Power for Airborne Systems: MIL-STD-704

This technical brief discusses a widely used military standard for power-supply characteristics.

A critical part of any electronic design is ensuring that the power-supply circuitry can tolerate the input voltage that is made available to the system. Ideally, the electronics will function more or less normally under all expected conditions, but at minimum you want the boards to function most of the time and tolerate occasional deviations without malfunctioning in a dangerous way or experiencing physical damage.

The MIL-STD

If you’ve worked on military systems, or even if you’ve spent time around people who work on military systems, it’s quite possible that you’ve heard or seen the term “MIL-STD-704.” This is a very important document because it gives designers the information that they need to ensure that an airborne military system will be compatible with the power supply provided by the aircraft.

The first thing to understand about MIL-STD-704 is that it is not a “requirements document” for the characteristics of the power supply circuitry in an electronic system. Rather, MIL-STD-704 tells you what you can expect from the aircraft power supply itself, and then this information is used to design (or buy) an appropriate power supply for the system.

The current MIL-STD-704 revision is “F,” and you can view the document here. It’s not overwhelmingly long, and actually it is quite approachable because much of the information is presented in graphical format. It covers both AC and DC supplies; DC systems can be either 28 V or 270 V. In my experience the 28 V DC specs are the most widely used, so we’ll focus on those for the rest of the article.

Transients, Undervoltage, Overvoltage

The 28 V aircraft supply can deviate significantly from the nominal voltage, and the system’s power-supply circuitry must be able to cope with this.

Plot taken from the MIL-STD-704F document.

Under normal conditions, the voltage can be as high as 50 V and as low as 18 V. These extreme deviations will not occur for long periods of time, though, and the graph readily conveys the general deviation vs. duration trend: the maximum duration of the transient decreases as the magnitude of the transient increases.

The next plot addresses overvoltage and undervoltage. As you can see from the values on the horizontal axis, the difference between an “overvoltage/undervoltage condition” and a “transient” is duration; both events refer to a voltage that is higher or lower than the nominal value, but overvoltage and undervoltage are long-term conditions.

Plot taken from the MIL-STD-704F document.

Here we see that the supply can stay at 20 V for up to 100 seconds. This is a rather long time considering that 20 V is only ~71% of the nominal voltage, but you also have to keep in mind that the “28 V supply” isn’t exactly a 28 V supply: Table IV indicates that the steady-state voltage can be anywhere from 22 V to 29 V.

Designing for MIL-STD-704

It would not be easy to design and fully verify a MIL-STD-704-compliant power supply circuit, and my guess is that most systems that are required to be officially compliant will rely on modules sold by companies that have expertise in this field. However, incorporating transient suppression (for overvoltage transients) and extra capacitance (for smoothing ripples and maintaining adequate supply voltage when the input drops below nominal) are not unreasonable tasks, and the specs given in the document certainly provide enough information to guide the design process.