MIL- STD-883F 2004 TEST METHOD STANDARD MICROCIRCUITS - 第169页

MIL-STD-883F METHOD 1032.1 29 May 1987 1 METHOD 1032.1 PACKAGE INDUCED SOFT ERROR TEST PROCEDURE ( DUE TO ALPHA PARTICLES) 1. PURPOSE . This test method defi nes the pr ocedure f or tes ting i ntegrated c irc uits under …

MIL-STD-883F

METHOD 1031

25 August 1983

This page intentionally left blank

MIL-STD-883F

METHOD 1032.1

29 May 1987

METHOD 1032.1

PACKAGE INDUCED SOFT ERROR TEST PROCEDURE

(DUE TO ALPHA PARTICLES)

1. PURPOSE

. This test method defines the procedure for testing integrated circuits under known test conditions for

susceptibility to alpha induced errors. This test was specifically designed to measure the device's ability to withstand alpha

particle impact. In addition, the procedure will determine the effectiveness of a "die-coating" shield. The test objective is to

determine the rate that failures are induced due to alpha radiation sourced from the device package, die and die-coat

material.

1.1 Definitions

. The following definitions were created to be specific and relevant within the confines of this method.

1.1.1 DUT

. Device under test.

1.1.2 Soft error

. Any error induced by alpha particle impact resulting in either a transient error or an error in data storage

witnessed at the DUT's output.

1.1.3 Source

. A foil of Thorium-232. (Note: This foil generates particles which have an alpha energy spectrum of 0

through 10 MeV).

1.1.4 Soft error rate (SER)

. Failures per unit time under normal conditions of package environment.

1.1.5 Accelerated soft error rate (ASER)

. Failures per unit time induced by exposure to a known alpha particle source.

1.1.6 Failure in time (FIT)

. 1 FIT = 1 failure in 10

device-hours.

1.1.7 Package flux

. The total number of alpha particles impinging on the die surface per unit of time and area, due to

package material impurities (i.e., lid, die material, sealants, and optional alpha barrier material). Normal units of

measurement: alpha/cm

-hr.

1.1.8 Modified package flux

. The total number of alpha particles impinging on the die surface per unit of time and area,

when a die coat is in place. Normal units of measurement: alpha/cm

-hr.

1.1.9 Source flux

. The total number of alpha particles impinging on the die surface per unit of time and area, due to the

calibrated source. Normal units of measurement: alpha/cm

-s.

2. APPARATUS

. The apparatus will consist of electrical test instrumentation, test circuit board(s), cabling, interconnect

boards, or switching systems and a Thorium-232 foil (optional). Precautions will be observed to obtain an electrical

measurement system with adequate shielding, low electrical noise induction, and proper grounding.

2.1 Radiation source

. The radiation source used in this test shall be a Thorium-232 foil with dimensions large enough to

cover the entire exposed die cavity. The plated source shall be within the range of 0.01 - 5.0 µCi and shall produce the

same energy spectrum as the package impurities. Radiation sources must be controlled according to state and federal

regulations. The sources shall be certified periodically and decay rates used to determine the actual flux values at the time

of use. This source must be processed at least one year before being used. Caution: These sources should not be

exposed to heat.

2.2 Electrical test instruments

. Electrical test instruments will be standard test instruments normally used for testing the

DUT. They must be capable of establishing the required test conditions and measuring the required electrical parameters.

All instruments shall be periodically calibrated in accordance with the general requirements of this test method standard.

MIL-STD-883F

METHOD 1032.1

29 May 1987

2.3 Test circuits

. The test circuit shall contain the DUT, wiring, and auxiliary components as required. Connection will

allow for the application of the specified test conditions to obtain the specified outputs. Provision will be made for monitoring

and recording the specified outputs. Any loading of the output(s), such as resistors or capacitors, shall be specified. The

test circuit must not exhibit permanent changes in electrical characteristics as a result of exposure to the radioactive source.

Shielding will be incorporated to prevent such effects from occurring if necessary.

2.4 Cabling

. Cabling, if required, shall be provided to connect the test circuit board containing the DUT to the test

instrumentation. All cables will be as short as possible. Care will be exercised to reduce electrical noise induced by the

cable by using shielded cable, triax, zipper tubing, or other shielding methods.

3. PROCEDURES

. Two methods of testing are allowed by this procedure. The first is a long term test (sometimes

referred to as a system test) which does not incorporate a source but which accumulates a statistically valid amount of test

time to determine the SER directly. This method is self explanatory and must be accomplished using the same parameters

outlined in 3.1 (test plan). To determine the SER from this method, the following formula should be used and the result

converted to FIT's.

SER = Total number of errors/Total test time

The second method incorporates the use of the source outlined in 2.1 (radiation source). The procedure for testing with an

accelerated flux provided by the source is given below. These steps will be followed for each test outlined in 3.1.

a. The flux that the surface of the die would receive without a die coat will be determined. This is designated as the

package flux.

b. If the device has a die-coat it should be left in place for the next portion of the test. The DUT will be delidded and

the source placed directly over the die cavity at the same distance as the package lid was from the die.

NOTE: The distance between the foil and the die must be less than 50 mils and the foil must cover the entire die-

cavity opening in order to assure all angles of incidence will be maintained.

NOTE: If the DUT has an inverted die configuration (e.g., flip-chip) a test jig must be implemented which will

expose the active surface of the die to the irradiating source.

c. The testing outlined in 3.1 will be performed at this time with the configuration in b. above, in order to determine

the SER for each test performed.

d. Recorded for each test performed will be the following:

(1) Total number of errors recorded during each test.

(2) Time to accumulate the errors.

(3) SER

, calculated from the following formulas:

ASER

= Total number of errors/test time

SER

= ASER

x (Package flux/source flux)

e. If no die-coating has been applied, the SER

will be reported as the measured rate of failure. However, if a die

coat exists, steps 3.f through 3.j will also be performed.