Page images
PDF
EPUB

APPENDIX D

Schafft, H. A., Failure Analysis of Wire Bonds, presented as part of a Workshop on Failure Analysis of Semiconductor Devices and Packages, 1973 Reliability Physics Symposium, Las Vegas, Nevada, April 4, 1973.

Sher, A. H., and Kessler, H. K., Microelectronic Interconnection Bonding with Ribbon Wire, presented at Third Annual Symposium on Hybrid Microelectronics, Baltimore, May 3, 1973.

Buehler, M. G., Thermally Stimulated Measurements: The Characterization of Defects in Silicon p-n Junctions, Semiconductor Silicon/1973, H. R. Huff and R. R. Burgess, eds., pp. 549-560 (Electrochemical Society, Princeton, New Jersey, 1973).

Leedy, K. O., Scanning Electron Microscope Examination of Wire Bonds from High-
Reliability Devices, NBS Tech. Note 785, (July, 1973).

Schafft, H. A., Methods for Testing Wire-Bond Electrical Connections, to be presented at the Third Symposium on Reliability in Electronics, Budapest, November 13-16, 1973.

Use of Capacitor Microphones to Study Vibrations in Microelectronics Ultrasonic Bonding Equipment, Ultrasonics 11, No. 1, 5-6 (January, 1973). (Based on NBS Technical Note 573).

[merged small][merged small][merged small][merged small][merged small][merged small][ocr errors]

IN THE PAST SEVERAL YEARS ELECTRICAL MEASUREMENTS BY SPREADING RESISTANCE PROBES HAVE BEEN INCREASINGLY USED TO CHARACTERIZE SEMICONDUCTOR MATERIALS, DEVICES, AND PROCESSING. THE PURPOSE OF THIS SYMPOSIUM IS TO PROVIDE AN EXCHANGE OF INFORMATION AND TO REVIEW THE STATE OF THE ART OF MEASUREMENT BY SPREADING RESISTANCE PROBES.

Papers are solicited on the following subjects:

1. Theory

1.1 Models of contacts and current flow

1.2 Correction factors and their computation for thin and multilayer structures

[blocks in formation]

Persons intending to contribute papers should submit a descriptive title by
September 1, 1973, followed by a 500 to 1000 word summary before December
15, 1973 to enable the reviewers to determine the suitability of the paper
for the symposium.

Send to: Dr. James Ehrstein

National Bureau of Standards

Bldg. 225, Room B-346

Washington, D. C. 20234

Phone: 301-921-3625

Camera ready copies of the paper must be submitted to Dr. Ehrstein prior to
oral delivery of the paper.

The proceedings of the symposium will be published shortly after the meeting.
Only those papers presented at the symposium will be published.

JOINTLY SPONSORED BY ASTM COMMITTEE F-1 ON ELECTRONICS
AND THE NATIONAL BUREAU OF STANDARDS.

APPENDIX F

SWITCHING TRANSIENTS IN TRANSISTORS

The turn-off transient which occurs when the emitter of a transistor is opened rapidly while the base and collector remain connected can result in emitter-base junction breakdown (NBS Tech. Note 773, pp. 30-31). This breakdown may cause deterioration [1] or catastrophic failure of the transistors. The breakdown occurs even though the emitter current is in the direction appropriate to a forward biased emitter-base junction. It arises because of a potential drop along the emitter-base junction induced by the flow of majority carriers out of the base as required to bring the base region back to equilibrium following reduction of electron injection from the emitter. Breakdown can be avoided if the emitter is opened slowly enough that the base can remain in a near neutral condition without rapid flow of charge and the resulting potential build up. Alternatively, the emitter-base voltage can be limited to a value below the breakdown voltage with a protective circuit based on the use of a zener diode. The charge flow in the base during the turn-on and turn-off transients and of the operation of the protective circuit are described in detail below.

To elucidate the details of the charge flow during switching an extensive series of experiments was conducted. In these experiments, the current in various branches of the circuit was measured with a current probe and the voltage across the appropriate transistor terminals was measured with an oscilloscope. The circuit used for these measurements is shown in figure 23. This circuit also contains a protective zener diode circuit (NBS Tech. Note 754, p. 27) which can be incorporated by the switch S1.

[merged small][ocr errors][ocr errors][merged small][merged small][ocr errors][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small]

APPENDIX F

The low-level emitter current supply, usually included when measuring thermal resistance, was omitted during this series of experiments; no significant differences, except one which will be noted subsequently, were observed when the low-level emitter current was included.

Turn-on Transient

[ocr errors]

Consider first the conditions in the on state which are set up during turn on. For explicitness the argument is developed for an n-p-n transistor, but it can easily be extended to a p-n-p transistor. In the quiescent state the base is grounded and the collector-base junction is reverse biased (collector positive). At a time t = 0, the emitter supply is connected to the emitter contact to foward bias the emitter-base junction (emitter negative) which results in electrons being injected into the base. The injected electrons are minority carriers in the base.

s.

They diffuse toward the collector-base junction with a time constant that can be the order of microseconds, but as soon as they enter the base they upset the charge neutrality in the base. To maintain charge neutrality, majority holes flow rapidly into the base from the base contact. The return to neutrality in the base occurs with a time constant determined by the dielectric relaxation time in the base. This time constant is the order of 10-12 to 10-13 Thus, as soon as an electron is injected into the base a hole is created to neutralize it. This process results in a build up of the electron-hole pair density in the base and continues until equilibrium is reached. The turn-on current transients for a 35-W, single diffused transistor with a 1.5 mil (60 μm) wide base are shown in figure 24. During the period of charge build up, the electron current into the base from the emitter is much larger than the electron current leaving the base for the collector, and holes flow into the active region of the base, under the emitter-base junction. Outside the active region the base is thin and long relative to its width. The charge flow across the extrinsic base

[graphic][merged small][ocr errors][merged small][merged small][merged small]

APPENDIX F

resistance associated with this region results in a potential drop such that the edge of the active region of the transistor is negative with respect to the base contact. Lateral hole flow across the intrinsic base resistance within the active region, causes an additional potential drop across the emitter-base junction; this potential drop is, of course, in the same direction. It acts to debias the center of the junction and causes the familiar current crowding phenomenon which forces the emitter current to the outer periphery of the junction.

Equilibrium is established when the excess hole-electron pair density in the base region reaches the value appropriate to the injected electron current from the emitter. In equilibrium, holes still flow from the base contact to the active region in order to supply holes for recombination with electrons in the base (base transport losses) and for injection into the emitter (injection efficiency losses). The current associated with these losses is much smaller than either the peak base current during turn-on or the steady-state emitter current. In the example shown in figure 24, the equilibrium base current is 15 mA while the peak base current is approximately 450 mA and the steady-state emitter and collector currents are about 1 A.

Turn-off Transient When the transistor emitter circuit is opened, electron injection into the base from the emitter decreases. However, the excess carrier density which causes the concentration gradient associated with the diffusion of electrons toward the collector does not change instantaneously; hence, electrons continue diffusing toward the collector-base junction where they are swept out of the base region. This results in a reduction of the electron density in the base which must be accompanied by a reduction of the hole density in order to maintain charge neutrality. Consequently, holes must flow out from the active region of the transistor to the base contact. This flow is in the opposite direction to the flow during the turn-on transient and sets up a potential gradient across the extrinsic base resistance such that the active region of the transistor is positive with respect to the contact. The most positive portion of the base can approach the potential of the collector region.

EB

The simplest turn-off transient occurs for collector-base voltage well below the breakdown voltage of the emitter base junction. To illustrate this case, the current and base emitter voltage turn-off transients are shown for a collector voltage of about 10 V in figure 25. The switch S1 in the test circuit (figure 23) was set to position 1 so that the protective circuit is not connected. When turn-off begins, as evidenced by a decrease in begins to rise toward zero from a small negative potential. At the same time, ig decreases from a small positive value to zero and then increases to a negative value of approximately 165 mA. During this same period ig begins to decrease rather rapidly, but the collector current does not begin to decrease rapidly until about 0.2 us after ip starts to decrease. During this 0.2-μs period, i increases in the reverse direction in such a way as to be always equal to

E

« PreviousContinue »