Particle Impact Noise Detection

Are you looking for more information regarding this and other subjects? Look no further than Semitracks' Online Training. Semitracks' Online Training contains short courses and other material useful for any Engineer trying to learn new subjects or brush up on old ones.

What is a PIND Test?

PIND stands for Particle Impact Noise Detection. The purpose of this test is to measure the amount of electrical noise produced by the device while under vibration. If loose particles are present in the cavity of the package, failures such as short circuits may occur to sabotage the reliability of the component.

Why Perform a PIND Test?

First of all, the test, governed by MIL-STD-883C Method 2020, is required to guarantee certain levels of component reliability for weapon systems and other military applications. The test may also be useful in intermittent failures such as loose bond wires or suspected particle induced high leakage currents or a short. A variable frequency PIND test is the most sensitive to detect the presence of conductive particles. The device can be made susceptible to potential shorting problems by performing the test using maximum bias voltage conditions to electrostatically attract the particles to exposed conductors.

How is a PIND Test Performed?

The necessary test equipment must be capable of providing a required variable frequency vibration, a calibrated high impedance voltmeter for noise measurement during test, and the necessary optical and electronic equipment for post-test measurements.

The device under test is attached to a transducer head with a grease-like substance. When the transducer detects a particle by the noise which is produced, a signal is produced proportional to the noise.

The device is rigidly fastened on a vibration platform, and the leads are adequately secured. The range of frequency is 20 to 2 kHz with an amplitude of 0.06 inch or a constant peak acceleration of 20 g minimum. The test does not last longer than 4 minutes. The orientation of vibration is performed along each axis (x, y, and z). Devices with an internal cavity containing parts or elements subject to possible movement or breakage during vibration should be further examined by delidding or opening them and performing an internal visual examination of 30X to reveal damage or dislocation. X-ray techniques could also be used to assess any internal damage if open the package may inhibit further analysis.

When is a PIND Test Performed?

PIND tests are only applicable to components with unfilled cavities. The PIND test is performed when the analyst suspects high leakage currents or a short. The test may also be useful in intermittent failures such as loose bond wires. It is important to note that the PIND test cannot differentiate between conductive and non-conductive particles; it can only detect loose particles. Internal electrostatic fields, often located where they can do the most damage to the die surface, attract particles of small mass.

MIL STD Procedures for PIND

MIL-STD 883C

METHOD 2020.3

PARTICLE IMPACT NOISE DETECTION TEST

  1. PURPOSE. The purpose of this test is to detect loose particles inside a device cavity. The test provides a nondestructive means of identifying those devices containing particles of sufficient mass that, upon impact with the case, excite the transducer. Because of the limited efficiency of this test method, it may be desirable to subject devices to several sequences of this test in order to achieve desired confidence.
  2. APPARATUS. The equipment required for the particle impact noise detection (PIND) test shall consist of the following (or equivalent):
    1. A dual beam oscilloscope capable of 500 kHz response minimum, and a sensitivity of 20 mV/cm for visual display of the particle noise and of the threshold detector. Alternatively, a single beam oscilloscope may be used in conjunction with a lamp indicator for the threshold detection circuit.
    2. A threshold detector to detect particle noise voltage exceeding a preset threshold of 5 +/- 1 millivolt peak above system peak noise. See figure 2020-3 for an acceptable circuit to perform the threshold detection function. The threshold detector may be disabled during application of co-test shock for a period of time not to exceed 100 ms from initiation of the shock pulse.
    3. An audio system with speaker to monitor the audio signal from the PIND electronics. If headphones are used, the system shall provide safeguards against loud noise bursts.
    4. A vibration shaker and driver assembly with a payload consisting of the DUT, (PIND) transducer, the transducer isolator, preamplifier (when included), co-test shock mechanism (when included), a portion of the transducer cable and its restraints, capable of providing essentially sinusoidal motion at:
      1. Condition A - 20g peak at 40 to 250 Hz.
      2. Condition B - 10g peak at 60 Hz.
    5. PIND transducer, calibrated to a peak sensitivity of -77.5 +/- 3 dB per one volt per microbar at a point within the frequency of 150 to 160 kHz.
    6. A sensitivity test unit (STU) (see figure 2020-2) for periodic assessment of the PIND system performance. The STU shall consist of a transducer with the same tolerances as the PIND transducer and a circuit to excite the transducer with a 250 microvolt ~20 percent pulse. The STU shall produce a pulse of about 20 mV peak on the oscilloscope when the transducer is coupled to the PIND transducer with attachment medium.
    7. PIND electronics, consisting of an amplifier with a gain of +60 +/- 2 dB centered at the frequency of peak sensitivity of the PIND transducer to amplify the transducer signal to a usable level for threshold detection, audio detection and oscilloscope display. The noise at the output of the amplifier shall not exceed 10 mV peak.
    8. Attachment medium. The attachment medium used to attach the DUT to the PIND transducer shall be either a viscous acoustic couplant such as Automation Industries No. 50A4084 (or equivalent) or double-faced tape such as Permacel P50 (or equivalent).
    9. Co-test shock mechanism or tool, consisting of the integral co-test shock mechanism of 2.d. above (when included), or a six-inch solid AWG No. 10 copper rod with rounded end, or other mechanism capable of imparting shock pulses between 200 and 1800g to the DUT. The duration of the main shock shall not exceed 100 microseconds. If an integral co-test shock system is used the shaker vibration may be interrupted or perturbed for period of time not to exceed 250 ms from initiation of the shock pulse.
    10. Special mounting adapters for devices which have irregular surfaces (see 3.3.2).
    11. Isolator material between the PIND transducer and the vibration shaker and driver when required to reduce background noise. The isolator shall have no resonance within the test frequency range.
    12. A pre-test shock fixture capable of imparting shock pulses between 200 and 1800g to the DUT. The duration of the main shock shall not exceed 100 microseconds. A co-test shock mechanism integral to the shaker and driver may be used for this purpose.
  3. PROCEDURES.

    3.1 Test equipment setup. The test equipment shall be connected as indicated in figure 2020-1 and set-up in a low background noise area. Critical settings to provide proper detection sensitivity, unless otherwise specified, are as follows:

    1. Audio output volume shall be adjusted to a comfortable noise level output.
    2. Shaker drive frequency shall be adjusted in accordance with table I for condition A, or at 60 Hz for condition B.
    3. Shaker drive amplitude shall be 20g (condition A) or 10g (condition B) with DUT and mounting adapter (if any) in place.
    4. Oscilloscope vertical deflection primary beam sensitivity (displaying PIND electronics output) shall be 20 millivolts/centimeter. Secondary beam sensitivity (if displaying threshold detector output) shall produce approximately a 2 centimeter deflection difference between the two states of the threshold detector. The secondary beam display (without horizontal deflection) shall be centered vertically and approximately 1 centimeter to the left or right of the primary beam display.
    5. Oscilloscope horizontal deflection shall be adjusted to 4 cm and shall obtain drive from the sine generator/amplifier, amplified accelerometer, or a time base (2 ms/cm) triggered from the accelerometer output.

    3.2 Test equipment checkout. The test equipment checkout shall be performed to assure proper system operation, when any of the following occurs:

    1. After a change of vibration frequency.
    2. System shut-down for any reason.
    3. Change of operators.
    4. Work shift change.
    5. Prior to and after testing group(s) of devices or every 4 hours during the test operating period, whichever comes first. System deficiencies shall be corrected prior to test. Failure of the system to meet checkout requirements shall require retest of all devices tested subsequent to the last successful system checkout.
    • 3.2.1 Shaker drive system checkout. The drive system shall achieve the shaker frequency specified in 3.1.b. and shaker amplitude specified In 3.1.c. The drive system shall be calibrated so that the frequency settings are within +/- 8 percent and the amplitude vibration settings are within +/-10 percent of the nominal values. If a visual displacement monitor is affixed to the transducer, it may be used for amplitudes between 0.04 and 0.12 Inch (1.02 and 3.05 mm). An accelerometer may be used over the entire range of amplitudes and shall be used below amplitudes of 0.040 inch (1.02 mm).
    • 3.2.2 Detection system checkout. With the shaker deenergized, the STU transducer shall be mounted face-to-face and coaxial with the PIND transducer using the recommended attachment medium. The STU shall be activated several times to verify low level signal pulse visual and threshold detection on the oscilloscope (approximately 20 millivolt peak or 10 millivolt peak above system noise).

      NOTE: Not every application of the STU will produce the required amplitude but the majority of applications will do so.

    • 3.2.3 System noise verification. For proper system operation, no extraneous noise can be permitted to exist in the system. During proper operation, the normal system noise, as observed on the oscilloscope, will appear as a fairly constant band and must not exceed 10 millivolts zero to peak. Extraneous noise is defined as noise in the system other than the permissible background noise that is present with no device on the transducer. Such noise can be due to a number of sources which must be eliminated or their effects guarded against, since those non-signal noise spikes can trigger the threshold detector and appear as signals on the other indicators. Common sources of electromagnetic Interference from such sources as fluorescent lighting, heater elements, soldering irons, and other switching transients, line transients and especially, less than optimum installation and support of the transducer cabling. The latter source normally may be eliminated by redressing the cable, tightening or cleaning the connector at the transducer, or even replacing the transducer or transducer cable. To verify that no extraneous noise exists in the system, observe the oscilloscope while turning on the shaker and Increasing the drive amplitude from zero to the desired acceleration level (see 3.1 c.) while applying the co-shock (see 3.3.4). This noise is usually present as pulses which remain in a fixed position on the oscilloscope trace. If extraneous noise is observed, correct the problem by shielding or other precautions, such as those suggested above and re-run the entire noise check.

    3.3 Test sequence. The following sequence of operations (a through i) constitute one test cycle or run.

    1. Pre-test shock.
    2. Vibration 3-5 seconds.
    3. Co-test shock.
    4. Vibration 3-5 seconds.
    5. Co-test shock.
    6. Vibration 3-5 seconds.
    7. Co-test shock.
    8. Vibration 3-5 seconds.
    9. Accept or reject.
    • 3.3.1 Pre-test shock. Prior to vibrating the device, it shall receive a pre- test shock (see 2.L).
    • 3.3.2 Mounting requirements. Special precautions (e.g., in mounting, grounding of DUT leads, or grounding of test operator) shall be taken as necessary to prevent electrostatic damage to the DUT. All devices shall be mounted in an inverted position without adapters except for the following:
      1. Stud-mounted devices shall be mounted In suitable adapters.
      2. Axial diodes shall be mounted without adapters and with the leads in a horizontal plane.
      3. Double-ended resistance welded packages (i.e., optical isolator) shall be mounted using a suitable adapter and with the leads horizontal.
      4. Most part types will mount directly to the transducer via the attachment medium. Parts shall be mounted with the largest flat surface against the transducer at the center or axis of the transducer for maximum sensitivity. When so mounted, the leads of the part will point up (e.g., TO-5) or horizontal (e.g., flat packs). Where more than one large surface exists, the one that is the thinnest in section or has the most uniform thickness shall be mounted toward the transducer, e.g., flat packs are mounted top down against the transducer. Small axial-lead, right circular cylindrical parts are mounted with their axis horizontal and the side of the cylinder against the transducer. Parts with unusual shapes may require special fixtures. Such fixtures shall have the following properties:
        1. Low mass.
        2. High acoustic transmission (aluminum alloy 7075 works well).
        3. Full transducer surface contact, especially at the center.
        4. Maximum practical surface contact with test part.
        5. No moving parts.
        6. Suitable for attachment medium mounting.

        Leads on the parts shall be dressed, as necessary, so they will not strike each other or the transducer during vibration. Long or thin section leads shall be observed for signs of resonance, indicated by motion exceeding 3 or 4 diameters. Such resonance may give extraneous noise during test even though the leads do not strike each other. In these cases, the leads may have to be shortened (if permitted by the application) or special fixturing or frequency changes may be required.

        NOTE: Some especially long-leaded TO-5 packages have been observed to be close to resonance at the test frequency.

    • 3.3.3 Test monitoring. To avoid false indications, the DUT shall be inspected for any attached foreign matter or leads which are touching each other. The DUT shall be mounted on the center of the transducer using attachment medium and if necessary, a mounting adapter. To provide maximum signal transmissibility with a viscous couplant, a sufficient amount of couplant shall be used and the DUT shall be firmly mounted so that any excess couplant can be squeezed out. When double-faced tape is used, it shall be changed at the start of a test group and after each 25 units or less thereafter. Devices shall be put on and removed from the attachment medium with a slight twisting motion. Device orientation for each package type shall be as specified in 3.3.2. The shaker input frequency shall be set in accordance with 3.1.b. and the shaker drive amplitude shall be increased to the level specified in 3.1.c. All detection systems shall be monitored for evidence of loose particles. Any device which gives a particle indication shall be considered a reject. Particle indications can occur in any one or combinations of the three detection systems as follows:
      1. Visual indication of high frequency spikes which exceed the normal constant background white noise level.
      2. Audio indication of clicks, pops, or rattling which is different from the constant background noise present with no DUT on the transducer.
      3. Threshold detection shall be indicated by the lighting of a lamp or by deflection of the secondary oscilloscope trace.
      4. If no particles are observed in 3 to 5 seconds, a co-test shock (see 3.3.4) shall be applied to the DUT while the shaker is operating. It is permissible to interrupt or perturb the vibration for a period not to exceed 250 milliseconds to provide for the application of an integral co-test shock. The audio, oscilloscope, and threshold detection systems are to be closely monitored during the time period immediately after each shock application as well as for an additional 3 - 5 seconds to detect particles which may lock up quickly. If no particles are detected with the first co-test shock application, the test shall be repeated two times. If there is no indication of particles within 3-5 seconds after the third co-test shock (see 3.3.4), the device is acceptable.
    • 3.3.4 Co-test shock application. When using the copper rod shock tool (see 2.1.), the shock shall be applied to the DUT by bringing at least 1/4 to 1/2 inch of the free end of the shock tool into momentary contact with the vibrating DUT. The tool shall be held lightly and freely between the thumb and forefinger opposite the free end. Striking or hammering motions shall not be used. The shock shall be only the result of the mass inertia of the freely supported shock tool being struck by the vibrating DUT. The tool shall be held approximately horizontal and shall contact the DUT on a portion of the upper surface of its case. The duration of this contact is on the order of one-half second and results in several impacts of random shock to the DUT. The tool shall not contact the leads, other than minor accidental brushing of the leads along and parallel to their axis and shall not contact any glass portion of the case, except for all glass envelope diodes. If any other co-test shock device is used, its mode of operation shall be in accordance with procedures supplied by the equipment manufacturer. In systems that disable the threshold detector during the co-test shock, the period of time from shock pulse to reinitiation of threshold detection shall not exceed 100 milliseconds.

    3.4 Failure criteria. Any noise bursts as detected by any of the three detection systems exclusive of background noise, except those caused by the shock blows, during the monitoring periods shall be cause for rejection of the device. Rejects shall not be retested (see 3.3.3) except for retest of all devices in the event of test system failure as provided in 3. If additional cycles of testing on a lot are specified, the entire test procedure (equipment set-up and checkout mounting, vibration, and co-shocking) shall be repeated for each retest cycle. Reject devices from each test cycle shall be removed from the lot and shall not be retested in subsequent lot testing.

    * 3.5 Lot acceptance. Unless otherwise specified, the inspection lot (or sublot) shall be submitted to 100 percent PIND testing a maximum of three times in accordance with condition A herein. PIND prescreening shall not be performed. The lot may be accepted on any of the three runs if the percentage of defective devices is less than 1 percent (or one device, whichever is greater). All defective devices shall be removed after each run. Lots which do no meet the 1 percent PDA on the third run, or exceed 25 percent defectives cumulative, shall be rejected and resubmission is not allowed.

    TABLE I. Package height vs test frequency for 20G acceleration (condition A).

    Average internal cavity height Frequency
    Mils mm Hz
    <40 <1.02 130
    40-50 1.02-1.27 120
    50-60 1.27-1.52 110
    60-70 1.52-1.78 100
    70-80 1.78-2.03 90
    80-90 2.03-2.29 80
    90-100 2.29-2.54 70
    >100 >2.54 60

    Note: The approximate average internal package height shall be measured from the floor of the package cavity or the top of the major substrate for hybrid or multichip assemblies and shall exclude the thickness of the die mounted inside the package.

  4. SUMMARY. The following details shall be specified in the applicable detail specification:
    1. Test condition letter A or B (see 2.d. and 3.1.c.).
    2. Lot acceptance/rejection criteria (if applicable).
    3. The number of test cycles, if other than one.
    4. Attachment medium, if other than that specified (see 2.h.).
    5. Pre-test shock level and co-test shock level, if other than specified in 2.l. and 2.i., respectively.

NOTICE 7

* 3.3.2 Test monitoring. Each test cycle (see 3.3) shall be continuously monitored except for the period during co-test shocks and 250 ms maximum after the shocks. Particle indications can occur in any one or combinations of the three detection systems as follows:

  1. Visual indication of high frequency spikes which exceed the normal constant background white noise level.
  2. Audio indication of clicks, pops, or rattling which is different from the constant background noise present with no DUT on the transducer.
  3. Threshold detection shall be indicated by the lighting of a lamp or by deflection of the secondary oscilloscope trace.

3.4 Failure criteria. Any noise bursts as detected by any of the three detection systems exclusive of background noise, except those caused by the shock blows, during the monitoring periods shall be cause for rejection of the device. Rejects shall not be retested except for retest of all devices in the event of test system failure. If additional cycles of testing on a lot are specified, the entire test procedure (equipment setup and checkout mounting, vibration, and co-shocking) shall be repeated for each retest cycle. Reject devices from each test cycle shall be removed from the lot and shall not be retested in subsequent lot testing.

3.5 Lot acceptance. Unless otherwise specified, the inspection lot (or sublot) shall be submitted to 100 percent PIND testing a maximum of five times in accordance with condition A herein. PIND prescreening shall not be performed. The lot may be accepted on any of the five runs if the percentage of defective devices is less than 1 percent for one device, whichever is greater). All defective devices shall be removed after each run. Lots which do not meet the 1 percent PDA on the fifth run, or exceed 25 percent defectives cumulative, shall be rejected and resubmission is not allowed.

TABLE I. Package height vs test frequency for 20G acceleration (condition A).

Average internal cavity height Frequency
Mils mm Hz
<40 <1.02 130
41-50 1.02-1.27 120
51-60 1.27-1.52 110
61-70 1.52-1.78 100
71-80 1.78-2.03 90
81-90 2.03-2.29 80
91-100 2.29-2.54 70
>100 >2.54 60

Note: The approximate average internal package height shall be measured from the floor of the package cavity or the top of the major substrate for hybrid or multichip assemblies and shall exclude the thickness of the die mounted inside the package.

4. SUMMARY. The following details shall be specified in the applicable detail specification:

  1. Test condition letter A, B, or C.
  2. Lot acceptance/rejection criteria (if applicable).
  3. The number of test cycles, if other than one.
  4. Pre-test shock level and co-test shock level, if other than specified.

MIL-STD-750C, method 2052.

References on PIND

R.E. McCullough, J.C. Burru, and W.R. Reynolds, "PIND's Role as a Failure Analysis Tool," Proceeding of AFTA, 1979, pp. 23-26.