)

$\×(1+(\mathrm{pv}1+\frac{\mathrm{pv}1l}{l-\mathrm{dl}}+\frac{\mathrm{pv}1w}{w-\mathrm{dw}})\×|V_{\mathrm{SS}}\left|\right)$

In formula: reff represents the DC resistance of three end resistance, unit is ohm; Rsh represents square resistance, and unit is ohms/square; L represents the length of three end resistance, and w represents the width of three end resistance, and dl represents the length offset amount of three end resistance, and dw represents the width offset of three end resistance, and the unit of l, w, dl and dw is micron; L-dl represents the effective length of three end resistance, and w-dw represents the effective width of three end resistance; V _sDrepresent voltage between three end resistance sources and drain terminal, unit is volt; Pvc1 represents V _sDfor the first order influence coefficient of three end resistances, and pvc1 is relevant to the effective length of three end resistance; Pvc2 represents V _sDfor the second order influence coefficient of three end resistances, and pvc2 is relevant to the effective length of three end resistance; V _sSrepresent the voltage between three end resistance sources and substrate terminal, unit is volt; Pv1l represents V _sSfor the first order influence coefficient of three end resistances, and pv1l is relevant to the effective length of three end resistance; Pv1w represents V _sSfor the first order influence coefficient of three end resistances, and pv1w is relevant to the effective width of three end resistance; Pv1 represents V _sSto the first order influence coefficient of three end resistances; Wherein, rsh, dl, dw, pvc1, pvc2, pv1l, pv1w and pv1 are the model parameters in the model of three end resistance DC characteristic.

The model parameter of the model of the DC characteristic to three end resistance is given initial value, then be based upon the model file of putting forward the three end resistance DC characteristic that mould software can identify, this model file comprises initial value, formula (1), formula (2) and the formula (3) of model parameter.

In step 501) in, the model parameter of the model of the DC characteristic of three end resistance is given initial value, is preferably: rsh=1, dl=0u, dw=0u, tc1=0, tc2=0, pvc1=0, pvc2=0, pv1=0, pv1l=0, pv1w=0.

Step 502) create and carry the three end Resistance model for prediction files that mould software can be identified, this model file comprises macro model and the step 501 that step 10) is set up) in the model of three end resistance DC characteristic in model parameter, and by this model file and step 501) model file of the three end resistance DC characteristic that obtain is placed under same file folder.

Step 503) carrying in mould software, first preference pattern type is three end Resistance model for prediction, is written into step 30) the three end resistance test data files that obtain, carry mould software and form three end resistance test curves according to three end resistance test data files; Then be written into step 502) the middle three end Resistance model for prediction files that create, carry mould software and automatically generate three end resistance simulation curves according to this model file, then set-up procedure 502) in model parameter in the three end Resistance model for prediction files that create, change three end resistance simulation curves, until the root-mean-square error that three end resistance simulation curves and three end resistance test curves obtain is less than three end Resistance Fitting setting values, finally preserve step 502) in three end Resistance model for prediction files in model parameter, this model parameter is the model of integrated circuit by the DC characteristic of three end resistance.

Three end Resistance Fitting setting values can, according to the design needs of integrated circuit, be predetermined a numerical value.When the root-mean-square error obtaining when three end resistance simulation curves and three end resistance test curves is less than this three ends Resistance Fitting setting value, just stop adjusting the model parameter in three end Resistance model for prediction files, and preserve the model parameter in three end Resistance model for prediction files now.

Step 60) set up three end Resistance model for prediction files: by step 40) model and the step 50 of the integrated circuit AC characteristic of three end resistance that obtains) the integrated circuit model of the DC characteristic of three end resistance that obtains integrates, make the three end Resistance model for prediction files that comprise three end resistance DC and AC characteristics for integrated circuit simulating software, in this three ends Resistance model for prediction file, the girth of two parasitic diodes is w+l, and area is w × l/2.

Step 70) measure the resistance characteristic of three end resistance in integrated circuit: in integrated circuit simulating software, for three end resistance in integrated circuit simulating, invocation step 60) the three end Resistance model for prediction files that obtain, the resistance characteristic of three end resistance under alternating current-direct current signal function in integrated circuit measured.

Below by the technical program and prior art are contrasted, the technical program premium properties is described.

Embodiment 1.

As shown in Figure 3, be that 6 μ m, length are the diode of 60 μ m to width, in the time that probe temperature is 25 DEG C, adopt test curve and the simulation curve fitted figure of " capacitance-voltage " characteristic that measuring method of the present invention obtains.Horizontal ordinate represents voltage, and unit is volt; Ordinate represents electric capacity, and unit is farad.In figure, [M] represents test curve, and [S] represents simulation curve.In Fig. 3, the root-mean-square error RMS that matching test curve and simulation curve reach is 0.731%.

Embodiment 2.

As shown in Figure 4, be that 6 μ m, length are the diode of 60 μ m to width, in the time that probe temperature is 25 DEG C, adopt test curve and the simulation curve fitted figure of " current-voltage " characteristic that measuring method of the present invention obtains.Horizontal ordinate represents voltage, and unit is volt; Ordinate represents electric current, and unit is ampere.In figure, [M] represents test curve, and [S] represents simulation curve.In Fig. 4, the root-mean-square error RMS reaching when matching test curve and simulation curve is 1.332%.

Embodiment 3.

As shown in Figure 5, be that 5 μ m, length are the three end resistance of 100 μ m to width, be 25 DEG C at probe temperature, V _sSbe respectively 0V ,-2V and-when 4V, adopt step 50 in measuring method of the present invention) three end resistance " resistance-V of obtaining _sD" test curve and the simulation curve fitted figure of characteristic.Horizontal ordinate represents voltage V _sD, unit is volt; Ordinate represents resistance, and unit is ohm.In figure, [M] represents test curve, and [S] represents simulation curve, RMS[V _sS=0V] expression V _sSduring for 0V, the root-mean-square error of test curve and simulation curve matching.RMS[V _sS=-2V] expression V _sSduring for-2V, the root-mean-square error of test curve and simulation curve matching.RMS[V _sS=-4V] expression V _sSduring for-4V, the root-mean-square error of test curve and simulation curve matching.In Fig. 5, V _sSthe root-mean-square error that test curve during for 0V and simulation curve matching reach is 0.826%; V _sSthe root-mean-square error that test curve during for-2V and simulation curve matching reach is 0.432%; V _sSthe root-mean-square error that test curve during for-4V and simulation curve matching reach is 0.440%.

Embodiment 4.

As shown in Figure 6, be that 6 μ m, length are the three end resistance of 40 μ m to width, be 25 DEG C at probe temperature, V _sSbe respectively 0V ,-2V and-when 4V, adopt step 50 in measuring method of the present invention) three end resistance " resistance-V of obtaining _sD" test curve and the simulation curve fitted figure of characteristic.Horizontal ordinate represents voltage V _sD, unit is volt; Ordinate represents resistance, and unit is ohm.In figure, [M] represents test curve, and [S] represents simulation curve.RMS[V _sS=0V] expression V _sSduring for 0V, the root-mean-square error of test curve and simulation curve matching.RMS[V _sS=-2V] expression V _sSduring for-2V, the root-mean-square error of test curve and simulation curve matching.RMS[V _sS=-4V] expression V _sSduring for-4V, the root-mean-square error of test curve and simulation curve matching.In Fig. 6, V _sSthe root-mean-square error that test curve during for 0V and simulation curve matching reach is 0.401%; V _sSthe root-mean-square error that test curve during for-2V and simulation curve matching reach is 0.283%; V _sSthe root-mean-square error that test curve during for-4V and simulation curve matching reach is 0.232%.

Embodiment 5.

As shown in Figure 7, be that 5 μ m, length are the three end resistance of 100 μ m to width, be-40 DEG C at probe temperature, V _sSbe respectively 0V ,-2V and-when 4V, adopt step 50 in measuring method of the present invention) three end resistance " resistance-V of obtaining _sD" test curve and the simulation curve fitted figure of characteristic.Horizontal ordinate represents voltage V _sD, unit is volt; Ordinate represents resistance, and unit is ohm.In figure, [M] represents test curve, and [S] represents simulation curve, RMS[V _sS=0V] expression V _sSduring for 0V, the root-mean-square error of test curve and simulation curve matching.RMS[V _sS=-2V] expression V _sSduring for-2V, the root-mean-square error of test curve and simulation curve matching.RMS[V _sS=-4V] expression V _sSduring for-4V, the root-mean-square error of test curve and simulation curve matching.In Fig. 7, V _sSthe root-mean-square error that test curve during for 0V and simulation curve matching reach is 1.051%; V _sSthe root-mean-square error that test curve during for-2V and simulation curve matching reach is 0.692%; V _sSthe root-mean-square error that test curve during for-4V and simulation curve matching reach is 0.834%.

Comparative example 1.

As shown in Figure 8, be that 5 μ m, length are the three end resistance of 100 μ m to width, be 25 DEG C at probe temperature, V _sSbe respectively 0V ,-2V and-when 4V, adopt the measuring method of three end resistance for the integrated circuit based on model R3cmc, the three end resistance " resistance-V that obtain _sD" test curve and the simulation curve fitted figure of characteristic.Horizontal ordinate represents voltage V _sD, unit is volt; Ordinate represents resistance, and unit is ohm.In figure, [M] represents test curve, and [S] represents simulation curve, RMS[V _sS=0V] expression V _sSduring for 0V, the root-mean-square error of test curve and simulation curve matching.RMS[V _sS=-2V] expression V _sSduring for-2V, the root-mean-square error of test curve and simulation curve matching.RMS[V _sS=-4V] expression V _sSduring for-4V, the root-mean-square error of test curve and simulation curve matching.In Fig. 8, V _sSthe root-mean-square error that test curve during for 0V and simulation curve matching reach is 1.781%; V _sSthe root-mean-square error that test curve during for-2V and simulation curve matching reach is 1.033%; V _sSthe root-mean-square error that test curve during for-4V and simulation curve matching reach is 2.791%.

Comparative example 2.

As shown in Figure 9, be that 6 μ m, length are the three end resistance of 40 μ m to width, be 25 DEG C at probe temperature, V _sSbe respectively 0V ,-2V and-when 4V, adopt the measuring method of three end resistance for the integrated circuit based on model R3cmc, the three end resistance " resistance-V that obtain _sD" test curve and the simulation curve fitted figure of characteristic.Horizontal ordinate represents voltage V _sD, unit is volt; Ordinate represents resistance, and unit is ohm.In figure, [M] represents test curve, and [S] represents simulation curve, RMS[V _sS=0V] expression V _sSduring for 0V, the root-mean-square error of test curve and simulation curve matching.RMS[V _sS=-2V] expression V _sSduring for-2V, the root-mean-square error of test curve and simulation curve matching.RMS[V _sS=-4V] expression V _sSduring for-4V, the root-mean-square error of test curve and simulation curve matching.In Fig. 9, V _sSthe root-mean-square error that test curve during for 0V and simulation curve matching reach is 0.908%; V _sSthe root-mean-square error that test curve during for-2V and simulation curve matching reach is 1.153%; V _sSthe root-mean-square error that test curve during for-4V and simulation curve matching reach is 1.188%.

Comparative example 3.

As shown in figure 10, be that 5 μ m, length are the three end resistance of 100 μ m to width, be-40 DEG C at probe temperature, V _sSbe respectively 0V ,-2V and-when 4V, adopt the measuring method of three end resistance for the integrated circuit based on model R3cmc, the three end resistance " resistance-V that obtain _sD" test curve and the simulation curve fitted figure of characteristic.Horizontal ordinate represents voltage V _sD, unit is volt; Ordinate represents resistance, and unit is ohm.In figure, [M] represents test curve, and [S] represents simulation curve, RMS[V _sS=0V] expression V _sSduring for 0V, the root-mean-square error of test curve and simulation curve matching.RMS[V _sS=-2V] expression V _sSduring for-2V, the root-mean-square error of test curve and simulation curve matching.RMS[V _sS=-4V] expression V _sSduring for-4V, the root-mean-square error of test curve and simulation curve matching.In Figure 10, V _sSthe root-mean-square error that test curve during for 0V and simulation curve matching reach is 0.933%; V _sSthe root-mean-square error that test curve during for-2V and simulation curve matching reach is 1.764%; V _sSthe root-mean-square error that test curve during for-4V and simulation curve matching reach is 1.857%.

Can find out from Fig. 3 and Fig. 4: the AC characteristic modeling of diode, the simulation curve that the integrated circuit that uses the present invention to propose obtains by the measuring method of three end resistance resistance characteristics all can overlap substantially from the test curve that test under different size, different probe temperature obtains, the root-mean-square error that matching reaches, all in 1% left and right, can reach very high degree of fitting.

Can find out from Fig. 5, Fig. 6 and Fig. 7: three end resistance DC characteristic modelings, the simulation curve that the integrated circuit that uses the present invention to propose obtains by the measuring method of three end resistance resistance characteristics all can overlap substantially from the test curve that test under different size, different probe temperature obtains, the root-mean-square error that matching reaches nearly all, in 1%, can reach very high degree of fitting.

Thus, the method that the present invention proposes can be measured the resistance characteristic of integrated circuit with three end resistance very accurately.

Can find out from Fig. 8, Fig. 9 and Figure 10: based on the measuring method of model R3cmc, the root-mean-square error arriving when simulation curve and test curve matching is almost all in 2% left and right.

Comprehensive Correlation Fig. 3 to Figure 10, the method that the present invention proposes, compared with the measuring method based on model R3cmc, is not only carried mould parameter still less, and measuring process is simple to operation, and has very high accuracy.Therefore the integrated circuit that, the present invention proposes is more suitable for applying in engineering by the measuring method of three end resistances.

Claims (2)

1. a measuring method for three end resistances for integrated circuit, is characterized in that: this measuring method comprises the following steps:

Step 10) set up the macro model of integrated circuit with three end resistance: this macro model is a circuit that contains source, drain terminal and three ports of substrate terminal, wherein, between source and drain terminal, be connected a resistance, between source and substrate terminal and between drain terminal and substrate terminal, be connected respectively a parasitic diode, this macro model forms three end resistance;

Step 20) build the integrated circuit test structure of three end resistance: according to step 10) macro model of the integrated circuit three end resistance of foundation, build the test structure of integrated circuit three end resistance, this test structure comprises three end resistance of different size, and different size, with three end resistance with under technique, complete and there is the diode of same structure with the parasitic diode of three end resistance;

Step 30) set up Test Diode data file and three end resistance test data files: according to step 20) in the test structure of integrated circuit three end resistance that builds, be made into chip, then test diode in this chip and the electrology characteristic of three end resistance, form Test Diode data file and three end resistance test data files; Test the electrology characteristic of the diode in this chip, comprise " electric current-voltage " family curve and " electric capacity-voltage " family curve that test different size diode obtains under different probe temperatures, form Test Diode data file;

Described step 30) in, it is the voltage V that 0-5V, step-length are 0.1V that the electrology characteristic of testing three end resistance in this chip is included in loading range between source and drain terminal _sD, between substrate terminal and source, loading range is that 0--5V, step-length are-the voltage V of 1V _sS, the test different size three " resistance-Vs of end resistance under different probe temperatures _sD" family curve, form three end resistance test data files;

Step 40) set up the model of integrated circuit by the AC characteristic of three end resistance: carrying in mould software, first preference pattern type is diode model, carry mould software and automatically generate diode simulation curve according to diode model, then be written into step 30) the Test Diode data file that obtains, carry mould software and form Test Diode curve according to Test Diode data file; Then adjust the model parameter in diode model, change diode simulation curve, until the root-mean-square error that diode simulation curve and Test Diode curve obtain is less than diode matching setting value, finally preserve the model parameter of diode model, the model parameter of this diode model is the model of integrated circuit by the AC characteristic of three end resistance;

Step 50) set up the model of integrated circuit by the DC characteristic of three end resistance:

Step 501) three end resistances are carried out to the correction of temperature, size, substrate bias, set up the model file of three end resistance DC characteristic;

Tempc=temper-25 (formula 1)

Teff=1+tc1 × tempc+tc2 × tempc ²(formula 2)

In formula, tc1 represents the single order fitting parameter of temperature, tc2 represents the second-order fit parameter of temperature, temper represents the probe temperature of three end resistance, tempc represents the difference of three end resistance test temperature and room temperature, the unit of temper and tempc is degree Celsius, and teff represents the influence coefficient of temperature to three end resistances; Wherein, tc1 and tc2 are the model parameters in the model of three end resistance DC characteristic;

$\begin{array}{c}\mathrm{reff}=\mathrm{rsh}\×\frac{l-\mathrm{dl}}{w-\mathrm{dw}}\×\mathrm{teff}\×(1+\frac{\mathrm{pvc}1}{(l-\mathrm{dl})\×{10}^6}\×|V_{\mathrm{SD}}|+\frac{\mathrm{pvc}2}{{(l-\mathrm{dl})}^2\×{10}^{12}}\×{V_{\mathrm{SD}}}^2)\\ \×(1+(\mathrm{pvl}+\frac{\mathrm{pv}1l}{l-\mathrm{dl}}+\frac{\mathrm{pv}1w}{w-\mathrm{dw}})\×|V_{\mathrm{SS}}\left|\right)\end{array}$ (formula 3)

In formula: reff represents the DC resistance of three end resistance, unit is ohm; Rsh represents square resistance, and unit is ohms/square; L represents the length of three end resistance, and w represents the width of three end resistance, and dl represents the length offset amount of three end resistance, and dw represents the width offset of three end resistance, and the unit of l, w, dl and dw is micron; L-dl represents the effective length of three end resistance, and w-dw represents the effective width of three end resistance; V _sDrepresent voltage between three end resistance sources and drain terminal, unit is volt; Pvc1 represents V _sDfor the first order influence coefficient of three end resistances, and pvc1 is relevant to the effective length of three end resistance; Pvc2 represents V _sDfor the second order influence coefficient of three end resistances, and pvc2 is relevant to the effective length of three end resistance; V _sSrepresent the voltage between three end resistance sources and substrate terminal, unit is volt; Pv1l represents V _sSfor the first order influence coefficient of three end resistances, and pv1l is relevant to the effective length of three end resistance; Pv1w represents V _sSfor the first order influence coefficient of three end resistances, and pv1w is relevant to the effective width of three end resistance; Pv1 represents V _sSto the first order influence coefficient of three end resistances; Wherein, rsh, dl, dw, pvc1, pvc2, pv1l, pv1w and pv1 are the model parameters in the model of three end resistance DC characteristic;

The model parameter of the model of the DC characteristic to three end resistance is given initial value, then be based upon the model file of putting forward the three end resistance DC characteristic that mould software can identify, this model file comprises initial value, formula (1), formula (2) and the formula (3) of model parameter;

Step 502) create and carry the three end Resistance model for prediction files that mould software can be identified, this model file comprises step 10) macro model and the step 501 set up) in the model of three end resistance DC characteristic in model parameter, and by this model file and step 501) model file of the three end resistance DC characteristic that obtain is placed under same file folder;

Step 503) carrying in mould software, first preference pattern type is three end Resistance model for prediction, be written into step 30) the three end resistance test data files that obtain, carry mould software and form three end resistance test curves according to three end resistance test data files, then be written into step 502) the middle three end Resistance model for prediction files that create, carry mould software and automatically generate three end resistance simulation curves according to this model file, then set-up procedure 502) in model parameter in the three end Resistance model for prediction files that create, change three end resistance simulation curves, until the root-mean-square error that three end resistance simulation curves and three end resistance test curves obtain is less than three end Resistance Fitting setting values, finally preserve step 502) in three end Resistance model for prediction files in model parameter, this model parameter is the model of integrated circuit by the DC characteristic of three end resistance,

Step 60) set up three end Resistance model for prediction files: by step 40) model and the step 50 of the integrated circuit AC characteristic of three end resistance that obtains) the integrated circuit model of the DC characteristic of three end resistance that obtains integrates, make the three end Resistance model for prediction files that comprise three end resistance DC and AC characteristics for integrated circuit simulating software, in this three ends Resistance model for prediction file, the girth of two parasitic diodes is w+l, and area is w × l/2;

Step 70) measure the resistance characteristic of three end resistance in integrated circuit: in integrated circuit simulating software, for three end resistance in integrated circuit simulating, invocation step 60) the three end Resistance model for prediction files that obtain, the resistance characteristic of three end resistance under alternating current-direct current signal function in integrated circuit measured.

2. the measuring method with three end resistances according to integrated circuit claimed in claim 1, it is characterized in that: described step 501) in, the model parameter of the model of the DC characteristic of three end resistance is given initial value and is: rsh=1, dl=0, dw=0, tc1=0, tc2=0, pvc1=0, pvc2=0, pv1=0, pv1l=0, pv1w=0.