MIL-PRF-55365 Capacitor, Fixed, Electrolytic \(Tantalum\), Chip, Established Reliability, Nonestablished Reliability, And High Reliability, General Specification ForThis specification covers the general requirements for non-established reliability \(non-ER\), established reliability \(ER\), and high reliability tantalum dielectric, fixed chip capacitors, primarily intended for use in thick and thin film hybrid circuits or surface mount applications for filter, bypass, coupling, and other applications where the alternating current \(ac\) component is small compared to the direct current \(dc\) rated voltage and where supplemental moisture protection is available \(see 6.1\). The established reliability capacitors have reliability ratings established on the basis of life tests performed at specified voltage at +85??C for failure rate levels \(FRL\) ranging from: a. 1.0 percent per 1,000 hours to 0.001 percent per 1,000 hours in accordance with MIL-STD-690. These FRL?╞s are established at a 60-percent confidence level and are maintained at a 10-percent producer?╞s risk \(exponential distribution\). b. 0.1 percent per 1,000 hours to 0.001 percent per 1,000 hours or 1 FIT \(FIT = failure unit = one failure per 109 device hours\) at 90-percent confidence level \(Weibull distribution\).

Pages 1 - 2 - 3 - 4 - 5 - 6 - 7 - 8 - 9 - 10 - 11

MIL-PRF-55365H

4.7.20 Weibull FRL grading (see 3.23). Capacitors shall be tested in accordance with

method 108 of MIL-STD-202. The following details and exceptions shall apply:

a. Distance of the heating source from specimens, in inches: Not applicable.

b. Method of mounting: Capacitors shall be mounted by their terminations.

c. Test temperature and tolerance: +85�C +4�C, -0�C.

d. Operating conditions: Accelerated dc voltage, +4 percent, -0 percent, as applicable (see table XII), shall be

applied gradually (not to exceed 5 minutes by a slow buildup of the voltage). Maximum nominal acceleration

factor (see table XII) shall be 20,000:1. Only the capacitors used in 4.7.20e must be fused. Voltage shall be

applied continuously, except for failure count periods. The impedance of the voltage source, as seen from

the terminals of each capacitor, shall not exceed 1 ohm. An electronic power supply capable of supplying at

least 5 amperes when a capacitor is shorted shall be used. A 1 ampere to 2 ampere fuse shall be

connected in series with each capacitor. Slow-blow fuses shall not be used. The manufacturer shall be able

to demonstrate that all monitored and unmonitored parts receive correct voltage. When separate equipment

is used for testing the monitored sample and the unmonitored sample (rest of the lot), the following is

required:

(1) Equipment shall be cross calibrated for temperature, voltage, and test duration to ensure equivalent

test conditions.

(2) The ramp rate of voltage application at the start of test shall be identical for monitored and

unmonitored test groups.

e. Minimum sample size for monitoring at the beginning of test prior to infant mortality period: 300 pieces, or

100 percent, whichever is less. If the sample size is less than 100 percent, the remainder shall be subject to

the same accelerated dc voltage for the same amount of time.

f. Duration of test: 40 hours minimum. The manufacturer shall record the test start and stop times for each lot

tested.

Timing:

Infant mortality period

0 .................................0.......................... 0 ..........................0 ......................... 0

5-15

See

40 hours

See

4.7.20 i

minimum

4.7.20 i

minutes

g. Failure definition: A failure is defined as a blown fuse or equivalent.

h. Capacitors from positions with blown fuses: Capacitors from positions with blown fuses shall be removed

and shall not be delivered on the contract. These capacitors may be used for solderability testing or other

destructive testing.

Failure count during test: The lot size (see 4.6.1.1.2) to be graded is established after removal of gross

defectives (infant mortality) (5 minutes - 15 minutes). The first failure count shall be performed at 2 hours

�0.3 hour after the test was started. If there are no failures at time x1, the manufacturer may use one of the

following options:

(1) Complete a minimum of 40 hours and compute the failure from MIL-STD-690, table II FRSP-90, based

on the number of failures at time x2, or;

(2) Extend time x1 from 2 hours to a maximum of 10 hours. A failure cannot be assumed. If there are still

no failures, option 4.7.20i(1) shall be used, or;