space-rha.comShort courseChapter 4 · Mechanisms and technologies

Chapter 4 · Lesson 5 of 6 · about 14 minutes

Bipolar devices and ELDRS

Some bipolar and linear devices degrade more at a low dose rate than a high-rate test would suggest. ELDRS depends on the technology and on the test and use conditions; there is no single correction factor that covers it. The assurance case has to say why the exposure that was chosen represents the mission.

ELDRS laboratory

The space-rha.com ELDRS model: two identical bipolar oxide blocks at the same dose and two dose rates, with hydrogen and temperature controls.

An original space-rha.com laboratory, embedded unchanged with its own controls; its onset values, gains and depths are teaching assumptions, not limits for flight hardware.

Equal dose at a faster rate may not be conservative.

Dose-rate enhancement and ordinary annealing are different things.

Annealing between exposure and measurement can change a comparison between two tests without any ELDRS being involved. In a susceptible bipolar structure, the processes in the oxide and at the interface can increase recombination in the base and reduce the gain. Bias, temperature, hydrogen in the package and the process history can all affect the response.

The evidence should be low-dose-rate data that apply to the part, or an accepted acceleration method that has been demonstrated for it, with the conditions and limits of that method stated. A generic margin on a high-dose-rate result does not take the place of that evidence.

Evidence to retain. Keep the low-dose-rate data or the validated accelerated data, the dose-rate history, the bias, the package and process identity, the temperature, any anneal and the linear-circuit parameters the application depends on.

Mission context. An analog feedback path can lose gain or pick up an offset over a long mission even though it survived a short high-rate exposure in the laboratory.

Going deeper

Reaction and transport compete near the emitter–base surface.

Holes generated by the radiation and hydrogen-related species take part in reactions in the oxide and in the buildup of interface traps. The transport and the reaction kinetics depend on the field, the time and the precursors available. The added surface recombination raises the base current, which lowers β at a given collector current.

The one-population annealing law shown in the laboratory leaves out the hydrogen reaction network on purpose. It illustrates one trade in the time history. It does not describe the physics of ELDRS, and it is not an acceleration law for qualification.

Takeaway

The dose rate of a test is one of the conditions that decide whether its result applies, and the choice of dose rate needs a physical and an experimental justification.

Check yourself

A bipolar operational amplifier passed a high-dose-rate Co-60 test with a generous margin over the mission dose, and the project proposes to accept it on the strength of that margin. Why is that not the evidence the course asks for?

Some bipolar and linear devices degrade more at a low dose rate than a high-rate test suggests, and ELDRS depends on the technology and on the test and use conditions, so there is no single correction factor and a generic margin on a high-dose-rate result does not take the place of the evidence. What is needed is low-dose-rate data that apply to the part, or an accepted acceleration method that has been demonstrated for it, with the conditions and limits of that method stated. The teaching amplifier of lesson 8.3 shows the size the effect can have: its low-dose-rate leg failed at about 50 % of the high-dose-rate failure dose.

Two reports on the same bipolar part, taken at different dose rates, show different degradation at the same dose, and a reviewer attributes the difference to ELDRS. What else could explain it, and what would you check?

Dose-rate enhancement and ordinary annealing are different things: annealing between exposure and measurement can change a comparison between two tests without any ELDRS being involved, and bias, temperature, hydrogen in the package and the process history can all affect the response. Before attributing the difference, compare the dose-rate history, the bias during irradiation, the package and process identity, the temperature and any anneal for both tests, and confirm that the linear-circuit parameters measured were the same ones.

In a susceptible bipolar part, why does the damage show up as a loss of gain at a given collector current?

Holes generated by the radiation and hydrogen-related species take part in reactions in the oxide and in the buildup of interface traps near the emitter-base surface, and the transport and the reaction kinetics depend on the field, the time and the precursors available. The added surface recombination raises the base current, so the same collector current needs more base current and β falls. The mission consequence is an analog feedback path that loses gain or picks up an offset over a long mission even though it survived a short high-rate exposure in the laboratory.

Why does the course say that the decision to run a low-dose-rate exposure has to be made at the start of the program?

Because the low-dose-rate condition turns a half-hour exposure into weeks (lesson 6.2), so the result arrives in time to change the part or the circuit only if the test starts early. The dose rate of a test is one of the conditions that decide whether its result applies, and the assurance case has to say why the exposure that was chosen represents the mission, with both a physical and an experimental justification; a high-rate exposure can be finished quickly and, for a susceptible part, still not be conservative.

The ELDRS laboratory on the page shows two identical oxide blocks at the same dose and two dose rates ending in different states. Why does the course warn against reading an acceleration factor off that model?

The one-population annealing law shown in the laboratory leaves out the hydrogen reaction network on purpose; it illustrates one trade in the time history, it does not describe the physics of ELDRS, and it is not an acceleration law for qualification. An acceleration method is credited only when it has been demonstrated for the part in question, with its conditions and limits stated, and the alternative is low-dose-rate data that apply to the part. The laboratory's onset values and gains are teaching assumptions, not limits for flight hardware.

Further reading