Switch-mode power supplies (SMPS) continue to replace linear-regulator types in a host of applications. As the need for more efficient electronics accelerates and as a result of their size, weight, and energy-saving advantages, SMPS are being widely used in applications such as LCD TV monitors, PC/ laptop displays, portable electronics chargers, printers, DVD recorders, and even automotive electronics and industrial.
Yet because these new SMPS lack the inherent resistance of linear-regulator designs, there’s a growing need for proper circuit protection. Engineering schools generally do not emphasize selection and application of circuit protection devices. As a result, too often many SMPS lack adequate protection.
Whether they are external or internal, forward or fly-back conversion units, continuous or discontinuous, each type of SMPS is subject to regulatory requirements. For instance, a power supply for a telecom installation will potentially need to comply with Telcordia or ITU requirements, depending on the target region and application. In the consumer sector, IEC, UL, and CSA standards govern at the equipment level, and a host of other safety testing will be required. There are also IEC, UL, and CSA standards that apply specifically to circuitprotection components.
Fuses are ideal candidates for overcurrent protection in SMPS because of their proven safety, reliability, low resistance, small size, and cost effectiveness. Just as there are system-level requirements for the SMPS, there are regulatory requirements for safety and performance at the component level. For example, IEC-60127 gives specific dimensional requirements and also specifies a series of fuse tests.
CIRCUIT PARAMETERS OVERVIEWFigure 1 shows the location of a fuse in the ac-mains input of an ac-dc SMPS, such as in a cell-phone charger. The fuse positively interrupts the current flow when a component such as the radio-frequency interference (RFI) choke or filter capacitor fails in the short circuit mode.A metal oxide varistor (MOV) in the ac-mains input suppresses transient voltages associated with lightning or load switching. Additional common components, transient-suppression diodes on the internal dc bus, further suppress any transients, providing a higher degree of protection for the dcdc converter circuitry.
Figure 2 shows an embedded ac-dc SMPS that may be found in a server. In addition to the fuse at the ac mains input, fuses find employment on the high-voltage dc bus and in the housekeeping power supply.
To select the proper circuit protection device, it is necessary to accurately define the key parameters of the application. These include circuit voltage, maximum normal operating current, maximum potential fault current, maximum operating temperature, pulse currents, and mounting/form factor.
Circuit voltage is the source voltage driving the circuit. For safety concerns, it is critical to know because the fuse’s voltage rating must be equal to or greater than the circuit voltage. It is also extremely important that fuses in dc applications have an adequate dc-voltage rating.
Maximum normal operating current is the maximum RMS current under full load in normal operation. Maximum potential fault current is the expected maximum current with the source voltage shorted out. Obviously, maximum operating temperature is the anticipated operating temperature of the circuitry near the protection device, under full load, with all shields and covers in place, at maximum ambient temperature.
Pulse currents are the transients induced by normal switching events in the circuitry as well as those coupled to the ac mains from lightning and load switching. Both the magnitude and duration of the transients and the anticipated number of transients over the lifetime of the equipment need careful consideration.
Finally, choosing the mounting method and/or form factor of the fuse comes into play. Options vary to include surface-mount devices, pin-through-hole, pigtail leads, or simply a fuse and a holder. Also critical is the amount of space available for mounting the device.
FUSE SELECTION The fuse selection process usually begins by satisfying three basic selection criteria. First, what standards must be satisfied? For most SMPS products, the answer is a wide variety of international standards, which leads to a fuse that complies with IEC-60127. Several families of fuses have European, Asian, and North American safety agency approvals to IEC-60127.Second, what is the maximum operating source voltage? The most common answer for ac-mains input protection for a SMPS destined for the international market is 250 V ac. Third, what is the desired mounting/form factor? The most popular choice for a SMPS is radial or axial lead for direct circuit-board attachment, as small as possible, driven largely by economic and available space concerns.
The next step in the selection process is to determine the maximum potential fault current. It is best to do this by analyzing or measuring the path impedance across the input to the SMPS with the SMPS disconnected from the ac mains. This impedance, when added to the estimated impedance of the ac mains, enables one to calculate the maximum potential fault current.
Continue to page 2
Let’s assume that this drives the decision to use a high-interrupting-capacity 5- by 20-mm fuse with an interrupting rating of 1500 A at 250 V ac. A fast-acting fuse is tentatively the favored component since it is the most economical and provides the closest degree of protection.
Once we select a fuse family, we need to estimate the current rating within the family. The initial estimate will involve the calculation or direct measurement of the highest rating that will still provide the required protection. Direct measurement can be accomplished by observing the actual current waveform with a storage oscilloscope on the line side of the acmains input using a Pearson current probe or a shunt resistor and differential probe.
To measure the maximum normal operating current, the SMPS should be operating at full load. Figure 3 shows typical voltage and current levels for a desktop computer SMPS as they would be viewed on the oscilloscope. The yellow waveform (channel one) is the voltage across a 0.01-O shunt, which translates into 0.763 ARMS.
Since fuses that meet IEC-60127 operate right at their current rating, the next highest rating in the fuse family is 0.8 A. It is important to note that using a fuse meeting UL/CSA standards means a factor of 0.75 would apply since those fuses operate at 75% or less of their rated current, which means that a 1-A fuse would be appropriate.
Two additional factors require consideration before one can finalize the fuse’s ampere rating—the effects of pulse cycle withstand and temperature. Referring back to the current waveform in Figure 3, the current pulses have a magnitude of 2 A and a duration of 2 ms, which can be approximated by a 2-A by 1-ms square pulse with an I2t value of 0.004 A2s. This is approximately 2% of the nominal melting I2t of the chosen fuse, which does not pose a pulse withstand issue. (Refer here to the manufacturer’s pulse cycle withstand capability information.)
There is also a requirement that the ac input survive five simulated lightning surges of 8 by 20 µs at 1000-A peak with an I2t value of about 5.6 A2s. This drives the fuse selection away from an 0.8-A fastacting fuse to a 3.15-A fast-acting fuse with a nominal melting I2t of 7.9 A2s or a 2-A time lag fuse with a nominal melting I2t of 7A2s. The 2-A time lag fuse is more reliable as it is closer to the normal operating current and will supply better protection in the event of an overload fault.
For information on the effects of ambient temperature on current rating, refer to the manufacturer’s temperature de-rating curves. Re-rating based on temperature varies with the characteristic of the device. Apply the appropriate rerating factor on the type of fuse selected and the actual measured temperature near the circuit protection device, under full load, with all shields and covers in place, at maximum ambient temperature.
TESTING THE APPLICATION Upon fuse selection, one needs to test to prove the concept. One approach is to monitor load current through the application, and the fuse, via a Pearson current probe and simultaneously monitor the voltage across the fuse using a differential probe.While operating the load at maximum current draw, monitor the current to ensure consistent operation and monitor voltage drop across the fuse to ensure a minimal shift. Check the cold resistance of the fuse both before and after testing to ensure the fuse has not opened after testing.
While a fuse does not prevent a fault from occurring, it will operate quickly to prevent further damage. By doing so, it maintains the safety of the application and can help to prevent collateral equipment damage. The fuse will also limit the extent of the damage to the application to as small a portion of it as possible, making repairs less costly. Proper selection of the fuse is critical. With an understanding of both the application and the fuse, finding a suitable match should be a quick and easy task.