“Thermal fuses have been used successfully for over 150 years as a basic circuit protection device. Thermal fuses are effective, reliable, and easy to use, and are available in a variety of values ​​and versions to meet different design goals. However, the inevitable disadvantage of thermal fuses for designers looking to cut current at extremely fast rates is their ability to reset themselves and operate at relatively low currents. For these designers, Electronic fuses (often denoted eFuse or e-Fuse) are a good solution, sometimes replacing thermal fuses, but often in addition to thermal fuse functionality.

“

By Bill Schweber

Thermal fuses have been used successfully for over 150 years as a basic circuit protection device. Thermal fuses are effective, reliable, and easy to use, and are available in a variety of values ​​and versions to meet different design goals. However, the inevitable disadvantage of thermal fuses for designers looking to cut current at extremely fast rates is their ability to reset themselves and operate at relatively low currents. For these designers, electronic fuses (often denoted eFuse or e-Fuse) are a good solution, sometimes replacing thermal fuses, but often in addition to thermal fuse functionality.

The eFuse is based on the simple concept of sensing the current by measuring the voltage across a known resistor and then cutting it off through a field effect transistor (FET) when the current exceeds a design limit. eFuse has features, flexibility and functionality that thermal fuses cannot achieve.

This article will describe how electronic fuses work. Then, discuss the features, additional functions, and effective use of active circuit fuses. At the same time, this article will introduce electronic fuses and their effective use with solutions from Texas Instruments, Toshiba Electronic Devices and Storage, and STMicroelectronics as examples.

How do eFuses work?

The working principle of traditional thermal fuses is well known and simple: when the current through the fusible link exceeds the design value, the element is heated enough to melt. In this way, the current path is cut off and the current returns to zero. Depending on the fuse rating, type, and magnitude of overcurrent, thermal fuses can respond and open the current path within a few hundred milliseconds to a few seconds. Of course, as with all active and passive components, there are many variations, details, and shaded operations to choose from for this simple, purely passive device.

In contrast, electronic fuses work quite differently. Electronic fuses have some of the same functions, but also add different new functions and new features. The basic concept of the eFuse is also straightforward: the load current flows through the FET and a sense resistor, and the voltage across the sense resistor is monitored. When this voltage exceeds a preset value, the control logic turns off the FET and cuts off the current path (Figure 1). The FET is in series with the power line and load and must have very low on-resistance so it does not cause excessive current resistance (IR) drop or power loss.

Figure 1: In an electronic fuse, when the current from the power source to the load passes through the sense resistor, it is monitored by the voltage on the resistor; when the measured voltage exceeds the set value, the control logic turns off the FET to block the Current flows to the load. (Image credit: Texas Instruments)

It appears that the eFuse is just a more complex active version of the classic passive thermal fuse. Nonetheless, electronic fuses also have some unique properties:

Speed: fast response, its disconnection response time is microseconds, and some designs can reach nanoseconds. This property is important for today’s circuits using relatively sensitive ICs and passive components.

Low Current Operation: Electronic fuses can not only be designed for low current operation (approximately 100 milliamps (mA) or less), but they can also operate properly at low single-digit voltages. At these levels, thermal fuses often do not get enough self-heating current to cause the fusible link to blow.

Resettable: Depending on the model, the eFuse can choose to remain disconnected after activation (called latch mode), or resume normal operation after the current fault disappears (auto-restart mode). The latter setting is especially useful in transient inrush current situations without “hard” faults, such as when a board is plugged into a live bus. Also useful in situations where fuse replacement is difficult or expensive.

Reverse Current Protection: Electronic fuses can also provide reverse current protection, which thermal fuses cannot. Reverse current occurs when the system output voltage is higher than its input voltage. This is the case, for example, when a set of redundant power supplies is connected in parallel.

Overvoltage Protection: With some additional circuitry, the eFuse can also provide overvoltage protection against surge or inductive tripping, i.e., when the input voltage exceeds a set overvoltage trip point, the FET is turned off and maintained for the duration of the overvoltage condition in disconnected state.

Reverse Polarity Protection: eFuse can also provide reverse polarity protection, which quickly cuts off the current if the power supply is connected in reverse. For example, a car battery is briefly reversed due to accidental cable contact.

Incremental slew rate: Some advanced electronic fuses can also provide specified power-down/power-up current slew rates by external control or by using fixed elements to control the on/off switching of passive element FETs.

Therefore, eFuses are an attractive current control solution. Although in some cases these devices can replace thermal fuses, the two are often used in pairs. In this layout, eFuses are used to provide localized fast-response protection for subcircuits or PC boards, such as in hot-swap systems, automotive applications, programmable logic controllers (PLCs), and battery charge and discharge management; complementary thermal fuses Provides system-level protection against large-scale critical failures requiring hard permanent shutdown.

In this way, designers can achieve the best of both worlds, which is all the functionality of an electronic fuse plus the well-defined action of a thermal fuse. There are no technical trade-offs to achieve this, and there are no flaws. Of course, as with any design decision, there are some tradeoffs to consider. In this case, the increased footprint and bill of materials (BOM) will be slightly larger.

Select eFuse: Features and Applications

There are some basic parameters to consider when choosing an electronic fuse. Obviously, the primary consideration is the current level at which the fuse operates. Current levels are typically from under 1 ampere (A) to around 10A, and the highest terminal voltage the fuse can withstand. For some electronic fuses, the current level is fixed, while for other devices the current level can be set by the user through an external resistor. Other selection factors include response speed, quiescent current, size (package), and the number and type (if any) of external auxiliary components required. In addition, the designer must also consider any additional features and functions that different electronic fuse models may have.

For example, a PLC is an application where electronic fuses are useful in their various subcircuits, which are prone to sensor I/O and power misconnections. Additionally, current surges can occur when connecting wires or hot swapping boards. For example, the TPS26620 electronic fuse from Texas Instruments is commonly used in these types of 24V applications. As shown in Figure 2, the set current limit is 500mA. The fuse operates from 4.5V to 60V and has a maximum current of 80mA with programmable current limit, overvoltage, undervoltage and reverse polarity protection. The IC also controls inrush current and provides robust reverse current and field miswiring protection for PLC I/O modules and sensor power supplies.

Figure 2: The TPS26620 electronic fuse from Texas Instruments is shown with a 500mA trip current setting in this 24V DC PLC application. (Image credit: Texas Instruments)

The timing diagram for the Toshiba TCKE805 (18V, 5A electronic fuse) in Figure 3 shows how one supplier implements auto-restart and latch-up mode. In auto-restart mode (set by the EN/UVLO package pin), an overcurrent protection function protects the electronic fuse and load from damage by suppressing power dissipation during fault conditions.

Figure 3: The Toshiba TCKE805 18V, 5A electronic fuse uses a “test and repeat cycle” sequence to assess whether the recovery current is safe. (Image credit: Toshiba)

If the output current set by the external resistor (RLIM) exceeds the current limit (ILIM) due to load failure or short circuit, the output current and output voltage drop, limiting the power consumption of the IC and the load. When the output current reaches a preset limit and an overcurrent is detected, the output current is clamped so that the current flowing does not exceed ILIM. If the overcurrent problem is not resolved at this stage, this current clamping state is maintained and the temperature of the electronic fuse continues to rise.

When the temperature of the electronic fuse reaches the operating temperature of the thermal shutdown function, the eFuse MOSFET is turned off and the current is completely cut off. The auto-restart feature attempts to restore current flow by blocking this current, which reduces the temperature and relieves thermal shutdown. If the temperature rises again, repeat the above actions and stop operation until the overcurrent condition is removed.

Conversely, latch mode clamps the output until the electronic fuse is reset via the IC’s enable (EN/UVLO) pin (Figure 4).

Figure 4: Unlike auto-restart mode, Toshiba e-fuses in latch mode only reset when commanded from the IC enable pin. (Image credit: Toshiba)

Some electronic fuses are specially configured to overcome problems related to sense current on the resistor, such as IR drop, which reduces the supply rail voltage on the output side. For example, STMicroelectronics’ 3.3V STEF033AJR has a maximum nominal current and FET on-resistance values ​​of 3.6A and 40 milliohms (mΩ), respectively, for the DFN package version, and 2.5A and 25mΩ for the flip-chip package version. . In the conventional connection shown in Figure 5, even a modest IR drop of about 15 millivolts (mV) in the power rail through the on-resistance can be noticeable and worrying at higher currents.

Figure 5: In the conventional wiring of the STEF033AJR, the resistor R-lim used to determine the current limit is placed between the two designated terminals. (Image credit: STMicroelectronics)

In traditional wiring, a resistor is placed between the positive-side limit connection and the output voltage connection (VOUT/Source), which can be modified to achieve a Kelvin sense layout that compensates for IR drop (Figure 6).

Figure 6: To reduce the effect of the current sense IR drop, the negative side of the current limiting resistor is connected to the voltage output (VOUT/Source). (Image credit: STMicroelectronics)

Note that while electronic fuses are semiconductor devices that operate at single-digit voltages, they are not limited to this low-voltage region. For example, the Texas Instruments TPS2662x series of electronic fuses are rated to operate from 4.5V to 57V.

eFuse: Build or buy?

In principle, a basic eFuse can be built from discrete components with several FETs, a resistor, and an Inductor. The first electronic fuses were made this way, where the inductor served two purposes: filtering the DC output, and using the DC resistance of its windings as a sense resistor.

However, a robust electronic fuse that takes into account component characteristics and actual operating conditions requires more than a few discrete components. Even if components are added, only the basic functions of electronic fuses can be realized (Figure 7).

Figure 7: The inherent limitations of electronic fuses that use discrete components for their basic functions must be foreseen and overcome. (Image credit: Texas Instruments)

In fact, with the proliferation of active and passive discrete components, not only do devices quickly become bulky, but they are also prone to performance variations between individual products, as well as differences from initial tolerances, component aging, and temperature. Drift and other related issues. In summary, there are a number of limitations to DIY “manufactured” discrete solutions:

Discrete circuits generally use P-channel MOSFETs as switching elements. P-channel MOSFETs are more expensive than N-channel MOSFETs in terms of achieving the same on-resistance value (RDS(ON)).

The discrete solution is inefficient because it causes power dissipation in the diodes and a responsive board temperature rise.

For discrete circuits, it is difficult to provide adequate thermal protection for passive component FETs. Therefore, this critical improvement cannot be made, or the design size has to be substantially increased to provide a suitable Safe Operating Area (SOA).

A comprehensive discrete circuit requires many components and considerable board space, and achieving robustness and reliability of the protection circuit requires additional components.

Although the output voltage slew rate in discrete designs can be adjusted using resistive-capacitor (RC) components, the dimensions of these components must be determined with a careful understanding of the gate characteristics of passive FETs.

Even if the discrete component approach is acceptable, its functionality will still be limited compared to the IC approach. The latter may include some or all of the additional functions described above, as shown in the electronic fuse block diagram in Figure 8. In addition, IC solutions are smaller, have more stable performance when fully characterized, and can be implemented with “peace of mind” at a lower cost, which is not possible with multi-device solutions. It is worth noting that the TPS26620 data sheet provides dozens of performance and timing diagrams covering a wide range of operating conditions, which are difficult to provide with discrete “manufacturing” methods.

Figure 8: The simplicity of a full-featured electronic fuse belies its internal complexity, which is not possible with discrete components. (Image credit: Texas Instruments)

Another key reason to buy a standard e-fuse IC instead of going the DIY discrete route: regulatory approval. Many fuses (thermal fuses and electronic fuses) are used for safety-related functions to prevent excessive current flow that could cause components to overheat and possibly catch fire, or cause injury to the user.

All conventional thermal fuses are approved by various regulatory agencies and standards and provide fail-safe current shutdown when used appropriately. However, obtaining the same approvals for a discrete solution is very difficult and time-consuming, if not impossible.

In contrast, many electronic fuse ICs have been approved. For example, the TPS2662x series of electronic fuses are recognized by UL 2367 (“Solid State Overcurrent Protectors for Special Purposes”) and IEC 62368-1 (Audio/Video, Information and Communication Technology Equipment – Part 1: Safety Requirements). The series also complies with IEC61000-4-5 (“Electromagnetic Compatibility (EMC) – Part 4-5: Test and Measurement Techniques – Surge Immunity Test”). To be certified, these electronic fuses have passed performance tests for their essential role under conditions including minimum and maximum operating temperatures, minimum and maximum storage and transport temperatures, extensive anomaly and durability testing, and thermal cycling.

Summary of this article

The eFuse uses an active circuit instead of a fusible link to interrupt the current flow and is designed to help designers meet requirements such as fast shutdown, self-reset, and reliable operation at low current conditions. Electronic fuses can also have various protection functions, as well as adjustable slew rates. Therefore, such devices are an important addition to an engineer’s circuit and system protection device package.

As mentioned above, electronic fuses can replace traditional thermal fuses, although in many cases they are only used for local protection and supplemented by thermal fuses. Like conventional thermal fuses, many electronic fuses are certified for safety-related functions, extending versatility and applicability.