Mean Time Between Failures (MTBF) is the reliability metric that defense procurement engineers live by. It appears in contract specifications, it drives logistics planning, and it determines whether a power supply is considered mission-capable or a potential risk. Suppliers can put a number in a datasheet, but what separates ACT from most of the field is how that number is derived, what engineering decisions drive it, and how ACT designs its products to sustain it across programs that run a decade or longer.
MTBF quantifies the average time a repairable system will operate before a failure requires intervention. In practice, there are two ways to arrive at that number, and understanding the difference is essential to evaluating any supplier’s reliability claims.
Calculated MTBF is a prediction. ACT engineers use parts count and stress analysis methods to model the failure rate of every component in a design: resistors, capacitors, semiconductors, magnetics, connectors. Individual component failure rates are summed to produce a total failure rate expressed in failures per million hours, which is then converted to an MTBF figure. This analysis happens after a prototype has been fielded and prior to the design verification release to eliminate any components that are not properly derated.
Demonstrated MTBF is a measurement. It comes from accelerated life testing, controlled stress testing, or field tracking of actual units in service, and it captures what actually happens when manufacturing variances, environmental stresses, and real operating conditions are factored in. A supplier whose calculated and demonstrated MTBF numbers align closely has both good models and good manufacturing. A supplier whose numbers diverge significantly has a problem somewhere in that chain.
The standard most commonly associated with defense reliability prediction is MIL-HDBK-217F. The problem with relying on it exclusively is that its underlying failure rate data was established in the 1970s and 1980s, a period when many modern semiconductor devices were in early development and carried substantially higher failure rates than today. As technology improved over subsequent decades, the handbook’s failure rate tables did not keep pace. Calculations based solely on MIL-HDBK-217F can produce reliability predictions that are systematically pessimistic for modern components, skewing the MTBF figure away from what the design will actually deliver.
ACT engineers reliably supplement MIL-HDBK-217F with Telcordia (formerly known as Bellcore), which maintains more current base failure rates for modern components. The result is a prediction that reflects the actual reliability of the parts being used more accurately. For customers whose programs are evaluated against MTBF targets, the difference between an outdated prediction method and a current one can be meaningful in terms of whether the design appears to meet the requirement on paper and whether it actually does in the field.
Reliability prediction is not a passive exercise at ACT. The analysis actively identifies components operating under elevated stress conditions and gives engineers specific targets for improvement. Two factors dominate the outcome: thermal management and component quality level.
Thermal stress is embedded in the reliability calculation for every component in the design. Components running at elevated junction temperatures accumulate failure rates faster than components running at lower junction temperatures. ACT’s thermal design practice, which includes component placement, heatsinking, and derating to keep junction temperatures well below maximum ratings, translates into improved MTBF figures and longer service in the field. A power supply that runs cool is simply more reliable.
Component quality has an equally significant effect. Failure rates for capacitors and semiconductors can vary significantly. So, ACT engineers specify components at the quality level the application requires. For military and aerospace programs, MIL-PRF-38535 qualified integrated circuits are the benchmark: parts screened through rigorous production controls, lot testing, and a qualified manufacturer list that ensures consistency. Where MIL-PRF-38535 parts are unavailable, ACT engineers evaluate automotive-grade alternatives qualified to AEC-Q100 or AEC-Q101, which impose thermal cycling, vibration, and lot testing requirements at increased batch sizes, ensuring a reduction in premature failure risk compared to a standard industrial component.
The tantalum capacitor is a useful illustration of why component quality decisions matter. Early tantalum capacitors using manganese dioxide dielectric showed a well-documented failure mode when operated near their voltage rating: thin dielectric layers with microscopic impurities would break down, causing the capacitor to fail. The engineering response was a 50 percent voltage derating rule that became standard practice. ACT engineers apply derating rules like this systematically because the failure modes are real and the consequences of ignoring them are serious.
Defense programs run long. A power supply qualified today may be expected to remain in service for ten years or more, supporting a platform through its full operational life. MTBF requirements set at the beginning of the program reflect that reality, and ACT’s engineering process is built around it.
ACT engineers design to keep every component operating below its derated electrical and thermal limits across the full range of possible operating conditions. This is a quantifiable decision where every degree of temperature reduction and every reduction in voltage/current stress has a calculable effect on the component’s failure rate. ACT engineers make those tradeoffs explicitly, with the reliability model in hand.
For programs with lifecycle requirements extending beyond a decade, this discipline delivers benefits that compound. A design engineered with margin against those stresses from the outset sustains its reliability performance through the whole program lifecycle.
Some suppliers calculate MTBF after the fact, just to satisfy a contract deliverable. Others use it throughout development as an active design tool. At ACT, we take the second approach. Reliability prediction helps us identify which components are driving the failure rate, flag parts operating outside their derated limits, and quantify the reliability benefit of proposed design changes before they’re built into hardware. The MTBF number on the final datasheet is the result of that process, not a number calculated afterward to check a box.
For defense engineers evaluating suppliers, this distinction is meaningful. Ask how the MTBF was calculated, which standard was used, what component quality levels were assumed, and how thermal stress was factored in. The answers reveal whether the number reflects genuine reliability engineering or a calculation performed to meet a specification requirement. ACT’s engineers can walk through every element of that analysis because they built it into the design from the beginning.
If your program has specific MTBF requirements, contact us to discuss how our reliability engineering process applies to your application.
