MIL- STD-883F 2004 TEST METHOD STANDARD MICROCIRCUITS.pdf - 第530页

MIL-STD-883F METHOD 3018 29 May 1987 2 2. APPARATUS . The apparatus us ed for c ros stal k measur ements s hall i nclude a s uitabl e sourc e generator (s ee 2.1), wideband osc illosc ope (see 2.2) , low c apacit ance pr…

MIL-STD-883F

METHOD 3018

29 May 1987

METHOD 3018

CROSSTALK MEASUREMENTS FOR DIGITAL MICROELECTRONICS DEVICE PACKAGE

1. PURPOSE

. This method establishes the means of measuring the level of cross-coupling of wideband digital signals

and noise between pins in a digital microcircuit package. The method may be used to gather data that are useful in the

prediction of the package's contribution to the noise margin of a digital device. The technique is compatible with multiple

logic families provided that the drive and load impedance are known.

1.1 Definitions

1.1.1 Crosstalk

. Signal and noise waveforms coupled between isolated transmission lines, in this case, package

conductors.

1.1.2 Coupling capacitance

. The effective capacitance coupling between a pair of conductors in a package as measured

by the time constant of the charge pulse applied on one line and measured on the other.

1.1.3 Noise pulse voltage

. The voltage of a crosstalk measured at the minimum noise pulse width as measured on a

receiver input line.

1.1.4 Peak noise voltage

. The peak value of the noise pulse measured on a receiver input line.

1.2 Symbols

. The following symbols shall apply for the purpose of this test method and shall be used in accordance with

the definitions provided (see 1.2.1 and 1.2.2).

1.2.1 Logic levels

(max): The maximum output low level specified in a logic system.

(min): The minimum output high level specified in a logic system.

(max): The maximum allowed input low voltage level in a logic system.

(min): The minimum allowed input high level in a logic system.

1.2.2 Noise pulse width

: The low level noise pulse width, measured at the V

(max) level (see method 3013).

: The high level noise pulse width, measured at the V

(min) level (see method 3012).

1.2.3 Transition times (see method 3004)

tLH

: Rise time. The transition time of the output from the 10 percent to the 90 percent of the high voltage levels

with the output changing from low to high.

tHL

: Fall time. The transition times from the 90 percent to the 10 percent of the high voltage level with the output

changing from high to low.

1.2.4 Crosstalk parameters

: Coupling capacitance (see 1.1.2).

: Noise pulse voltage (see 1.1.3).

NPK

: Peak noise voltage (see 1.1.4).

MIL-STD-883F

METHOD 3018

29 May 1987

2. APPARATUS. The apparatus used for crosstalk measurements shall include a suitable source generator (see 2.1),

wideband oscilloscope (see 2.2), low capacitance probe (see 2.3) and load resistors (see 2.4).

2.1 Source generator

. The source generator for this test shall be capable of duplicating (within 5 percent) the transition

times, V

and V

levels of the logic system(s) being considered for application using the package style under evaluation.

The source generator shall have a nominal characteristic source impedance of 50 .

2.2 Wideband oscilloscope

. The oscilloscope used to measure the crosstalk pulse shall have a display risetime that is

less than 20 percent of the risetime of the logic systems being considered for application in the package style under

evaluation. A sampling-type oscilloscope is recommended.

2.3 Low capacitance probe

. The interface between the oscilloscope and the unit under test shall be a high impedance low

capacitance probe. The probe impedance shall be 10 k, minimum and the capacitance shall be 5 pF, maximum, unless

otherwise specified in the acquisition document.

2.4 Load resistor

. The load resistors specified for this test shall be low inductance, low capacitance, chip style resistors

with a tolerance of ±5 percent. Load resistor values(s) shall be specified by the acquisition document to match the load

impedance levels of the application logic family for a single receiver load.

3. PROCEDURE

. The test equipment configuration shall be as shown on figure 3018-1 using a source generator,

oscilloscope, probe and loads as specified (see 2). Measurements shall be made of coupling capacitance, (see 3.2) and if

required by the acquisition document, of noise pulse voltage, peak noise pulse voltage, and noise pulse width (see 3.3).

3.1 General considerations

3.1.1 Package test configuration

. It is important to ground the package using the same pins as would be used in the

microcircuit application. If the package has an internal ground plane or ground section, this should be connected via

package pin(s) to the exterior test set-up ground plane. The package should be connected to the test set-up with coaxial

cable or stripline. Unshielded conductor medium should not be used between the signal source and package. Coaxial

shields must be grounded at both ends of the cable. Package sockets should not be used unless these are to be part of the

microcircuit application configuration. Package leads must be formed and trimmed as specified in the application.

Package-to-chip interconnecting media shall be installed in the package and used to connect to the load resistors.

3.1.2 Pin selection

. For simple packages with symmetrical, parallel pin conductors, only a sample of pin combinations

need be tested. Unless otherwise specified by the acquisition document, all combinations adjacent to the ground pin(s) and

combinations opposite the ground pin(s) shall be tested, as a minimum. Complex packages with nonparallel conductors or

multilayer wiring shall be tested for all adjacent-pair combinations, unless otherwise specified.

3.2 Coupling capacitance measurements

. Connect the test equipment as shown on figure 3018-1. Use a 50  chip

resistor load in the driven pin channel, unless otherwise specified. For the pick-up channel, use the load resistor value(s) as

specified by the acquisition document. (Load resistor values should be set such that the parallel combination of load

resistance and probe impedance matches as closely as practical the specified load impedance of a single receiver in the

logic system to be used in the microcircuit application.) Check the residual cross-coupling of the measuring set-up by

touching the probe to the pick-up channel load before the pick-up pin is connected to the resistor. Measure and record the

peak pulse voltage observed. This peak pulse reading must be less than 50 percent of the reading observed with the pin

connected to the resistor for a reading to be valid. Adjust the test set-up cable orientation and configuration to minimize this

residual cross-coupling.

MIL-STD-883F

METHOD 3018

29 May 1987

Connect the pick-up pins to the load resistor and adjust the pulse width so that the time required to charge the coupling

capacitance to 0 V can be observed. Measure the time at the 63 percent voltage point on the waveform (T) and calculate

coupling capacitance (C

) as follows:

Determine

Total

Probe Load

Total

C Total Probe

Values of C

can be used as a relative measure for comparison of potential crosstalk among several packages to a standard

package. The coupling capacitance (C

) can also be used to predict levels of crosstalk for various logic systems or circuit

configurations by performing a pulse response analysis using a circuit simulator.

3.3 Noise pulse measurements

. Using the same test setup as in 3.2, measure the crosstalk noise pulse voltage at the

minimum noise pulse width specified for the logic system or as specified by the acquiring agency.

Measure the peak noise voltage value of the coupled crosstalk.

4. SUMMARY

. The following details, when applicable, shall be specified in the acquisition document:

a. C

b. V

(max).

c. V

(min).

d. V

(max).

e. V

(min).

f. t

g. t

h. t

tLH

i. t

tHL

j. V

k. V

NPK