scope:

The primary objective of this document is to provide guidelines and easy to follow procedures for the preparation of detailed device specifications for the procurement of microcircuits and semiconductor devices where radiation Hardness Assurance (RHA) is required. The guidelines are applicable to MIL-M-38510, MIL-S-19500 microcircuit and semiconductor device detailed specifications as well as to other specifications such as Source Control Drawings (SCD), Selected Item Drawings (SID), Specification Control Drawings (SCD), and Standardized Military Drawings (SMD). Recommended procedures are provided for characterizing the radiation response of a part and for obtaining post-irradiation end-point limits for qualification and lot Acceptance Tests (LAT).

These guidelines address radiation response measurement and hardness assurance questions at the piecepart level. In keeping with the present scheme for MIL-STD RHA device specifications, which only addresses ionizing radiation dose effects measured at 200, plus or minus 100 rads(Si)/second and displacement damage effects due to neutrons (see table 1), these guidelines emphasize these two radiation environments. However, because the addition of dose rate and Single Event Upset (SEU) specifications is now under consideration, brief mention of these two radiation environments is also included. The principles discussed are applicable to both bipolar and MOS silicon transistors and integrated circuits, and to devices made from gallium arsenide and other semiconductor materials. They have, furthermore, been presented in terms of the intrinsic performance characteristics of the part independent of any system in which it may be used. For MIL-STD specifications, the levels shown in table I are used as reference points. The general principles may be applied to system specific requirements.

The procedures for measuring radiation response characteristics and for calculating LAT end-point limits are discussed in terms of parts whose design and production processes are mature. That is, these guidelines do not attempt to discuss the case where the part is still under development and its characteristics are still undergoing change. It is recognized that this latter case is important and occurs frequently because system designers are interested in obtaining the most advanced parts for their systems. The process of obtaining LAT end-point limits for such parts, however, involves testing and iterative end-point adjustments by the part manufacturer and the system parts engineers and designers which would be difficult to formulate as a set of generalized steps which could be applied to a variety of system needs. Because MIL-STD procurements of RHA devices use attribute lot Tolerance Percent Defective (LTPD) LAT tests exclusively, the end-point limit discussions here emphasize LTPD tests.

If sample costs are affordable, the LAT methods discussed will be performed on samples of the devices themselves. In the case of Very Large Scale Integration (VLSI) devices for which only low yields can be achieved, one possible option might be to use devices for LAT which are acceptable from an electrical performance standpoint but do not meet all the normal visual acceptance criteria. Such devices are said to be selected according to "alternate" visual criteria. Another possible option might be to use test structures for LAT which have been processed on the same wafers as the lot under consideration. By test structures are meant simpler and less expensive microcircuits which have been designed specifically to correlate with the radiation response characteristics and failure levels of the subject VLSI circuit; radiation tests on the test structure can then be used to estimate the performance characteristics of the VLSI circuit. The use of test structures for LAT cannot be recommended until data and experience show definitively that test structure responses correlate reliably with the actual devices.

A fundamentally different approach to quality assurance, from that based mostly on testing the end product, is now being implemented. An overly simplified description of this approach is to say that it is based on tests which will continuously measure the quality of the starting materials and of the production process itself and allow the material purity and the process to be controlled so that overall quality is not only maintained but is improved with time. This approach, which uses a Qualified Manufacturers' List (QML), is not yet in place and the way in which hardness assurance will be achieved under it has not yet been determined. With respect to the military standard procurement or "JAN" system this approach is also termed "Generic Qualification" (GQ) because it will allow a vendor to qualify a particular production line and then to ship a variety of part types from that line without having to qualify each new part type separately (as is required under the present qualified parts list or QPL scheme). GQ will be important because it is expected to improve radiation response uniformity over extended periods of time. Figure 1 shows some of the features of the proposed generic qualification system. A new draft controlling document has recently been issued by Rome Air Development Center (RADC) (see MIL-I-38535). This document includes RHA requirements in all the appropriate sections but does so only in general terms. A "strawman" plan entitled: "Methodology for Including Radiation Hardness Assurance in the Generic qualification Program" is presently undergoing review.