Heavy-ion Testing
Heavy ion test facilities:
​
There are currently only a handful of operational heavy ion test facilities in the US: Texas A&M University (TAMU), Lawrence Berkeley National Laboratory (LBNL), and Brookhaven National Laboratory (BNL), which has the Tandem Van de Graaff and the NASA Space Radiation Laboratory (NSRL) facility [1]-[4].
​
Testing logistics:
​
Heavy ion testing is typically conducted in-situ, which requires measurement capabilities of the device operation during irradiation. This often means that the test equipment have to be positioned in close proximity to the beam source. Heavy ion test facilities have a separate beam cave and control room. The test samples can be placed in vacuum or in air. Testing in air degrades the beam uniformity, and one needs to correct for the energy degradation of the ion passing through air. Operating in vacuum also introduces another layer of complexity to the test setup. Cable feed-throughs and longer cable lengths can introduce noise. Also, it may not be feasible to test high power devices in vacuum due to heat dissipation issues. Another important consideration is whether the range of ions sufficiently penetrates the device sensitive volume.
These considerations will impact the selection of the beam energy and test facility. For example, the NSRL facility at Brookhaven is a relatively high energy source facility. The energy is sufficiently high that the test samples do not need decapsulation in most cases. NSRL would be ideal for difficult to prepare package types (i.e. flip-chip packages), and/or for part types that require long ion range. The drawback is that the hourly rate at NSRL is also factors higher than other test facilities.
​
Heavy ion Test Methods and Analysis:
​
JEDEC JESD57A describes test methods and guidelines for heavy ion SEE testing [6]. European organizations typically use the ESA standard ESCC No. 25100 [7].
Heavy ion testing is often carried out for a range of LETs, with the goal of determining the upset threshold LET, saturating cross section, and a well defined shape for the cross section.
​
-
The experimenter should know within reasonable accuracy the LET through the device sensitive volume. The test facility typically reports the initial LET and surface LET as the ion exits the source. However, the experimenter should take care to understand beam degradation through air and other mediums before the sensitive volume. Overburden layers can be significant in some high-density modern ICs. Also, some device types have deep structures that require a long ion range to penetrate the sensitive volume, in order to trigger some destructive effects. So, it is always beneficial to have information on the device dimensions, or be conservative in the beam energy and ion range.
-
The desired particle fluence and the number of events or SEUs collected should be determined to reflect statistical representation of the sensitive volume(s). Test standards and guidelines typically recommend a fluence of 1E07 cm-2 for SEL testing, although one should always evaluate for the actual technology and architecture of the test article.
-
Other beam parameters of interest include the particle flux and angle. In some devices, a relatively high flux can result in successive ion strikes that are too close together (spatially or temporally), leading to bus contention effects. Such effects should be identified and taken out from the SEU rate analysis, since they are the artifact of cyclotron testing, and not likely to occur in a low flux space environment.
-
One should always evaluate the device response at multiple angles, since ion interaction in space is anisotropic. Irradiation at sharp incident angles can potentially produce significantly different cross sections than normal incident irradiation, also revealing the true dimensions of the sensitive volume(s).
​
Finally, the cross section data represents a cumulative probability distribution for the failure rate. Tools such as CREME96 can be used to calculate upset rates in a relevant radiation environment [5]. Our SEE analysis tool assists the user to plot cross section data, and find appropriate Weibull fits and quickly calculate figure-of-merit event rates.
​
Reference:
-
JEDEC, "Test procedure for the management of single-event effects in semiconductor devices from heavy ion irradiation," JESD57A, Nov 2017
-
ESA "Single Event Effects Test Method and Guidelines," ESCC No. 25100, Oct 2014.
​
Go back to the topics front page or read more about proton testing or laser testing.
​