Downloaded from https://www.everyspec.com MIL-HDBK-816 NOTICE 1 possibility that the displacement damage which high energy electrons produce may interfere with the measurements. Flash x-ray sources which, typically, deliver the dose at a very high dose rate, are not recommended because the results may be difficult to compare with Co-60. Low energy x-ray sources are not recommended for characterization measurements unless a valid correlation to Co-60 is established. 5.1.9.3 Bias. Bias must be applied on the device during the irradiations and the bias conditions must be in accordance with the test plan.
If measurements are to be made “in flux,” then the test plan should specify these conditions as well. Because time dependent effects can be very important for ionizing radiation dose, a complete time history of the irradiations and the post-radiation or in-flux measurements should be recorded. This time history should include the start and stop times of both the irradiations and the measurements as well as the rate at which the ionizing radiation dose was delivered. A pre- test evaluation of the time dependent effects may also be necessary to make sure the times used for the irradiations and the measurements will give meaningful results.
5.1.9.4 Transient ionization (dose rate) exposures and measurements. High energy electron Linacs are the facility of choice for dose rate upset measurements. If flash x-ray machines are used, extra care is required with the dosimetry to make results obtained at one facility comparable to those obtained at another. 5.1.9.5 Single event effects irradiations and measurements. SEE measurements are sufficiently complex so that they should be made in collaboration with one of the several groups in the U.S. that are making such measurements routinely.
A full characterization measurement for single event effects produced by heavy ions requires that the number of upsets or the latch-up in a particular device be measured as a function of the Linear Energy Transfer (LET) of the ion. Heavy ion irradiations are usually performed at cyclotrons or at Tandem Van de Graaff accelerators but may also be performed at other high energy heavy ion accelerators (see ASTM F1192, EIA/JEDEC-57, JESD-89).
For single event upsets induced by protons, a measurement of the number of upsets per unit fluence at a proton energy of 60 MeV or higher should be adequate as an input data point to a theoretical model that can then be used for estimating the total number of device upsets to be expected in some particular proton environment. (See Bendel and Petersen, 1983.) High energy protons for SEE tests can be obtained from a number of cyclotron accelerators. Laser and Californium-252 techniques are being studied but have not yet advanced to the point where they can be used routinely for SEE characterization measurements. 5.1.10 Post-irradiation measurements and data examination.
The parameters which will be measured after irradiation and the measurement conditions, and the equipment if possible, should be the same as those for the pre-irradiation measurements. Table III (see 5.1.5) and table V (see 5.1.7.5.1) list the information required. The test plan will contain the actual parameters to be measured and the corresponding operating conditions. However, the post-irradiation measurement procedure will have to take into account the possibility of time dependent effects. In the case of ionizing radiation dose, time dependent effects can be especially serious and must be evaluated for the device being tested.
MIL-STD-883 or MIL-STD- 750 method 1019 can be used as a guide until an improved method for taking time dependent effects into account is developed. Bias must be applied during the irradiation. Bias may not be necessary at all times after the irradiation but the bias conditions and other measurement conditions and time intervals should be as consistent as possible from one run to another. Bias or shorting conditions on the device between the end of a radiation exposure and the start of electrical measurements should be recorded. Again, the measurement results should be examined for bad devices or outliers and, appropriate action taken if necessary.
This examination is important in preventing bad data from being included in the characterization report and, possibly later, from entering a database. SUPERSEDES PAGE 23 OF MIL-HDBK-816 23