What is a power supply with an active PFC Power Factor Correction module? How to choose a power supply for your computer

Good afternoon, friends!

Surely many of you have seen the mysterious letters “PFC” on your computer power supply. Let’s say right away that these letters most likely will not be on the cheapest blocks. Do you want me to tell you this terrible secret? Listen!

What is PFC?

PFC is an acronym for Power Factor Correction. Before we decipher this term, let’s remember what types of power there are.

Active and reactive power

Back in our school physics course, we were told that power can be active and reactive.

Active power does useful work, particularly by releasing it as heat.

Classic examples are an iron and an incandescent lamp. An iron and a light bulb are almost purely resistive loads; the voltage and current across such a load are in phase.

But there is also a load with reactivity - inductive (electric motors) and capacitive (capacitors). In reactive circuits there is a phase shift between current and voltage, the so-called cosine φ (Phi).

The current can lag behind the voltage (in an inductive load) or lead it (in a capacitive load).

Reactive power does not produce useful work, but only dangles from the generator to the load and back, uselessly heating the wires .

This means that the wiring must have a spare cross-section.

The greater the phase shift between current and voltage, the more power is wasted uselessly on the wires.

Reactive power in the power supply

But in a building where a hundred or a thousand computers are installed, reactive power must be taken into account!

The typical value of the cosine Phi for computer power supplies without correction is about 0.7, i.e. the wiring must be designed with a 30% power reserve.

However, the matter is not limited to excessive load on the wires!

In the power supply itself, the current flows through the high-voltage input circuits in the form of short pulses. The width and amplitude of these pulses can vary depending on the load.

Large current amplitudes adversely affect high-voltage capacitors and diodes, reducing their service life. If rectifier diodes are selected “back to back” (which often happens in cheap models), then the reliability of the entire power supply is further reduced.

How is power factor correction performed?

To combat all these phenomena, devices that increase the power factor are used.

They are divided into active and passive.

The passive PFC circuit is a choke connected between the rectifier and the high-voltage capacitors.

A choke is an inductance that has reactive (more precisely, complex) resistance.

The nature of its reactivity is opposite capacitance capacitors, so some compensation occurs. The inductance of the inductor prevents the current from increasing, the current pulses are slightly stretched, and their amplitude decreases.

However, the cosine φ increases slightly and big win according to reactive power does not occur.

For more significant compensation they will apply active PFC circuits.

The active circuit increases the cosine φ to 0.95 and higher. The active circuit contains a boost converter based on inductance (inductor) and power switching elements, which are controlled by a separate controller. The throttle periodically either stores energy or releases it.

At the PFC output there is a filtering electrolytic capacitor, but of smaller capacity. A power supply with active PFC is less sensitive to short-term “drops” in the supply voltage I, which is an advantage. However, the use of an active circuit increases the cost of the design.

In conclusion, we note that the presence of PFC in a particular power supply can be identified by the letters “PFC” or “Active PFC”. However, there may be cases when the inscriptions do not correspond to reality.

Unambiguously judge the presence passive circuit it can be determined by the presence of a fairly weighty throttle, and an active one - by the presence of another radiator with power elements (there should be three of them in total).

That's it, friends! Tricky computer unit nutrition is arranged, isn't it?

All the best!

See you on the blog!

PFC- this is Power Factor Correction, which is translated from English. As "Power factor correction", the name "Reactive power compensation" is also found.
In relation to switching power supplies, this term means the presence in the power supply of a corresponding set of circuit elements, which is also commonly called “PFC”. These devices are designed to reduce the reactive power consumed by the power supply. Power supplies without PFC create powerful impulse noise on the electrical network for parallel-connected electrical appliances.
For quantification introduced distortions and interference there is a power factor (KM or Power Factor). Actually, the factor (or power factor) is the ratio of active power (power consumed irrevocably by the power supply) to total, i.e. to the vector sum of active and reactive powers. Essentially, the power factor (not to be confused with efficiency!) is the ratio of useful and received power, and the closer it is to unity, the better.

Varieties of PFC

PFC comes in two varieties - passive and active.
The simplest and therefore most common is the so-called passive PFC. Passive PFCs are made on a reactive element - a throttle. Unfortunately, to obtain acceptable efficiency, its dimensions are commensurate with the dimensions of the transformer version of this power supply, which is not economically profitable. The large geometric dimensions of the inductor are obtained because it must operate at a frequency of 50Hz (more precisely, 100Hz due to doubling the frequency after rectification) and it cannot in any way be smaller than the corresponding transformer for the same power. Quite often, under the guise of a “passive PFC”, a power supply unit hides a very small inductor. More precisely, there cannot be a choke of sufficient size due to the very limited space in the body of this power supply. Such a decorative PFC can spoil the dynamic characteristics of the power supply or cause unstable operation.

Active PFC represents another pulse source supply, and increasing the voltage.
In addition to the fact that active PFC provides a power factor close to ideal, also, unlike passive, it improves the performance of the power supply - it additionally stabilizes the input voltage of the main stabilizer of the unit - the unit becomes noticeably less sensitive to low mains voltage, also when using active PFC units are quite easily developed with universal power supply 110...230V, which do not require manual switching of the mains voltage.
Also, the use of active PFC improves the response of the power supply during short-term (fractions of a second) dips in the mains voltage - at such moments the unit operates using the energy of the high-voltage rectifier capacitors, the efficiency of which more than doubles. Another advantage of using an active PFC is the lower level of high frequency noise on the output lines, i.e. such power supplies are recommended for use in PCs with peripherals designed to work with analog audio/video material.

International organizations and PFC

The International Electrotechnical Commission (IEC) and the International Organization for Standardization (ISO) set limits on the content and levels of harmonics in the input current of secondary power supplies. The use of electrical appliances that do not meet the standards of these organizations is prohibited in many countries, so developers of serious equipment must remember this.

A little about power

Don't worry, you don't need a university degree in physics to understand how it works. We'll just explain what's different good block poor nutrition. If you know the basic principles of how it works, you are unlikely to make a bad purchase. So, let's move on.

Reactive current and reactive power

One of the important issues regarding power consumption when using switching power supplies is the "reactive" current caused by inductance. Please note that standby power consumption has nothing to do with idle mode. In addition, the load in this case does not overlap in any way with the power consumption at full load, but uses the same components. Reactive power must be significantly reduced (in best case scenario it should not exist at all) so that it does not lead to loss of energy in the resistance, which will be released in the form of heat. Such wasteful energy consumption must be reduced to almost zero by the internal circuits of switching power supplies.

Effective power and apparent power

Effective power is the opposite of reactive power in that it reflects actual power consumption. Apparent power is the sum of active and reactive power.

Power factor

This indicator is calculated as the ratio between effective power and apparent power and is between 0 (worst result) and 1 (ideal result). So, when buying a power supply, you need to make sure that it has a high power factor: this is one of the key quality indicators for power supplies.

Active PFC

Active Power Factor Correction (PFC) means active power factor correction. The power factor is important characteristic for the power supply, since it reflects the ratio between active and apparent power.

Advantages:

• An active power of about 99% can be considered ideal;
• High efficiency (less at low loads);
• Very stable power supply;
• Less energy consumption;
• Less heat generation;
• Less weight.

Flaws:

• Costs more;
• High probability of failure.

Passive PFC

Using passive power factor correction reactive currents can be reduced by using large inductors. This method is simpler and cheaper, but it is not the most effective.

Advantages:

• Costs less;
• Absence electromagnetic interference.

Flaws:

• Better cooling required;
• Not suitable for high loads;
• High energy consumption (energy loss);
• Heavier;
• Low active power (approximately 70% to 80%).

How to determine the efficiency of a power supply?

Basic principles, rules and regulations

One of the key performance indicators of a power supply is whether it meets Energy standards Star 5.0 and 80 PLUS. The latter will be a priority for computer technology and is a standard recognized throughout the world. In addition, if we are talking about European countries, then it is also necessary to check compliance with CE and ErP standards.

80 PLUS power supplies are more efficient.

The principles and specifications naturally influence the efficiency and quality of the food. A power supply marked with 80 PLUS certification will meet certain requirements, which are established through a set of tests. We would like to mention that the 80 PLUS stress testing conditions do not directly correspond to the ATX specification, however, they are performed under US conditions electrical networks power supplies operating at lower voltage. In the conditions of Russia and Europe, with 230 V networks, the efficiency of 80 PLUS power supplies will be slightly higher than in the USA.

The 80 PLUS concept has been expanded to include several performance levels, Platinum, Gold, Silver and Bronze, and each of these standards has its own set of specifications. Thus, an 80 PLUS Platinum or 80 PLUS Gold power supply will be more efficient than regular block nutrition. At the same time, these power supplies are more expensive.

Using the table below, you can see how a device's specification level affects its performance under a given load and evaluate each specific specification level.

Power consumption when the computer is turned off

When turning off the computer? The power supply usually continues to work. This is necessary to support some features like Wake-on-LAN. The power supply will waste some power even when the computer is turned off. Modern blocks power supplies, especially those sold in Europe, according to manufacturers, consume no more than 1 W in this mode. If saving is really important to you, then this decision will be right.

Linear and switching power supplies

Let's start with the basics. The power supply in a computer performs three functions. First, alternating current from the household power supply must be converted to direct current. The second task of the power supply is to reduce the voltage of 110-230 V, which is excessive for computer electronics, to the standard values ​​​​required by power converters of individual PC components - 12 V, 5 V and 3.3 V (as well as negative voltages, which we will talk about a little later) . Finally, the power supply plays the role of a voltage stabilizer.

There are two main types of power supplies that perform the above functions - linear and switching. The simplest linear power supply is based on a transformer, on which the voltage AC is reduced to the required value, and then the current is rectified by the diode bridge.

However, the power supply is also required to stabilize the output voltage, which is caused by both voltage instability in the household network and a voltage drop in response to an increase in current in the load.

To compensate for the voltage drop, in a linear power supply the transformer parameters are calculated to provide excess power. Then, at high current, the required voltage will be observed in the load. However, the increased voltage that will occur without any means of compensation at low current in the payload is also unacceptable. Excess voltage is eliminated by including a non-useful load in the circuit. In the simplest case, this is a resistor or transistor connected through a Zener diode. In a more advanced version, the transistor is controlled by a microcircuit with a comparator. Be that as it may, excess power is simply dissipated as heat, which negatively affects the efficiency of the device.

In the switching power supply circuit, one more variable appears, on which the output voltage depends, in addition to the two already existing: input voltage and load resistance. There is a switch in series with the load (which in the case we are interested in is a transistor), controlled by a microcontroller in pulse width modulation (PWM) mode. The higher the duration of the open states of the transistor in relation to their period (this parameter is called duty cycle, in Russian terminology the inverse value is used - duty cycle), the higher the output voltage. Due to the presence of a switch, a switching power supply is also called Switched-Mode Power Supply(SMPS).

No current flows through a closed transistor, and the resistance of an open transistor is ideally negligible. In reality, an open transistor has resistance and dissipates some of the power as heat. In addition, the transition between transistor states is not perfectly discrete. And yet, the efficiency of a pulsed current source can exceed 90%, while the efficiency of a linear power supply with a stabilizer reaches 50% at best.

Another advantage of switching power supplies is the radical reduction in the size and weight of the transformer compared to linear power supplies of the same power. It is known that the higher the frequency of alternating current in the primary winding of a transformer, the smaller the required core size and the number of winding turns. Therefore, the key transistor in the circuit is placed not after, but before the transformer and, in addition to voltage stabilization, is used to produce high-frequency alternating current (for computer power supplies this is from 30 to 100 kHz and higher, and as a rule - about 60 kHz). A transformer operating at a power supply frequency of 50-60 Hz would be tens of times more massive for the power required by a standard computer.

Linear power supplies today are used mainly in the case of low-power applications, where the relatively complex electronics required for a switching power supply constitute a more sensitive cost item compared to a transformer. These are, for example, 9 V power supplies, which are used for guitar effects pedals, and once for game consoles etc. But chargers for smartphones are already entirely pulsed - here the costs are justified. Due to the significantly lower amplitude of voltage ripple at the output, linear power supplies are also used in those areas where this quality is in demand.

⇡ General diagram of an ATX power supply

BP desktop computer is a switching power supply, the input of which is supplied with household voltage with parameters 110/230 V, 50-60 Hz, and the output has a number of lines DC, the main ones are rated 12, 5 and 3.3 V. In addition, the power supply provides a voltage of -12 V, and sometimes also a voltage of -5 V, required for the ISA bus. But the latter was at some point excluded from the ATX standard due to the end of support for the ISA itself.

In the simplified diagram of a standard switching power supply presented above, four main stages can be distinguished. In the same order, we consider the components of power supplies in the reviews, namely:

1. EMI filter - electromagnetic interference (RFI filter);
2. primary circuit - input rectifier (rectifier), key transistors (switcher), creating high-frequency alternating current on the primary winding of the transformer;
3. main transformer;
4. secondary circuit - current rectifiers from the secondary winding of the transformer (rectifiers), smoothing filters at the output (filtering).

⇡ EMF filter

The filter at the power supply input is used to suppress two types of electromagnetic interference: differential (differential-mode) - when the interference current flows in different directions in the power lines, and common-mode (common-mode) - when the current flows in one direction.

Differential noise is suppressed by capacitor CX (the large yellow film capacitor in the photo above) connected in parallel with the load. Sometimes a choke is additionally attached to each wire, which performs the same function (not on the diagram).

The common mode filter is formed by CY capacitors (blue drop-shaped ceramic capacitors in the photo), connecting the power lines to ground at a common point, etc. a common-mode choke (LF1 in the diagram), the current in the two windings of which flows in the same direction, which creates resistance for common-mode interference.

In cheap models they install minimum set filter parts, in more expensive ones the described circuits form repeating (in whole or in part) links. In the past, it was not uncommon to see power supplies without any EMI filter at all. Now this is rather a curious exception, although if you buy a very cheap power supply, you can still run into such a surprise. As a result, not only and not so much the computer itself will suffer, but other equipment connected to the household network - switching power supplies are a powerful source of interference.

In the filter area of ​​a good power supply, you can find several parts that protect the device itself or its owner from damage. There is almost always a simple fuse for short circuit protection (F1 in the diagram). Note that when the fuse trips, the protected object is no longer the power supply. If a short circuit occurs, it means that the key transistors have already broken through, and it is important to at least prevent the electrical wiring from catching fire. If a fuse in the power supply suddenly burns out, then replacing it with a new one is most likely pointless.

Separate protection is provided against short-term surges using a varistor (MOV - Metal Oxide Varistor). But there are no means of protection against prolonged voltage increases in computer power supplies. This function is performed by external stabilizers with their own transformer inside.

The capacitor in the PFC circuit after the rectifier may retain a significant charge after being disconnected from power. To prevent a careless person who sticks his finger into the power connector from receiving an electric shock, a high-value discharge resistor (bleeder resistor) is installed between the wires. In a more sophisticated version - together with a control circuit that prevents charge from leaking when the device is operating.

By the way, the presence of a filter in the PC power supply (and in the monitor power supply and almost any computer equipment it is also there) means that buying a separate " surge protector"instead of a regular extension cord, in general, to no avail. Everything is the same inside him. The only condition in any case is normal three-pin wiring with grounding. Otherwise, the CY capacitors connected to ground simply will not be able to perform their function.

⇡ Input rectifier

After the filter, the alternating current is converted into direct current using a diode bridge - usually in the form of an assembly in a common housing. A separate radiator for cooling the bridge is highly welcome. A bridge assembled from four discrete diodes is an attribute of cheap power supplies. You can also ask what current the bridge is designed for to determine whether it matches the power of the power supply itself. Although, as a rule, there is a good margin for this parameter.

⇡ Active PFC block

In an AC circuit with a linear load (such as an incandescent light bulb or an electric stove), the current flow follows the same sine wave as the voltage. But this is not the case with devices that have an input rectifier, such as switching power supplies. The power supply passes current in short pulses, approximately coinciding in time with the peaks of the voltage sine wave (that is, the maximum instantaneous voltage) when the smoothing capacitor of the rectifier is recharged.

The distorted current signal is decomposed into several harmonic oscillations in the sum of a sinusoid of a given amplitude (the ideal signal that would occur with a linear load).

The power used to perform useful work (which, in fact, is heating the PC components) is indicated in the characteristics of the power supply and is called active. The remaining power generated by harmonic oscillations of the current is called reactive. It does not produce useful work, but heats the wires and creates a load on transformers and other power equipment.

The vector sum of reactive and active power is called apparent power. And the ratio of active power to total power is called power factor - not to be confused with efficiency!

A switching power supply initially has a rather low power factor - about 0.7. For private consumers reactive power is not a problem (fortunately, it is not taken into account by electricity meters), unless he uses a UPS. It just fits the uninterruptible power supply full power loads. At the scale of an office or city network, excess reactive power created by switching power supplies already significantly reduces the quality of power supply and causes costs, so it is being actively combated.

In particular, the vast majority of computer power supplies are equipped with active power factor correction (Active PFC) circuits. A unit with an active PFC is easily identified by a single large capacitor and inductor installed after the rectifier. In essence, Active PFC is another pulse converter that maintains a constant charge on the capacitor with a voltage of about 400 V. In this case, current from the supply network is consumed in short pulses, the width of which is selected so that the signal is approximated by a sine wave - which is required to simulate a linear load . To synchronize the current consumption signal with the voltage sinusoid, the PFC controller has special logic.

The active PFC circuit contains one or two key transistors and a powerful diode, which are placed on the same heatsink with the key transistors of the main power supply converter. As a rule, the PWM controller of the main converter key and the Active PFC key are one chip (PWM/PFC Combo).

The power factor of switching power supplies with active PFC reaches 0.95 and higher. In addition, they have one additional advantage - they do not require a 110/230 V mains switch and a corresponding voltage doubler inside the power supply. Most PFC circuits handle voltages from 85 to 265 V. In addition, the sensitivity of the power supply to short-term voltage dips is reduced.

By the way, in addition to active PFC correction, there is also a passive one, which involves installing a high-inductance inductor in series with the load. Its efficiency is low, and you are unlikely to find this in a modern power supply.

⇡ Main converter

The general principle of operation for all pulse power supplies of an isolated topology (with a transformer) is the same: a key transistor (or transistors) creates alternating current on the primary winding of the transformer, and the PWM controller controls the duty cycle of their switching. Specific circuits, however, differ both in the number of key transistors and other elements, and in quality characteristics: Efficiency, signal shape, noise, etc. But here too much depends on the specific implementation for this to be worth focusing on. For those interested, we provide a set of diagrams and a table that will allow you to identify them in specific devices based on the composition of the parts.

In addition to the listed topologies, in expensive power supplies there are resonant versions of Half Bridge, which are easily identified by an additional large inductor (or two) and a capacitor forming an oscillatory circuit.

⇡ Secondary circuit

The secondary circuit is everything that comes after the secondary winding of the transformer. In most modern power supplies, the transformer has two windings: 12 V is removed from one of them, and 5 V from the other. The current is first rectified using an assembly of two Schottky diodes - one or more per bus (on the highest loaded bus - 12 V - in powerful power supplies there are four assemblies). More efficient in terms of efficiency are synchronous rectifiers, which use field-effect transistors instead of diodes. But this is the prerogative of truly advanced and expensive power supplies that claim the 80 PLUS Platinum certificate.

The 3.3V rail is typically driven from the same winding as the 5V rail, only the voltage is stepped down using a saturable inductor (Mag Amp). A special winding on a transformer for a voltage of 3.3 V is an exotic option. Of the negative voltages in the current ATX standard, only -12 V remains, which is removed from the secondary winding under the 12 V bus through separate low-current diodes.

PWM control of the converter key changes the voltage on the primary winding of the transformer, and therefore on all secondary windings at once. At the same time, the computer's current consumption is by no means evenly distributed between the power supply buses. In modern hardware, the most loaded bus is 12-V.

To separately stabilize voltages on different buses, additional measures are required. The classic method involves using a group stabilization choke. Three main buses are passed through its windings, and as a result, if the current increases on one bus, the voltage drops on the others. Let's say the current on the 12 V bus has increased, and in order to prevent a voltage drop, the PWM controller has reduced the duty cycle of the key transistors. As a result, the voltage on the 5 V bus could go beyond the permissible limits, but was suppressed by the group stabilization choke.

The voltage on the 3.3 V bus is additionally regulated by another saturable inductor.

A more advanced version provides separate stabilization of the 5 and 12 V buses due to saturable chokes, but now this design has given way to DC-DC converters in expensive high-quality power supplies. In the latter case, the transformer has a single secondary winding with a voltage of 12 V, and the voltages of 5 V and 3.3 V are obtained thanks to DC-DC converters. This method is most favorable for voltage stability.

Output filter

The final stage on each bus is a filter that smoothes out voltage ripple caused by the key transistors. In addition, the pulsations of the input rectifier, whose frequency is equal to twice the frequency of the power supply network, penetrate to one degree or another into the secondary circuit of the power supply.

The ripple filter includes a choke and large capacitors. High-quality power supplies are characterized by a capacitance of at least 2,000 uF, but manufacturers of cheap models have reserves for savings when they install capacitors, for example, of half the nominal value, which inevitably affects the ripple amplitude.

⇡ Standby power +5VSB

A description of the power supply components would be incomplete without mentioning the 5 V standby voltage source, which makes the PC sleep mode possible and ensures the operation of all devices that must be turned on at all times. The “duty room” is powered by a separate pulse converter with a low-power transformer. In some power supplies there is also a third transformer used in the circuit feedback to isolate the PWM controller from the primary circuit of the main converter. In other cases, this function is performed by optocouplers (an LED and a phototransistor in one package).

⇡ Methodology for testing power supplies

One of the main parameters of the power supply is voltage stability, which is reflected in the so-called. cross-load characteristic. KNH is a diagram in which the current or power on the 12 V bus is plotted on one axis, and the total current or power on the 3.3 and 5 V buses is plotted on the other. At the intersection points at different meanings Both variables determine the voltage deviation from the nominal value on a particular bus. Accordingly, we publish two different KNHs - for the 12 V bus and for the 5/3.3 V bus.

The color of the dot indicates the percentage of deviation:

• green: ≤ 1%;
• light green: ≤ 2%;
• yellow: ≤ 3%;
• orange: ≤ 4%;
• red: ≤ 5%.
• white: > 5% (not allowed by ATX standard).

To obtain KNH, a custom-made power supply test bench is used, which creates a load by dissipating heat on powerful field-effect transistors.

Another equally important test is determining the ripple amplitude at the power supply output. The ATX standard allows ripple within 120 mV for a 12 V bus and 50 mV for a 5 V bus. A distinction is made between high-frequency ripple (at double the frequency of the main converter switch) and low-frequency (at double the frequency of the supply network).

We measure this parameter using a Hantek DSO-6022BE USB oscilloscope at maximum load on a power supply specified by specifications. In the oscillogram below, the green graph corresponds to the 12 V bus, the yellow graph corresponds to 5 V. It can be seen that the ripples are within normal limits, and even with a margin.

For comparison, we present a picture of ripples at the output of the power supply of an old computer. This block wasn't great to begin with, but it certainly hasn't improved over time. Judging by the magnitude of the low-frequency ripple (note that the voltage sweep division is increased to 50 mV to fit the oscillations on the screen), the smoothing capacitor at the input has already become unusable. High-frequency ripple on the 5 V bus is on the verge of permissible 50 mV.

The following test determines the efficiency of the unit at a load from 10 to 100% of rated power (by comparing the output power with the input power measured using a household wattmeter). For comparison, the graph shows the criteria for the various 80 PLUS categories. However, this does not cause much interest these days. The graph shows the results of the top-end Corsair PSU in comparison with the very cheap Antec, and the difference is not that great.

A more pressing issue for the user is the noise from the built-in fan. It is impossible to directly measure it close to the roaring power supply testing stand, so we measure the rotation speed of the impeller with a laser tachometer - also at power from 10 to 100%. The graph below shows that when the load on this power supply is low, the 135mm fan remains at low speed and is hardly audible at all. At maximum load the noise can already be discerned, but the level is still quite acceptable.

I’ll say right away that the article is intended for a simple PC user, although it was possible to go deeper into academic details.
Despite the fact that the diagrams are not mine, I give a description exclusively “on my own”, which does not pretend to be the only correct one, but aims to explain “at a glance” the work of such required device, like a computer power supply.

I felt the need to understand how APFC works in 2005, when I had a problem with my computer randomly rebooting. I bought the computer from a “soap” company without delving into the details. The service did not help: it works for the company, but it reboots for me. I realized that it was my turn to tense up... It turned out that the problem was home network, which in the evening dropped abruptly to 160V! I started looking for a circuit, increasing the capacitance of the input capacitors, it worked a little, but it didn’t solve the problem. While searching for information, I saw in the price lists the strange letters APFC and PPFC in the names of the blocks. Later I found out that I had PPFC and I decided to buy myself a unit with APFC, then I also bought an uninterruptible power supply. Other problems began - the uninterruptible power supply crashes when the system unit is turned on and the network disappears, the service team shrugs. I returned it, bought it 3 times more powerful, and it still works to this day without problems.

I will share my experience with you and hope you will be interested in learning a little more about the system component - the power supply unit, which is unfairly assigned almost the last role in the operation of a computer.

FSP Epsilon 1010 power supplies are high-quality and reliable devices, but given the problems of our networks and other accidents, they sometimes also fail. It would be a shame to throw away such a unit, and repairs could approach the cost of a new one. But there are also little things that, by eliminating them, can bring him back to life.

What FSP Epsilon 1010 looks like:

The most important thing is to understand the principle of operation and break down the block into pieces.

I will give an example of fragments of circuit diagrams of a standard FSP Epsilon unit that I dug up on the net. The diagrams were compiled manually by a very diligent and competent person, who kindly posted them for public access:

1. Basic diagram:
Figure 1:
Link to full size: s54.radikal.ru/i144/1208/d8/cbca90320cd9.gif

2. APFC controller circuit:
Figure 2:
Link to full size: i082.radikal.ru/1208/88/0f01a4c58bfc.gif

Modifications of power supplies of this series differ in the number of elements (additionally soldered into the same board), but the operating principle is the same.

So what is APFC?

PFC- this is power factor correction (PFC) - the process of bringing the consumption of an end device that has a low power factor when powered from an AC power network to a state in which the power factor complies with accepted standards. If you show it on three fingers, it looks like this:

We started the power supply, the capacitors began to charge - there was a peak in current consumption coinciding with the peak of the sine wave AC 220V 50Hz (too lazy to draw). Why matching? How will they charge at “0” volts closer to the time axis? No way! There will be peaks in each half-wave of the sine wave, since there is a diode bridge in front of the capacitor.
- the load of the unit drew current and discharged the capacitors;
- the capacitors began to charge and peaks of current consumption appeared again at the peaks of the sine wave.

And we see a “hedgehog” with which the sinusoid has grown, and which, instead of constant consumption, “pulls” the current in short jumps at narrow moments of time. What’s so scary about that, don’t let it bother you, you say. And this is where the Hound of the Baskervilles rummaged: these peaks overload electrical wiring and can even lead to a fire with the nominally calculated cross-section of the wires. What if we consider that there is more than one block in the network? Yes, and working on the same network electronic devices You’re unlikely to like such a “sawed-down” network with interference. Moreover, with the stated rated power of the power supply, you will pay more for light, since your network wires in the apartment (office) are already the load. The task arises of reducing the peaks of current consumption over time towards the dips of the sinusoid, that is, getting closer to a semblance of linearity and relieving the wiring.

PPFC - passive correction power factor. This means that in front of one power supply wire there is a massive inductor, the task of which is to reduce the peaks of current consumption during the charging of capacitors, taking into account the nonlinear properties of the inductor (that is, the fact that the current through it lags behind the voltage applied to it - remember school). It looks like this: at the maximum of the sine wave, the capacitor should be charged and it is waiting for this, but bad luck - they put a choke in front of it. But the inductor is not entirely concerned with what the capacitor needs - a voltage is applied to it and a self-induction current arises, which is directed in the opposite direction. Thus, the inductor prevents the capacitor from charging at the peak of the input sinusoid - there is a peak in the network, and the capacitor is discharged. Strange, isn't it? Isn't this what we wanted? Now the sinusoid drops, but the inductor behaves like most people: (we have it - we don’t value it, we lose it - we regret it) again a self-induction current arises only now coinciding with the decreasing current, which charges the capacitor. What we have: at the peak - nothing, at the dips - charge! Mission accomplished!
This is exactly how the PPFC circuit works by pulling peaks of current consumption onto the dips of the sine wave (rising and falling sections) using just one inductor. The power factor is close to 0.6. Not bad, but not perfect.

APFC- active power factor correction. This means using electronic components that require power. This power supply actually has two power supplies: the first is a 410V stabilizer, the second is a regular classic switching power supply. We will look at this below.

APFC and operating principle.

Figure 3:

We have just approached the principle of operation of active power factor correction, so we will immediately determine some points for ourselves. In addition to its main purpose (approaching the linearity of current consumption over time), APFC solves a triune problem and has the following features:

The power supply with APFC consists of two blocks: the first is a 410V stabilizer (APFC itself), the second is a regular classic switching power supply.
- APFC circuit provides a power factor of about 0.9. This is what we strive for - to "1".
- APFC circuit operates at a frequency of about 200KHz. Agree, pulling the current 200,000 times per second in relation to 50 Hz is practically at every moment of time, that is, linear.
- the APFC circuit provides a stable DC output voltage of about 410V and operates from 110 to 250V (in practice from 40V). This means that the industrial network has virtually no effect on the operation of internal stabilizers.

Circuit operation:

The operating principle of APFC is based on the accumulation of energy in the inductor and its subsequent release to the load.
When power is supplied through the inductor, its current lags behind the voltage. When the voltage is removed, the phenomenon of self-induction occurs. This is what the power supply eats up, and since the self-induction voltage can be close to twice the applied one - here you have work from 110V! The task of the APFC circuit is to dose the current through the inductor with a given accuracy so that the output always has a voltage of 410V, regardless of the load and input voltage.

In Figure 3 we see DC - source DC voltage after the bridge (not stabilized), storage choke L1, transistor switch SW1, which is controlled by a comparator and PWM. The circuit is made quite boldly at first glance, since the key actually short-circuits the socket at the moment of opening, but we will forgive it, given that the short circuit occurs for microseconds with a frequency of 200,000 times per second. But if there is a malfunction in the key control circuit, you will definitely hear and even smell, and maybe even see how the power keys in such a circuit burn out.

1. Transistor SW1 is open, current flows into the load as before through the inductor from “+ DC” - “L1” - “SW2” - “RL” to “-DC”. But the inductor resists the movement of current (self-induction begins), while energy accumulates in the inductor L1 - the voltage on it increases almost to the DC voltage, since this is a short circuit (albeit for a fraction of the time (as long as everything is fine). Diode SW2 prevents the discharge of capacitor C1 at the moment the transistor opens.
2. Transistor SW1 has closed... the voltage at the load will be equal to the sum of the voltages of the source DC1 and the inductor L1, which has just been slightly applied to the source and released a self-induction current with reverse polarity. The magnetic field of the inductor, disappearing, will cross it, inducing a self-inductive emf of the opposite polarity on it. Now the self-induction current has the same direction as the disappearing source current (self-induction end). Self-induction is the phenomenon of the occurrence of induced emf in an electrical circuit as a result of a change in current strength.
So, at the moment of self-induction after the transistor is closed, we get our addition to 410V due to the addition of energy from the inductor. Why the supplement? Remember back to school, how much will be the output of a bridge with a capacitor if the input is 220V? That's right, 220V multiplied by the root of two (1.41421356) = 311V. This is what would happen without the APFC circuit working. It is so at the point where we are waiting for 410V, while only the +5V control room is working and the unit itself is not running. Now there is no point in driving APFC, the duty room will have enough of its 2 Amps.
All this is strictly controlled by the control circuit using feedback from the 410V point. The level of self-induction is regulated by the opening time of the transistors, that is, by the time of energy accumulation L1 - this is pulse-width stabilization. The task of APFC is to stably maintain 410V output when external network factors and load change.

So it turns out that in a power supply with APFC there are two power supplies: a 410V stabilizer and the classic power supply itself.

The reduction in the dependence of current consumption peaks on sinusoid peaks is ensured by transferring these peaks to the operating frequency of the APFC circuit - 200,000 times per second, which approaches the linear current consumption at each moment of time of a 50Hz 220V sinusoid. Q.E.D.

Advantages of APFC:
- power factor about 0.9;
- work from any capricious network 110 - 250V, including unstable rural ones;
- noise immunity:
- high coefficient of stabilization of output voltages due to a stable input 410V;
- low ripple factor of output voltages;
- small filter sizes, since the frequency is about 200 KHz.
- high overall efficiency of the unit.
- small interference transmitted to industrial network;
- high economic effect in payment for light;
- electrical wiring is unloaded;
- in enterprises and telecommunications organizations that have 60V station batteries, you can do without a UPS at all to power critical servers - just plug the unit into the guaranteed 60V power supply circuit without changing anything or observing polarity (which does not exist). This will allow you to get away from those miserable 15 minutes of work from UPS to 10 hours from station batteries, so that the entire control system does not fall down in the event of a diesel failure. But many people don’t pay attention to this or haven’t thought about it until the diesel gets offended somehow... All the equipment will continue to work, but there will be nothing to control, since the computers will turn off after 15 minutes. The manufacturer presented an operating range of 90 - 265V due to the lack of such a power supply standard as 60V variable, but the practical operating limit was obtained at a value of 40V, there was no point in checking below.
Re-read the paragraph carefully again and evaluate the capabilities of your UPS for critical servers!

Disadvantages of APFC:
- price;
- difficulty in diagnosis and repair;
- expensive parts (transistors - about $5 per piece, and sometimes there are up to 5 pieces), often the cost of repairs is not justified;
- problems collaboration with uninterruptible power supplies (UPS) due to the high starting current. You need to choose a UPS with a double power reserve.

Now let's look at the FSP Epsilon 1010 power supply circuit in Fig. 1, 2.

In the FSP Epsilon 1010, the power part of the APFC is represented by three HGTG20N60C3 transistors with a current of 45A and a voltage of 600V, standing in parallel: www.fairchildsemi.com/ds/HG/HGT1S20N60C3S.pdf
In our standard diagram there are 2 Q10, Q11, but this does not change the essence. Our block is simply more powerful. The FPC OUT signal goes from leg 12 of the CM6800G chip to pin 12 of the control module in Fig. 2. Next, through resistor R8 to the gates of the keys. This is how APFC is controlled. The APFC control circuit is powered from +15V control room via optocoupler M5, resistor R82 - 8pin CB (A). But it starts only after the unit is turned on to the load via the PW-ON signal (green wire of the 24-pin connector to ground).

Typical faults:

Symptoms:
- the fuse blows with a bang;
- the unit “does not breathe” at all even after replacing the fuse, which is even worse. This means the damage threatens to result in more expensive repairs.

Diagnosis: APFC circuit failure.

Treatment:
It is difficult to make a mistake when diagnosing an APFC circuit failure.
It is generally accepted that a unit with APFC can be started without APFC if it fails. And we will think so, and even check it, especially when it comes to dangerous experiments with expensive HGT1S20N60C3S transistors. We solder the transistors.
The unit works successfully if the problem was only in the APFC circuit, but you need to understand that the power supply will lose power up to 30% and cannot be put into operation - only a test. Well, then we replace the transistors with new ones, but we turn on the unit in series through a 220V 100W incandescent lamp. For example, we load the block onto an old HDD. If the lamp is lit and the HDD has started (we touch it with our fingers), the fan on the unit is spinning - there is a possibility that the repair is complete. We start without a lamp with a fuse size reduced by 3 times. And now it’s not burned out? Well then, solder in the original F1 and go ahead for an hour-long test under a 300-500 watt equivalent load! A lamp burning at full intensity tells you about full opening key transistors or their dead state, we are looking for a problem in front of them.
If at any stage we are unlucky, we return to new purchase transistors, not forgetting to buy the CM6800G controller. We change the details, repeat everything again. Don't forget to visually inspect the entire board!

Symptoms:
- the unit starts up every other time or when it remains plugged in for 5 minutes;
- you have a faulty HDD out of nowhere;
- the fans are spinning, but the system does not boot, the BIOS does not beep at startup;
- condensers are swollen motherboard, video card;
- the system randomly reboots and freezes.

Diagnosis: Electrolytic capacitors have dried out.

Treatment:
- disassemble the unit and visually find swollen capacitors;
- best solution change everything to new ones, not just swollen ones;

Failure to start occurs due to dry duty capacitors C43, C44, C45, C49;
Component failures occur due to increased ripple in the +5V, +12V circuit due to drying of filter condensates.

Symptoms:
- the unit whistles or beeps;
- the tone of the whistle changes under load;
- the unit whistles only while it is cold or while it is hot.

Diagnosis: Cracks in the printed circuit board or missing parts.

Treatment:
- disassemble the block;
- visually inspect printed circuit board in places where key transistors and filter chokes are soldered, look for oval cracks at the soldering site;
- if we don’t find anything, then we still solder the legs of the power elements.
- we check and enjoy the silence.

There are a great many other faults, including internal breaks or interturn breakdowns, cracks in the board and parts, and so on. Temperature malfunctions are especially annoying when it works until it warms up or cools down.
Power supplies from other manufacturers have a similar operating principle, which will allow you to find and fix the problem.

Finally, a couple of tips on power supply:
1.Never unplug a power supply with APFC while it is running! First, park the system, and then unplug it from the socket or turn it off without the extension cord - otherwise you’ll end up playing…
If the voltage fails while the unit is operating, an arc stretches and sparking occurs, which leads to a bunch of harmonics other than 50Hz - this time, the voltage decreases and the APFC switches try to maintain a stable voltage at the output, while opening completely and longer time, causing even more current and arc - that's two. This leads to breakdown of open transistors with huge currents and uncontrolled harmonic voltages - that's three. It's easy to check if you want. Personally, I already checked... now I wrote this article and spent $25 on repairs. You can also write your own. By the way, on the FSP Epsilon 1010, the button on the case turns off not the power cord, but the control system, while all power elements remain energized - be careful! Therefore, if you urgently need to turn off the computer, then do it with the power button on the unit - everything is thought out here.

2. If you know in advance that you will work with an uninterruptible power supply, then buy a power supply with PPFC. This will save you from unnecessary problems.

In the story, I tried not to include unnecessary graphs, diagrams, formulas and technical terms, so as not to scare off the average tormentor of his PC on the fifth line, whose deeper understanding of the basics of nutrition will extend his time trouble-free operation.

Now is the time to disassemble the system unit and determine the model of your power supply, and at the same time shake out the dust from it. You have already prevented one malfunction. If it is clean, it will gratefully last longer. Lubricate the fan, this is also welcome.

Those who read the article to the end - thank you all!
Now your PSU is safe.