An previous noticed usually uttered by {an electrical} engineer (or their hapless techs) when an digital machine releases its magic smoke goes one thing like this: “The transistor bravely sacrificed itself to guard the fuse.”
This can be a succinct means of claiming that typical fuses (and most forms of circuit breakers) shouldn’t be counted on to guard a downstream load from full destruction within the occasion of a fault, however solely to (hopefully!) stop a hearth from overheated wiring, sparking, and so on.
This regrettable shortcoming is usually on account of physics and economics: a traditional fuse or thermal circuit breaker will take about 1,000 to 10,000 occasions too lengthy to guard most semiconductor units from overcurrent even when the latter are grossly over-sized (i.e. value far more than is economically justifiable). Specialised circuit breakers that mix a fast-acting magnetic journey together with the standard thermal one can enable for much less over-sizing of the downstream semiconductors, however they themselves are far more costly units, so as soon as once more this appears economically questionable.
To really defend the downstream semiconductor units from short-circuit faults with out affected by nuisance tripping throughout regular working circumstances requires an overcurrent protecting machine that responds to extreme overcurrents in 1 to 10 microseconds whereas nonetheless sustaining a extra typical fuse response to average overcurrents, which is exactly what an digital fuse can do—and doubtlessly at a aggressive value to the specialised magnetic-thermal circuit breaker, particularly if it permits using extra rationally-rated semiconductors.
Microchip’s Silicon Carbide E-Fuse demonstrator is on the market in six variants for 400–800 V battery methods.
To higher perceive the professionals and cons of digital fuses, a assessment of typical fuse and circuit breaker know-how is so as. All fuses basically depend on the melting of a fusible factor to interrupt the circuit throughout overcurrent circumstances. The period of time required is inversely proportional to the diploma of overcurrent. Because the heating of the fusible factor is the results of I2R losses, a doubling of present ought to lead to a quartering of the time to blow (that’s, following a sq. regulation). Certainly, fuses are most frequently characterised by I2t curves as a shorthand means of expressing the power required to rupture the fusible factor by leaving out the resistance, R, within the power equation of I2R over a given time, t.
E-fuses cannot solely reply far quicker to an overcurrent fault, they’ll additionally implement present vs time behaviors that may be impractical (or inconceivable) with typical applied sciences.
Numerous tips can be utilized to change this present vs time conduct—notably in lengthening the time that average overcurrents could be tolerated with out unduly compromising the response to shorts—however at larger overcurrents the period of time required for a fuse to open tends to an asymptotic minimal (as a result of want for the molten metallic to really fall far sufficient away to interrupt the circuit).
Moreover, if the fuse should break a DC circuit above 50 V or so, an arc will likely be developed because the fusible hyperlink melts, prolonging the period of time that present will likely be allowed by way of it (albeit with a voltage drop throughout stated arc). This arcing downside turns into more and more troublesome to handle with open-circuit DC voltage (it’s much less of a difficulty for AC, due to the periodic returning to 0 V), to the purpose that some type of compelled arc quenching is required above 300 VDC or so (by surrounding the fusible factor with quartz sand, or spring loading it, and so on).
Thermal circuit breakers additionally depend on I2R heating to journey, however this happens by way of the heating of a bimetallic strip that causes it to bend. Cheap thermal circuit breakers that aren’t anticipated to journey quite a lot of occasions of their operational life will make the most of the bimetallic strip instantly to maneuver the current-carrying contacts aside (by way of self-heating of the strip from load present), however extra sturdy designs—and particularly at larger currents—will use a separate heater coil across the strip which is able to itself actuate a toggling mechanism so that there’s a snap-action response upon reaching the crucial journey temperature, relatively than a extra gradual motion which might encourage arcing. The toggle mechanism permits for both a bistable motion, which means it’s steady in both the closed or open (tripped) positions, or a monostable motion, which means it’s going to mechanically return to the closed place after the bimetallic strip has cooled down.
Whatever the particular design of a thermal circuit breaker, it’s going to endure from some downsides that don’t afflict the easier fuse (or to not the identical diploma, anyway). For one factor, it’s going to take far longer to reply to a extreme overload or short-circuit than a fuse of the identical nominal present ranking. For one more, the journey present (or the time delay earlier than tripping happens) will likely be extra delicate to the ambient temperature, as a result of the temperature rise wanted to function the bimetallic strip should essentially be restricted in order to not create the very hearth hazard it’s supposed to stop.
Thermal-magnetic circuit breakers handle the extreme delay in responding to extreme overloads that’s attribute of purely thermal circuit breakers by passing the load present by way of a solenoid—that’s, a linear coil of wire round a shifting ferromagnetic plunger—which might independently function the toggling mechanism.
The magnetic power developed by a solenoid is proportional to the present flowing by way of it, so the upper the load present, the extra forcefully (and subsequently quicker) its plunger will transfer. The magnetic-hydraulic circuit breaker addresses the sensitivity of journey present to ambient temperature by changing the bimetallic strip with a viscous fluid that the solenoid plunger should function towards. This preserves the comparatively gradual response to average overloads typical of the thermal circuit breaker with the quicker response to short-circuits achieved by the magnetically-actuated mechanism.
The aforementioned shortcomings of typical fuse and circuit breaker designs could be addressed by going with an digital answer—that’s, by utilizing a semiconductor change to interrupt the circuit within the occasion of a fault situation—giving rise to what’s popularly known as an digital fuse, or e-fuse for brief (although it might be extra correct to explain it as an digital circuit breaker). E-fuses cannot solely reply far quicker to an overcurrent fault, they’ll additionally implement present vs time behaviors that may be impractical (or inconceivable) with typical applied sciences, in addition to defend towards extra hazards comparable to overvoltage, over- and under-temperature, and so on, at little to no additional value (i.e. with a couple of extra parts or extra traces of code). Moreover, e-fuses could value lower than typical electromechanical circuit breakers of comparable present interrupt ranking (although, admittedly, they examine much less favorably to standard fuses on this regard).
The only implementation of an e-fuse is a current-measuring circuit feeding a comparator that controls a semiconductor change. At working voltages as much as about 50 V and most currents as much as about 20 A, the present measuring circuit will sometimes be a resistive shunt that’s wired in sequence with the adverse line (aka the low facet), whose voltage drop is multiplied by an op-amp-based differential amplifier with a modest quantity of acquire (2x to 10x is typical right here) in order to not incur too excessive of an influence loss within the shunt. At larger voltages and/or currents, a present measuring circuit based mostly on a Corridor impact present sensor (or a competing know-how, comparable to big magnetoresistance) will likely be extra enticing, as a result of there isn’t a I2R loss as incurred by a shunt, and galvanic isolation is mechanically supplied. Both means, the output of the current-measuring circuit will likely be in comparison with a reference voltage by a comparator, which itself could be carried out with an analog IC (such because the venerable LM331 from Texas Devices), although it’s in all probability extra frequent to make use of a microcontroller IC as of late, particularly if different protecting features are to be carried out. A key requirement for the comparator perform is that it have hysteresis, or barely totally different turn-on and turn-off thresholds, to stop oscillation when the load present is close to the overcurrent journey level.
An e-fuse proof of idea for prime voltage 800 V automotive functions developed by Asahi Kasei Microdevices Company and Silicon Austria Labs GmbH.
The output of the comparator will probably be unable to instantly drive the semiconductor change, so some type of gate driver circuit or IC will likely be required, however this might merely be a complementary pair of bipolar transistors wired within the traditional push-pull configuration. The semiconductor change is mostly both a single low-side MOSFET (for blocking present circulate in a single course—i.e. solely discharge, from the angle of a battery), or two MOSFETs wired source-to-source (for bidirectional blocking). This latter configuration is often used within the BMS for Li-ion batteries (which, in spite of everything, incorporates the entire features of an e-fuse, along with its cell monitoring and balancing duties).
A much less apparent requirement for the semiconductor change is that the impedance of the upstream energy supply not be so low as to make it inconceivable for the e-fuse’s change to outlive a tough brief downstream.
At larger voltages, it’s extra traditional to place the change within the excessive facet (i.e. the optimistic line) and use both an remoted or level-shifting gate driver to manage it, particularly when the adverse line have to be earth grounded for security causes. One key requirement for the semiconductor change is that it should be capable of face up to the next peak present (or, extra appropriately, have the next fusing power—I2t once more) than the machine it’s defending; a much less apparent requirement is that the impedance of the upstream energy supply not be so low as to make it inconceivable for the e-fuse’s change to outlive a tough brief downstream.
Be aware that this latter requirement additionally applies to standard circuit breakers and fuses, however is much simpler to satisfy with units that don’t depend on energetic semiconductors. This means that the change(es) within the e-fuse will likely be beefier, and subsequently dearer, than the semiconductor units it’s supposed to guard, which might be economically questionable if not for the e-fuse’s prolonged safety capabilities and far quicker response velocity.
Lastly, e-fuses clearly require energy to function, and if the voltage of the upstream power supply isn’t instantly usable by the e-fuse circuitry (within the vary of three V to 30 V, say) then some sort of DC-DC converter may also be wanted, including to the general value whereas lowering reliability. This additionally signifies that an e-fuse can’t instantly change an current typical fuse or circuit breaker, because it wants one other wire again to the facility supply to finish its provide circuit (although that is sometimes simple to retrofit).
You will need to be aware that there are some energy circuit topologies during which no exterior fuse will be capable of defend its semiconductor units from failure, no matter fuse know-how or how briskly it could possibly open. The most typical instance of such are the switches in a 3-phase inverter, as they’re instantly preceded by a capacitor wired instantly throughout the provision rails. This capacitor basically presents a really low-impedance voltage supply to the switches, which implies it’s going to ship a really excessive peak present to them within the occasion of a fault comparable to overlapping conduction of each switches in a bridge leg or a turn-to-turn brief within the windings of the motor.
The one good answer for safeguarding the switches that are instantly fed by a voltage supply is energetic current-limiting on the particular person change stage (aka pulse-by-pulse present limiting). One other scenario during which an e-fuse would possibly fare poorly—particularly compared to its typical counterparts—is when subjected to a extreme voltage spike, comparable to from a close-by oblique lightning strike. That is primarily as a result of passive electrical (fuses) or electromechanical (circuit breakers) units are usually much more rugged than their digital counterparts (e-fuses). In conclusion, e-fuses aren’t a panacea, and so they gained’t prevent from dangerous design choices and part selections within the energy circuits they’re supposed to guard, however they might cut back general guarantee and technical assist prices, whereas probably costing much less up entrance, so that they’re positively value contemplating.
This text first appeared in Difficulty 69: July-September 2024 – Subscribe now.
