Power Supplies

3.4374460741977 (1159)
Posted by r2d2 03/03/2009 @ 13:14

Tags : power supplies, components, hardware, technology

News headlines
Power supplies hit by explosion - BBC News
CE Electric spokesman Michael Sargood said it was working as quickly as possible to restore power to the affected homes and it was hoped that would be done by Saturday night. Once supplies have been restored an investigation would start into the cause...
IEA:Love Affair With Gadgets Squeezing Power Supply - CNNMoney.com
LONDON -(Dow Jones)- The rapidly growing number of televisions, laptops and other electronic gadgets in households is putting a growing strain on electricity supplies and jeopardizing efforts to cut carbon emissions, the International Energy Agency...
Green IT Is Key to an Energy-Efficient Future: Report - Reuters
It's not just data centers that are benefiting from efficient power supplies: gadgets of all types, from personal computers to cell phones, use an estimated $17 billion in electricity every year. While power supplies are currently only 40 percent...
Selecting the right topology for energy-saving power supplies - Embedded.com
Initiatives to reduce energy wastage by electronic systems are forcing designers of single-phase AC input power supplies to use newer power supply technologies. For higher power levels, these initiatives call for efficiency levels of 87% and above....
Seasonic SS-400H1U & SS-460H2U Power Supply Review - PC Perspectives
Seasonic has an excellent reputation for producing quiet ATX style PC power supplies with high efficiencies and stable outputs. However, Seasonic also has a full line of server style power supplies in both 1U and 2U form factors....
EPSMA revises design guidelines for power supply design - EE Times Deutschland
Winchester, UK - The European Power Supply Manufacturers Association (EPSMA) has published three new guides designed to help power supply design engineers keep up to date with regard to the latest safety standards and certifications....
Frequent power failures hit industrial production - Daily Times
The association strongly protests against unscheduled disruption of power supplies to its member units. The industrial production in the city could plunge by as much as 30 percent. Sectors like textile, leather and salt appear to be among the worst hit...
Electronic ballast circuits enhance resonant-mode power supplies - EDN.com
This article explains the basic functionality of a typical electronic ballast circuit, highlights the similarities between the two applications, and describes a typical resonant-mode power-supply solution using a standard electronic ballast control IC....
Falling Gas Prices Deny Russia a Lever of Power - New York Times
Throughout his eight years as president of Russia, Vladimir V. Putin pursued the strategic goal of dominating natural gas supplies to Europe and the pipelines that deliver them. His success was underscored in January, when for the second time in three...

Modular Power Supply Unit

Modular Power Supply Unit, or "MPS", is the term used to describe a type of power supply in which the user can modify the power connectors to suit their personal needs. In a "Modular Power Supply", there are detachable power connectors that can be changed depending on what amount of power or type of connector that a component requires.

There are features that are found in "MPS"s that people may associate with them that are not useable features. Of the currently available "Modular Power Supplies", some have UV-reactive or LED components included. These lighting devices can be found in the detachable cables or even inside of the unit itself. While these types of features are interesting to some extreme system builders, they are far from necessities.

Some companies such as Kingwin and Ultra Products have patented this idea for the consumer market.

By allowing cables to be detached, the clutter of unused power connectors that may damage other components such as fans or that may just be a hassle to move around within the computer's case can be eliminated. Also, by removing unneeded cables, these devices (when installed in wholesale computers) are more cost-effective to companies and consumers.

Modular cables and connectors add electrical resistance between the power supply and the hardware components, which could affect the performance of the power supply, but in the real world the difference is not noticeable as long as the connections were designed to a good standard.

To the top

Power supply unit (computer)

The top cover has been removed to show the internals of a computer Power supply Unit.

A power supply unit (PSU) is the component that supplies power to a computer. More specifically, a power supply is typically designed to convert 100-120 V (North America and Japan) or 220-240 V (New Zealand, Europe, South America, Africa, Asia and Australia) AC power from the mains to usable low-voltage DC power for the internal components of the computer. Some power supplies have a switch to change between 230 V and 115 V. Other models have automatic sensors that switch input voltage automatically, or are able to accept any voltage between those limits.

The most common computer power supplies are built to conform with the ATX form factor. The most recent specification of the ATX standard PSU as of mid-2008 is version 2.31. This enables different power supplies to be interchangeable with different components inside the computer. ATX power supplies also are designed to turn on and off using a signal from the motherboard, and provide support for modern functions such as the standby mode available in many computers.

Computer power supplies are rated based on their maximum output power. Typical power ranges are from 300 W to 500 W (lower than 300 W for Small form factor systems). Power supplies used by gamers and enthusiasts sometimes range from 500 W to 1300 W, with the highest end units going up to 2 kW for servers and extreme performance computers with multiple processors, several hard disks and multiple graphics cards (ATI CrossFire or NVIDIA SLI). The power rating of a PC power supply is not officially certified and is self-claimed by each manufacturer.A common way to reach the power figure for PC PSUs is by adding the power available on each rail, which will not give a true power figure. This means that you cannot use the PSU maximum rating on one rail, but only as a total. Therefore you can overload a PSU on one rail without having to use the maximum rated power.

Most computer power supplies have the appearance of a square metal box, and have a large bundle of wires emerging from one end. Opposite the wire bundle is the back face of the power supply, with an air vent and C14 IEC connector to supply AC power. There may optionally be a power switch and/or a voltage selector switch. A label on one side of the box lists technical information about the power supply, including safety certifications maximum output wattage. Common certification marks for safety are the UL mark, GS mark, TÜV, NEMKO, SEMKO, DEMKO, FIMKO, CCC, CSA, VDE, GOST R and BSMI. Common certificate marks for EMI/RFI are the CE mark, FCC and C-tick. The CE mark is required for power supplies sold in Europe and India.

A RoHS or 80 PLUS can also sometimes be seen.

Dimensions of an ATX power supply are 150 mm width, 86 mm height, and typically 140 mm depth, although the depth can vary from brand to brand.

Inside the computer power supply is a complex arrangement of electrical components, including diodes, capacitors, transistors and transformers. Also, most computer power supplies have metal heat sinks and fans to dissipate the heat produced. The speed of the fan is often dependent on the temperature, or less often the power load. It may be dangerous to open a power supply even if it is not connected to an electrical outlet, as high voltages may still be present in charged capacitors. However, for most PSUs this can be fixed by unplugging the PSU and then pressing the power-on button, which will drain the capacitors. Still, care should be taken as some PSUs require a load on the output in order to discharge the capacitors fully. Even when the PC is turned off, a PSU will draw some power from the electrical outlet, most of it going to power the +5 VSB (standby voltage) rail.

Some models even include heat pipes to assist in heat dissipation.

There are two basic differences between AT and ATX power supplies: The connectors that provide power to the motherboard, and the soft switch. On older AT power supplies, the Power-on switch wire from the front of the computer is connected directly to the power supply.

On newer ATX power supplies, the power switch on the front of the computer goes to the motherboard over a connector labeled something like; PS ON, Power SW, SW Power, etc. This allows other hardware and/or software to turn the system on and off.

The motherboard controls the power supply through pin #14 of the 20 pin connector or #16 of the 24 pin connector on the motherboard. This pin (Usually the green wire but can be the grey wire Dependant on PSU manufacturer) carries 5V when the power supply is in standby. It can be grounded (connected to any black pins) to turn the power supply on without having to turn on the rest of the components. This is great for testing if you don't have a spare motherboard around, or don't want to connect a suspicious power supply to a working motherboard with risk of damaging it.

AT means Advanced Technology. ATX is not an acronym but is the actual trademark name.

Most portable computers have power supplies that provide 15 to 100 watts. In portable computers (such as laptops) there is usually an external power supply (sometimes referred to as a "power brick" due to its similarity, in size, shape and weight, to a real brick) which converts AC power to one DC voltage (most commonly 19 V), and further DC-DC conversion occurs within the laptop to supply the various DC voltages required by the other components of the portable computer.

Computer power supplies are generally about 70–75% efficient. That means in order for a 75% efficient power supply to produce 75 W of DC output it would require 100 W of AC input and dissipate the remaining 25 W in heat. Higher-quality power supplies can be over 80% efficient; higher energy efficient PSU's waste less energy in heat, and requires less airflow to cool, and as a result will be quieter. As of 2007, 93%-efficient power supplies are available.

It's important to match the capacity of a power supply to the power needs of the computer. The energy efficiency of power supplies drops significantly at low loads. Efficiency generally peaks at about 50-75% load. The curve varies from model to model (for examples of how this curve looks see the test reports of energy efficient models found on the 80 PLUS website). As a rule of thumb for standard power supplies it is usually appropriate to buy a supply such that the calculated typical consumption of your computer is about 60% of the rated capacity of the supply provided that the calculated maximum consumption of the computer does not exceed the rated capacity of the supply. Note that advice on overall power supply ratings often given by the manufacturer of single component, typically graphics cards, should be treated with great scepticism. These manufacturers wish to minimise support issues due to under rating the supply and are willing to advise you to overrate it to avoid this.

Various initiatives are underway to improve the efficiency of computer power supplies. Climate savers computing initiative promotes energy saving and reduction of greenhouse gas emissions by encouraging development and use of more efficient power supplies. 80 PLUS certifies power supplies that meet certain efficiency criteria, and encourages their use via financial incentives.

The DIY boom has led to power supply manufacturers marketing their products directly to end-users, often with grossly inflated specifications. Some of the main tricks employed are...

So if...

This tendency has led in turn to greatly overspecified power supply recommendations, and a shortage of high-quality power supplies with reasonable capacities. Very few computers require more than 300–350 watts maximum. Higher end computers such as servers and gaming machines with multiple high power GPUs are among the few exceptions.

A modular power supply is a relatively new approach to cabling, allowing users to omit unused cables. Whereas a conventional design has numerous cables permanently connected to the power supply, a modular power supply provides connectors at the power supply end, allowing unused cables to be detached from the power supply, producing less clutter, a neater appearance and less interference with airflow. It also makes it possible to supply a wider variety of cables, providing different lengths of Serial ATA power connectors instead of Molex connectors.

While modular cabling can help reduce case clutter, they have often been criticized for creating electrical resistance. Some third party websites that do power supply testing have confirmed that the quality of the connector, the age of the connector, the number of times it was inserted/removed, and various other variables such as dust can all raise resistance. However, the amount of this resistance in a good connector is small compared to the resistance generated by the length of the wire itself.

Power supply testers are available. These typically have a single socket for each common type of power supply connector, and use several LEDs to indicate if the power supply is working.

Power supplies can also fail when the electrolytic capacitors inside dry up and/or become defective. This is one of the most common reasons powers supplies fail.

Most desktop computer power supplies are equipped with a cooling fan, which helps to keep internal components cool and operating more efficiently. Abnormal fan noise is generally caused by dust, a lack of internal lubrication or a failing motor. Dust may be removed by carefully blowing air through the supply with an air pump or gas duster, or by opening the PSU and using a brush.

It's relatively easy and inexpensive to lubricate/replace a fan, but not issue free. Opening a power supply can create a slight risk of shock, and voids the warranty if one is in force.

NOTE: NEVER OPEN A PSU. The capacitors inside retain an electric charge even hours after the system has been off and powered down. The charge is strong enough to cause severe damage when touched. Blowing out the dust without opening the PSU is the best option to quiet a noisy fan.

To the top


ATX form motherboards became increasingly popular because of their advantages over older AT motherboards.

The ATX (for Advanced Technology Extended) form factor was created by Intel in 1995. It was the first big change in computer case and motherboard design in many years. ATX overtook AT completely as the default form factor for new systems. ATX addressed many of the AT form factor's annoyances that had frustrated system builders. Other standards for smaller boards (including microATX, FlexATX and mini-ITX) usually keep the basic rear layout but reduce the size of the board and the number of expansion slot positions. In 2003, Intel announced the BTX standard, intended as a replacement for ATX. As of 2009 the ATX form factor remains the industry standard for do-it-yourselfers; BTX has however made inroads into pre-made systems, being adopted by computer makers like Dell, Gateway, and HP.

The official specifications were released by Intel in 1995, and have been revised numerous times since, the most recent being version 2.2, released in 2004.

A full size ATX board is 305 mm wide by 244 mm deep (12" x 9.6" ). This allows many ATX form factor chassis to accept microATX boards as well.

AT-style computer cases had a power button that was directly connected to the system computer power supply (PSU). The general configuration was a double-pole latching mains voltage switch with the four pins connected to wires from a four-core cable. The wires were either soldered to the power button (making it difficult to replace the power supply if it failed) or blade receptacles were used.

An ATX power supply does not directly connect to the system power button, allowing the computer to be turned off via software. However, many ATX power supplies have a manual switch on the back to ensure the computer is truly off and no power is being sent to the components. With this switch on, energy still flows to the components even when the computer appears to be "off." This is known as soft-off or standby and can be used for remote wake up through Wake-on-Ring or Wake-on-LAN, but is generally used to power on the computer through a front switch.

The power supply's connection to the motherboard was changed. Older AT power supplies had two similar connectors that could be accidentally switched, usually causing short-circuits and irreversible damage to the motherboard. ATX used one large, keyed connector instead, making a reversed connection very difficult. The new connector also provided a 3.3 volt source, removing the need for motherboards to derive this voltage from one of the other power rails. Some motherboards, particularly late model AT form factor offerings, supported both AT and ATX PSUs.

If you are not working with an ATX motherboard you can only turn on the power by shorting the green wire from the ATX connector to a black wire on the connector (or ground), since it is the motherboard's power switch which the ATX PSU uses. This means that an old PC power supply can be used for tasks other than powering a PC, but one must be careful to observe the minimum load requirements of the PSU.

ATX was originally designed with the power supply drawing air into the case and exhausting it down onto the motherboard. The plan was to deliver cool air directly to the CPU's and power regulation circuitry's location, which was usually at the top of the motherboard in ATX designs. This was not particularly useful for a variety of reasons. Early ATX systems simply didn't have processors or components with thermal output that required special cooling considerations. Later ATX systems with significantly greater heat output would not be aided in cooling by a power supply, because it would be delivering its often significantly heated exhaust into the case. As a result, the ATX specification was changed to make PSU airflow optional.

The power distribution specification defined that most of PSU's power should be provided on 5V and 3.3V rails, because most of the electronic components (CPU, RAM, chipset, PCI, AGP and ISA cards) used 5V or 3.3V for power supply. The 12V rail was only used by fans and motors of peripheral devices (HDD, FDD, CD-ROM, etc.).

The original ATX power supply specification remained mostly unrevised until year 2000.

While designing the Pentium 4 platform in 1999/2000, the standard 20-pin ATX power connector was deemed inadequate to supply increasing electrical load requirements. So, ATX was significantly revised into ATX12V 1.0 standard (that is why ATX12V 1.x is sometimes inaccurately called ATX-P4). ATX12V 1.x was also adopted by Athlon XP and Athlon 64 systems.

This is a minor revision from August 2000. The power on 3.3V rail was slightly increased, among other much lesser changes.

A relatively minor revision from January 2002. The only significant change was that the −5V rail was no longer required (it became optional). This voltage was very rarely used, only on some old systems with some ISA add-on cards.

ATX12V 2.x brought a very significant design change regarding power distribution. When analyzing the then-current PC architectures' power demands, it was determined that it would be much easier (both from economical and engineering perspectives) to power most PC components from 12V rails, instead of from 3.3V and 5V rails.

This is a minor revision from June 2004. The −5V rail was completely removed from the specification.

This is a minor revision from March 2005. The power was slightly increased on all rails. Efficiency requirements changed.

Another minor revision, the main change was a call for higher-quality connectors on the motherboard power connectors.

This is an ATX12V power supply derivative made by AMD to power its Athlon MP (dual processor) platform. It was used only on high-end Athlon MP motherboards. It has a special 8-pin supplemental connector for motherboard, so an AMD GES PSU is required for such motherboards (those motherboards will not work with ATX(12V) PSUs).

Defined in SSI, and used by some (Xeon and Opteron) systems. It has 24 pin main connector, 8 pin secondary connector, optional 4 pin tertiary connector.

Because video card power demands have dramatically increased over the 2000s, some high-end graphics cards have power demands that exceed AGP or PCIe slot capabilities. For these cards, supplementary power was delivered through a standard 4-pin peripheral or floppy power connector. Midrange and high-end PCI Express-based video cards manufactured after 2004 typically use a standard 6 or 8-pin PCIe power connector directly from the PSU.

Although the ATX power supply specifications are all vertically compatible in both ways (both electrically and physically), it is not wise to mix old motherboards/systems with new PSU's, and vice versa.

Special note: Proprietary brand-name or high-end workstation/server designs do not fit into these guidelines. They usually require an exactly matching power supply unit.

Older Dell computers, particularly those from the Pentium II and III times, are notable for using proprietary power wiring on their power supplies and motherboards. While the motherboard connectors appear to be standard ATX, and will actually fit a standard power supply, they are not compatible. Not only have wires been switched from one location to another, but the number of wires for a given voltage has been changed. Thus, the pins cannot simply be rearranged.

The change affects not only 20-pin ATX connectors, but also auxiliary 6-pin connectors. Modern Dell systems may use standard ATX connectors. Dell PC owners should be careful when attempting to mix non-Dell motherboards and power supplies, as it can cause damage to the power supply or other components. If the power supply color coding on the wiring does not match ATX standards, then it is probably proprietary. Wiring diagrams for Dell systems are usually available on Dell's support page.

On the back of the system, some major changes were made. The AT standard had only a keyboard connector and expansion slots for add-on card backplates. Any other onboard interfaces (such as serial and parallel ports) had to be connected via flying leads to connectors which were mounted either on spaces provided by the case or brackets placed in unused expansion slot positions. ATX allowed each motherboard manufacturer to put these ports in a rectangular area on the back of the system, with an arrangement they could define themselves (though a number of general patterns depending on what ports the motherboard offers have been followed by most manufacturers). Generally the case comes with a snap out panel, also known as an I/O plate, reflecting one of the common arrangements. If necessary, I/O plates can be replaced to suit the arrangement on the motherboard that is being fitted and the I/O plates are usually included when purchasing a motherboard. Panels were also made that allowed fitting an AT motherboard in an ATX case.

ATX also made the PS/2-style mini-DIN keyboard and mouse connectors ubiquitous. AT systems used a 5-pin DIN connector for the keyboard, and were generally used with serial port mice (although PS/2 mouse ports were also found on some systems). Many modern motherboards are phasing out the PS/2-style keyboard and mouse connectors in favor of the modern standard of USB ports. Other legacy connectors that appeared on ATX motherboards but are being phased out include 25-pin parallel ports and 9-pin RS-232 serial ports. In their place are on-board peripheral ports such as Ethernet, Firewire, External SATA, audio ports (analog/S/PDIF), video (D-sub/DVI/HDMI), and extra USB ports.

There exist several ATX-derived form factors that use the same power supply, mountings and basic back panel arrangement, but set different standards for the size of the board.

In CeBIT 2008, Foxconn unveiled a Foxconn F1 motherboard prototype, the same length as a standard ATX motherboard, but wider to accommodate 10 slots. In January 2008, Lian Li unveiled Armorsuit PC-P80 case with 10 slots designed for the motherboard.

To the top

Power semiconductor device

Power semiconductor devices are semiconductor devices used as switches or rectifiers in power electronic circuits (switch mode power supplies for example). They are also called power devices or when used in integrated circuits, called power ICs.

Most power semiconductor devices are only used in commutation mode (i.e they are either on or off), and are therefore optimized for this. Most of them should not be used in linear operation.

Power semiconductor devices first appeared in 1952 with the introduction of the power diode by R.N. Hall. It was made of Germanium and had a voltage capability of 200 volts and a current rating of 35 amperes.

The thyristor appeared in 1957. Thyristors are able to withstand very high reverse breakdown voltage and are also capable of carrying high current. One disadvantage of the thyristor for switching circuits is that once it is 'latched-on' in the conducting state it cannot be turned off by external control. The thyristor turn-off is passive, i.e., the power must be disconnected from the device.

The first bipolar transistors devices with substantial power handling capabilities were introduced in the 1960s. These components overcame some limitations of the thyristors because they can be turned on or off with an applied signal.

With the improvements of the Metal Oxide Semiconductor technology (initially developed to produce integrated circuits), power MOSFETs became available in the late 1970s. International Rectifier introduced a 25 A, 400 V power MOSFET in 1978. These devices allow operation at higher frequency than bipolar transistors, but are limited to the low voltage applications.

The Insulated Gate Bipolar Transistor (IGBT) developed in the 1980s became widely available in the 1990s. This component has the power handling capability of the bipolar transistor, with the advantages of the isolated gate drive of the power MOSFET.

Some common power devices are the power diode, thyristor, power MOSFET and IGBT (insulated gate bipolar transistor). A power diode or MOSFET operates on similar principles to its low-power counterpart, but is able to carry a larger amount of current and typically is able to support a larger reverse-bias voltage in the off-state.

Structural changes are often made in power devices to accommodate the higher current density, higher power dissipation and/or higher reverse breakdown voltage. The vast majority of the discrete (i.e non integrated) power devices are built using a vertical structure, whereas small-signal devices employ a lateral structure. With the vertical structure, the current rating of the device is proportional to its area, and the voltage blocking capability is achieved in the height of the die. With this structure, one of the connections of the device is located on the bottom of the semiconductor .

A second classification is less obvious, but has a strong influence on device performance: Some devices are majority carrier devices (Schottky diode, MOSFET), while the others are minority carrier devices (Thyristor, bipolar transistor, IGBT). The former use only one type of charge carriers, while the latter use both (i.e electrons and holes). The majority carrier devices are faster, but the charge injection of minority carrier devices allows for better On-state performance.

Moreover, the transition between on and off states should be instantaneous.

In reality, the design of a diode is a trade-off between performance in on-state, off-state and commutation. Indeed, it is the same area (actually the lightly-doped region of a PiN diode) of the device that has to sustain the blocking voltage in off-state and allow current flow in the on-state. As the requirements for the two state are completely opposite, it can be intuitively seen that a diode has to be either optimised for one of them, or time must be allowed to switch from one state to the other (i.e slow down the commutation speed).

This trade-off between on-state, off-state and switching speed is the same for all power devices. A Schottky diode has excellent switching speed and on-state performance, but a high level of leakage current in off-state. PiN diodes are commercially available in different commutation speeds (so-called fast rectifier, ultrafast rectifier...), but any increase in speed is paid by lower performance in on-state.

The trade-off between voltage, current and frequency ratings also exists for the switches. Actually, all power semiconductors rely on a PiN diode structure to sustain voltage. This can be seen in figure 2. The power MOSFET has the advantages of the majority carrier devices, so it can achieve very high operating frequency, but can't be used with high voltages. As it is a physical limit, no improvement is expected from silicon MOSFET concerning their maximum voltage ratings. However, its excellent performance in low voltage make it the device of choice (actually the only choice) for applications below 200 V. By paralleling several devices, it is possible increase the current rating of a switch. The MOSFET is particularly suited to this configuration because its positive thermal coefficient of resistance tends to balance current between individual devices.

The IGBT is a relatively new component, so its performance improves regularly as technology evolves. It has already completely replaced the bipolar transistor in power applications, and the availability of power modules (in which several IGBT dice are connected in parallel) makes it attractive for power levels up to several megawatts, pushing further the limit where thyristors and GTO become the only option. Basically, an IGBT is a bipolar transistor driven by a power MOSFET: it has the advantages of being a minority carrier device (good performance in on-state, even for high voltage devices), with the high input impedance of a MOSFET (it can be driven on or off with a very low amount of power). Its major limitation for low voltage applications is the relatively high voltage drop it exhibits in on-state (2 to 4 V). Compared to the MOSFET, the operating frequency of the IGBT is relatively low (few devices are rated over 50 kHz), mainly because of a so-called 'current-tail' problem during turn-off. This problem is caused by the slow decay of the conduction current during turn-off resulting from slow recombination of large number of carriers, which flood the thick 'drift' region of the IGBT during conduction. The net result is that the turn-off switching loss of an IGBT is considerably higher than its turn-on loss. Generally, in datasheet, turn-off energy is mentioned as a measured parameter and one has to multiply that number with the switching frequency of the intended application to estimate the turn-off loss.

At very high power levels, thyristor-based devices (SCR, GTO, MCT) are still the only choice. Though driving a thyristor is somewhat complicated, as this device can only be turned on. It turns off by itself as soon as no more current flows through it. This requires specific circuit with means to divert current, or specific applications where current is known to cancel regularly (i.e Alternating Current). Different solution have been developed to overcome this limitation (Mos Controlled Thyristors, Gate Turn Off thyristor...). These components are widely used in power distribution applications.

Research is also ongoing on electrical issues such as reducing the parasitic inductance of packaging. This inductance limits the operating frequency as it generates losses in the devices during commutation.

Low-voltage MOSFETs are also limited by the parasitic resistance of the packages, as their intrinsic on-state resistance can be as low as one or two milliohms.

Some of the most common type of power semiconductor packages include TO-220, TO-247, TO-262, TO-3, D2Pak, etc.

IGBTs are still under development and we can expect increased operating voltages in the future. At the high-power end of the range, MOS-Controlled Thyristor are promising devices. A major improvement over conventional MOSFET structure is achieved by employing superjunction charge-balance principle to the design. Essentially, it allows the thick drift region of a power MOSFET to be heavily doped (thereby reducing the electrical resistance for electron flow) without compromising the breakdown voltage. An adjacent region of similarly doped (but of opposite carrier polarity - holes) is created within the structure. These two similar but opposite doped regions effectively cancel out their mobile charge and develop a 'depleted region' which supports the high voltage during off-state. On the other hand, during conducting state, the relatively higher doping of the drift region allows easier flow of carrier thereby reducing on-resistance. Commercial devices, based on this principle, have been developed by International Rectifier and Infineon in the name of CoolMOSTM.

The major breakthrough in power semiconductor devices is expected from the replacement of silicon by a wide band-gap semiconductor. At the moment, silicon carbide (SiC) is considered to be the most promising. SiC Schottky diodes with a breakdown voltage of 1200 V are commercially available, as are 1200 V JFETs. As both are majority carrier devices, they can operate at high speed. Bipolar devices are being developed for higher voltages, up to 20 kV. Among its advantages, silicon carbide can operate at higher temperature (up to 400°C) and has a lower thermal resistance than silicon, allowing better cooling.

To the top

Uninterruptible power supply

A small free-standing UPS. The unit in the photo has IEC connector inputs and outputs

An uninterruptible power supply (UPS), also known as a battery back-up, provides emergency power and, depending on the topology, line regulation as well to connected equipment by supplying power from a separate source when utility power is not available. It differs from an auxiliary or emergency power system or standby generator, which does not provide instant protection from a momentary power interruption. A UPS, however, can be used to provide uninterrupted power to equipment, typically for 5–15 minutes until an auxiliary power supply can be turned on or utility power is restored.

While not limited to safeguarding any particular type of equipment, a UPS is typically used to protect computers, data centers, telecommunication equipment or other electrical equipment where an unexpected power disruption could cause injuries, fatalities, serious business disruption or data loss. UPS units come in sizes ranging from units which will back up a single computer without monitor (around 200 VA) to units which will power entire data centers or buildings (several megawatts).

UPS units are divided into categories based on which of the above problems they address. Manufacturers categorize their products in accordance with the number of power related problems they address.

The general categories of modern UPS systems are on-line, line-interactive, and standby. An on-line UPS uses a "double conversion" method of accepting AC input, rectifying to DC for passing through the battery (or battery strings), then inverting back to AC for powering the protected equipment. A line-interactive UPS maintains the inverter in line and redirects the battery's DC current path from the normal charging mode to supplying current when power is lost. In a standby ("off-line") system the load is powered directly by the input power and the backup power circuitry is only invoked when the utility power fails. Most UPS below 1 kVA are of the line-interactive or standby variety which are usually less expensive.

For large power units, Dynamic Uninterruptible Power Supply are sometimes used. A synchronous motor/alternator is connected on the mains via a choke. Energy is stored in a flywheel. When the mains power fails, an Eddy-current regulation maintains the power on the load. DUPS are sometimes combined or integrated with a diesel-genset, forming a diesel rotary uninterruptible power supply, or DRUPS.

Fuel cell UPS have been developed in recent years using hydrogen and a fuel cell as a power source, potentially providing long run times in a small space. A fuel cell replaces the batteries used in other UPS designs.

The Offline / Standby UPS (SPS) offers only the most basic features, providing surge protection and battery backup. Usually the Standby UPS offers no battery capacity monitoring or self-test capability, making it the least reliable type of UPS since it could fail at any moment without warning. These are also the least expensive, selling for as little as US$40. The SPS may be worse than using nothing at all, because it gives the user a false sense of security of being assured protection that may not work when needed the most.

With this type of UPS, a user's equipment is normally connected directly to incoming utility power with the same voltage transient clamping devices used in a common surge protected plug strip connected across the power line. When the incoming utility voltage falls below a predetermined level the SPS turns on its internal DC-AC inverter circuitry, which is powered from an internal storage battery. The SPS then mechanically switches the connected equipment on to its DC-AC inverter output. The switch over time is stated by most manufacturers as being less than 4 milliseconds, but typically can be as long as 25 milliseconds depending on the amount of time it takes the Standby UPS to detect the lost utility voltage.

The Line-Interactive UPS is similar in operation to a Standby UPS, but with the addition of a multi-tap variable-voltage autotransformer. This is a special type of electrical transformer that can add or subtract powered coils of wire, thereby increasing or decreasing the magnetic field and the output voltage of the transformer.

This type of UPS is able to tolerate continuous undervoltage brownouts and overvoltage surges without consuming the limited reserve battery power. It instead compensates by auto-selecting different power taps on the autotransformer. Changing the autotransformer tap can cause a very brief output power disruption, so the UPS may chirp for a moment, as it briefly switches to battery before changing the selected power tap.

Autotransformers can be engineered to cover a wide range of varying input voltages, but this also increases the number of taps and the size, weight, complexity, and expense of the UPS. It is common for the autotransformer to only cover a range from about 90v to 140v for 120v power, and then switch to battery if the voltage goes much higher or lower than that range.

In low-voltage conditions the UPS will use more current than normal so it may need a higher current circuit than a normal device. For example to power a 1000 watt device at 120 volts, the UPS will draw 8.32 amps. If a brownout occurs and the voltage drops to 100 volts, the UPS will draw 10 amps to compensate. This also works in reverse, so that in an overvoltage condition, the UPS will need fewer amps of current.

The Online UPS is ideal for environments where electrical isolation is necessary or for equipment that is very sensitive to power fluctuations. Although once previously reserved for very large installations of 10kW or more, advances in technology have permitted it to now be available as a common consumer device, supplying 500 watts or less. The Online UPS is generally more expensive but may be necessary when the power environment is "noisy" such as in industrial settings, for larger equipment loads like data centers, or when operation from an extended-run backup generator is necessary.

The basic technology of the online UPS is the same as in a Standby or Line-Interactive UPS. However it typically costs much more, due to it having a much greater current AC-to-DC battery-charger/rectifier, and with the rectifier and inverter designed to run continuously with improved cooling systems. It is called a Double-Conversion UPS due to the rectifier directly driving the inverter, even when powered from normal AC current.

In an Online UPS, the batteries are always connected to the inverter, so that no power transfer switches are necessary. When power loss occurs, the rectifier simply drops out of the circuit and the batteries keep the power steady and unchanged. When power is restored, the rectifier resumes carrying most of the load and begins charging the batteries, though the charging current may be limited to prevent the high-power rectifier from overheating the batteries and boiling off the electrolyte.

The main advantage to the on-line UPS is its ability to provide an electrical firewall between the incoming utility power and sensitive electronic equipment. While the Standby and Line-Interactive UPS merely filters the input utility power, the Double-Conversion UPS provides a layer of insulation from power quality problems. It allows control of output voltage and frequency regardless of input voltage and frequency.

Recently there have been hybrid topology UPSs hitting the marketplace. These hybrid designs do not have an official designation, although one named used by HP and Eaton is Double Conversion on Demand. This style of UPS is targeted towards high efficiency applications while still maintaining the features and protection level offered by double conversion.

A hybrid (double conversion on demand) UPS operates as an offline/standby UPS when power conditions are within a certain preset window. This allows the UPS to achieve very high efficiency ratings. When the power conditions fluctuate outside of the predefined windows, the UPS switches to online/double conversion operation. In double conversion mode the UPS can adjust for voltage variations without having to use battery power, can filter out line noise and control frequency. Examples of this hybrid/double conversion on demand UPS design are the HP R8000, HP R12000, HP RP12000/3 and the Eaton BladeUPS.

Ferro-resonant units operate in the same way as a standby UPS unit with the exception that a ferro-resonant transformer is used to filter the output. This transformer is designed to hold energy long enough to cover the time between switching from line power to battery power and effectively eliminates the transfer time. Many ferro-resonant UPSs are 90-93% efficient and offer excellent isolation.

This used to be the dominant type of UPS and is limited to around the 15KVA range. These units are still mainly used in some industrial settings due to the robust nature of the UPS. Many ferro-resonant UPSs utilizing controlled ferro technology may not interact with power-factor-correcting equipment.

A UPS designed for powering DC equipment is very similar to an online UPS, except that it does not need an output inverter, and often the powered device does not need a power supply. Rather than converting AC to DC to charge batteries, then DC to AC to power the external device, and then back to DC inside the powered device, some equipment accepts DC power directly and allows one or more conversion steps to be eliminated. This equipment is more commonly known as a rectifier.

Many systems used in telecommunications use 48 volt DC power, because it is not considered a high-voltage by most electrical codes and is exempt from many safety regulations, such as being installed in conduit and junction boxes. DC has typically been the dominant power source for telecommunications, and AC has typically been the dominant source for computers and servers.

There has been much experimentation with 48v DC power for computer servers, in the hope of reducing the likelihood of failure and the cost of equipment. However, to supply the same amount of power, the current must be greater than an equivalent 120v or 240v circuit, and greater current requires larger conductors and/or more energy to be lost as heat.

High voltage DC (380 volts) is finding use in some data center applications, and allows for small power conductors, but is subject to the more complex electrical code rules for safe containment of high voltages.

A Rotary UPS uses the inertia of a high-mass spinning flywheel to provide short-term ride-through in the event of power loss. The flywheel also acts as a buffer against power spikes and sags, since such short-term power events are not able to appreciably affect the rotational speed of the high-mass flywheel. It is also one of the oldest designs, predating vacuum tubes and integrated circuits.

It can be considered to be online since it spins continuously under normal conditions. However, unlike an electronic double-conversion UPS, it is only capable of providing reserve power for a few seconds before the flywheel has slowed and the protection fails. It is traditionally used in conjunction with standby diesel generators, providing backup power only for the brief period of time the engine needs to start running and stabilize its output.

The Rotary UPS is generally reserved for applications needing more than 10,000 watts of protection, to justify the expense of an extremely large and heavy power system that can only be transported by forklift or crane. A larger flywheel or multiple flywheels operating in parallel will increase the reserve running time, but at greatly increasing cost due to the size and weight of the precision-balanced flywheels.

Because the flywheels are a mechanical power source, it is not necessary to use an electric motor or generator as an intermediary between it and a diesel engine designed to provide emergency power. By using a transmission gearbox, the rotational inertia of the flywheel can be used to directly start up a diesel engine, and once running, the diesel engine can be used to directly spin the flywheel. Multiple flywheels can likewise be connected in parallel through mechanical countershafts, without the need for separate motors and generators for each flywheel.

They are normally designed to provide very high current output compared to a purely electronic UPS, and are better able to provide inrush current for inductive loads such as motor startup or compressor loads, as well as medical MRI and cath lab equipment. It is also able to tolerate short-circuit conditions up 17 times larger than an electronic UPS, permitting one device to blow a fuse and fail while other devices still continue to be powered from the Rotary UPS.

Its life cycle is usually far greater than a purely electronic UPS, up to 30 years or more. But they do require periodic downtime for mechanical maintenance (ball bearing replacement), while solid-state designs, using batteries, do not require downtime if the batteries can be hot-swapped, which is usually the case for larger units.

In case #3 the motor generator can be synchronous/synchronous or induction/synchronous. The motor side of the unit in case #2 and #3 can be driven directly by an AC power source (typically when in inverter bypass), a 6-step double-conversion motor drive, or a 6 pulse inverter. Case #1 uses an integrated flywheel as a short-term energy source instead of batteries to allow time for external, electrically coupled gensets to start and be brought online. Case #2 and #3 can use batteries or a free-standing electrically coupled flywheel as the short-term energy source.

UPSs can be equipped with maintenance-free capacitors to extend service life .

Many computer servers offer the option of redundant power supplies, so that in the event of one power supply failing, one or more other power supplies are able to power the load. This is a critical point - each power supply must be able to power the entire server by itself.

Redundancy is further enhanced by plugging each power supply into a circuit (i.e. to a different circuit breaker).

While it is common practice by uninformed people to plug each of these individual power supplies into one single UPS, redundant protection can be extended further yet by connecting each power supply to its own UPS. This provides double protection from both a power supply failure and a UPS failure, so that continued operation is assured. This configuration is also referred to as 2N redundancy. If the budget does not allow for two identical UPS units then it is common practice to plug one power supply into mains power and the other into the UPS.

When a UPS system is placed outdoors, it should have some specific features that guarantee that it can tolerate weather with a 'minimal to none' effect on performance. Factors such as temperature, humidity, rain, and snow among others should be considered by the manufacturer when designing an outdoor UPS system. Operating temperature ranges for outdoor UPS systems could be around −40 °C to +55 °C.

Outdoor UPS systems can be pole, ground (pedestal), or host mounted. Outdoor environment could mean extreme cold, in which case the outdoor UPS system should include a battery heater mat, or extreme heat, in which case the outdoor UPS system should include a fan system or an air conditioning system.

UPS systems can be designed to be placed inside a computer chassis. There are two types of Internal UPS. The first type is a miniaturized regular UPS that is made small enough to fit into a 5.25″ CD-ROM slot bay of a regular computer chassis. The other type are re-engineered switching power supplies that utilize dual power sources of AC and/or DC as power inputs and have an AC/DC built-in switching management control units.

All these connections are connected to a common point called 'Earth'.

Some locations (hospitals, police stations, fire stations, etc.) have standby generators. The voltage and frequency of the power produced depends on the engine speed, and the speed is controlled by a system called a governor. Some are mechanical, some are electronic. The job of the governor is to keep the voltage and frequency constant as the load changes. In a large hospital, for example, the startup surge of an elevator can cause short "blips"in the frequency of the generator. In the USA the AM, FM and TV broadcast stations have generators, and since AM transmitters change load with the audio level, the generator is constantly trying to correct the output voltage and frequency as the load changes. And generators are rarely replaced. It's not uncommon to see 40 or 50 year old generators in regular use. And note that 85% of the broadcast transmitter sites are unmanned.

Many UPS units are incompatible with generators. The designers have written the microprocessor code to where they need EXACTLY 50.0Hz (or 60.0Hz) power. If it is not, they don't "see" the incoming power at all.

Here is a typical scenario where this behavior is is critically bad: Things are normal, the generator is off and the broadcast station is running on mains power. When the station loses power the UPS switches to its batteries and keeps its load running. The generator starts up, and a minute or so later the transfer switch moves the load (the station equipment) from the dead mains input to the generator output. The problem is that the load is constantly changing, and the generator frequency is constantly drifting plus and minus 1/2 cycle or so around 50.0Hz (or 60.0Hz). The UPS decides that the power is "bad" and it's load remains on the UPS batteries. When the UPS batteries run out the loads that are plugged into the UPS either get switched to the mains input or just die, depending on the UPS design. This is despite the fact that there is plenty of slightly out-of-spec input power from the generator that the UPS could use to make DC (which has no frequency at all) and both charge the batteries and run the UPS inverter.

To complete the picture a timer starts when mains power returns, and 15 or maybe 20 minutes later (after the incoming power is stable) the transfer switch connects the station load back to mains power.

Only then does the UPS see power that meets its its overly tight specification and apply power to the load, and begin to recharge its batteries.

So the station has two outages: one of a minute or so due to the start-up and stabilize time of the generator, and a second one lasting a few tens of milliseconds when the transfer switch connects the station back to the mains power and shuts down the generator. The equipment that is plugged into the UPS does not see the first one, but instead has a major outage that starts when batteries die, and last as long as it takes to get mains power back. So why have a UPS at all?

UPS manufacturers that are known to have this design problem include APC, TrippLite, and Best (as shipped from the factory). The Best "Fortress" line (and maybe others) can have its tolerance window expanded.

The problems related to input frequency should only affect UPS not designed as "double conversion" (APC, BEST). These UPS have to generate an output frequency identical to the input frequency. If the input frequency is out of tolerance, or not properly recognized, UPS must transfer to battery. UPS designed with the "double conversion" topology could adapt to any input frequency and most of them do. They will create an output frequency according to internal clock.

A problem in the combination of "double conversion" UPS and generator is the voltage distortion created by the UPS. The input of a double conversion UPS is essentially a big rectifier. The current drawn by the UPS is non-sinusoidal. This causes the voltage from the generator also to become non-sinusoidal. The voltage distortion then can cause problem in all electrical equipment connected to the generator, including the UPS itself! This level of "noise" is measured in a percentage of "Total Harmonic Distortion of the current" (THD(i)). Classic UPS rectifiers have a THD(i) level around 25-30%. To prevent voltage distortion, this requires generators more than twice as big as the UPS.

There are several solutions to reduce the THD(i) in double conversion UPS: Classic solutions such as passive filters reduce (THD(i) to 5-10% at full load. They are reliable, but big and only work at full load and have there own problems with generators. Newer solution is an active filter. THD(i) can go down to 5% over full power range. Active filters are smaller, but they are expensive, consume energy and add components, thus risk of failures. The newest technology in double conversion UPS is a rectifier that doesn't use classic rectifier components (Thyristors and Diodes) but high frequency components (IGBT's). A double conversion UPS with an IGBT rectifier can have a THD(i) as small as 2%. This eliminates completely the need to oversize the generator (and transformers), without additional filters, investment cost, losses, space.

To the top

Source : Wikipedia