PSUs 101: A Detailed Look Into Power Supplies

PSU Protections

In this section, we look at the various protections that a PSU has in order to avoid harm not only to the supply, but also to the system that it feeds with power. Many budget PSUs only have the necessary protections that the ATX specification demands (OCP, SCP, OVP), while higher-end units usually have much more protection.

Power Good Or PWR_OK Signal

As the ATX specification states, the PSU uses a power good or PWR_OK delay signal to indicate that the +5V, +3.3 V and +12V outputs are within the regulation thresholds of the power supply and that sufficient mains energy is stored by the converter to guarantee continuous operation within specification, for at least 17ms under full load (16ms for the AC loss to PWR_OK hold-up time). The PWR_OK delay period according to the ATX spec must be below 500ms and ideally less than 250ms. In any case, it should be equal to or above 100ms.

(OCP) Over-Current Protection

Over-current protection (OCP) is a popular protection found in all PSUs with multiple +12V rails, and in most cases, it also protects the minor rails. OCP kicks in when the current in the rails surpasses a certain limit. The ATX 2.2 specification states that if the load at each tested output rail reaches or exceeds 240VA, then OCP should interfere (paragraph 3.4.4). However, the ATX 2.31 specification omits this limit. In order to bypass it, some manufacturers implemented many virtual +12V rails, with each rail rated at 240VA. However, in most cases, the OCP trigger point was set much higher in order to withstand peak currents that some system components (like graphics cards) could draw.

To implement OCP in a PSU, two things are necessary: shunt resistors and a supervisor IC that supports OCP. The shunt resistors are low-resistance, high-precision resistors used to measure the current at the outputs of a PSU, utilizing the voltage drops those currents create across the resistors. By measuring the number of shunts in a PSU in the area where +12V wires are soldered, we are usually able to find the real number of +12V virtual rails. In some cases, when the manufacturer has initially built the PSU as a multiple +12V rail unit and afterward converted it to a single +12V rail unit, the shunt resistors are simply shorted together.

Shunt resistors used in the Corsair AX1200iShunt resistors used in the Corsair AX1200i

OVP/UVP (Over Voltage/Under Voltage Protection)

The ATX specification states that the over-voltage protection sense circuitry and reference should reside in packages that are separate and distinct from the regulator control circuitry and reference. So, no single point of fault should be able to cause a sustained over-voltage condition on any output. In other words, all PSUs must have an independent protections circuit and not count solely on the PWM controller to monitor the output voltages. We must also add that UVP is optional since it's not mentioned in ATX specification.

As you might have already guessed, OVP and UVP constantly check the voltages at each rail and kick in when these voltages surpass or go below a trigger point. The ATX spec provides a table with the minimum, nominal and maximum values for the OVP trigger points. The spec includes the 5VSB rail, although it states that OVP protection on this rail is recommended but not required. Below, you will find the corresponding table. 

OutputMinimum (V)Nominal (V)Maximum (V)
+12 VDC
(or 12V1DC & 12V2DC)
13.41515.6
+5 VDC5.746.37
+3.3 VDC3.764.24.3
5VSB (optional)5.746.37

As you can see, the trigger points are way too high. A manufacturer can set OVP to 15.6 for the +12V rails and still be within specification. Imagine 15.6V running through your system components!

Since UVP trigger points are not covered by the ATX specification, all manufacturers of IC protection circuits are free to set their own.

OPP (Over Power Protection)

The over-power protection (OPP) kicks in when the power we pull from a PSU exceeds its maximum rated capacity. Usually, the manufacturers give a little room for overpowering the PSU, so the OPP threshold is set to 50 to 100W (in some cases even more) above the maximum rated wattage of the PSU. In single +12V rail PSUs, where OCP is meaningless in most cases, OPP takes over its role and shuts down the PSU in case the +12V rail is overloaded.

OTP (Over Temperature Protection)

When the over-temperature protection (OTP) is present, we usually find a thermistor attached to the secondary heat sink (the fan control unit usually uses a thermistor in the same heat sink). The thermistor informs the protections circuit about the temperature of the heat sink, and if this exceeds a specified threshold, then the PSU shuts down. An excessive temperature may be a result of overloading or the cooling fan's failure, so OTP prevents (further) damage to the PSU. 

In some cases and because OTP isn't supported by most currently available supervisor ICs, it might be implemented through another method (for instance, by triggering another protection once excess temperature levels are detected in the PSU's internals). We consider OTP to be one of the most crucial protections in any PSU, though many models lack it.

SCP (Short Circuit Protection)

Short circuit protection (SCP) constantly monitors the output rails, and if it finds an impedance of less than 0.1Ω, it immediately shuts down the power supply. In other words, if somehow the output rails are short circuited, then this protection kicks in and shuts down the PSU to prevent damage or fire. According to the ATX 2.31 spec, each +12V rail should have a separate short circuit. This protection is present in almost all contemporary PSUs (at least the branded ones).

Create a new thread in the US Reviews comments forum about this subject
This thread is closed for comments
42 comments
Comment from the forums
    Your comment
  • Alexis Shaw
    In your list of top-tier capacitor manufacturers you missed out on some of the better american and european manufacturers, while these may not be used on many consumer-grade power supplies they are definitely top-tier and if you were to find them you would be happy. I suggest the addition of at least:
    Cornell Dubilier (USA)
    Illinois Capacitor (Now owned my Cornell Dubilier)
    Kemet Corporation (USA)
    ELNA (Japan)
    EPCOS (TDK company) (Germany)
    Vishay (USA)
    Würth Elektronik (Germany)
  • Aris_Mp
    Thank you very much for the list you provided. I am aware of almost all cap brands that you mentioned but unfortunately so far I found none of them inside a desktop/consumer grade PSU. I will think about it however (and also make a research on these cap brands), if I should include them as well inside my list.
  • InvalidError
    Anonymous said:
    Thank you very much for the list you provided. I am aware of almost all cap brands that you mentioned but unfortunately so far I found none of them inside a desktop/consumer grade PSU.

    There is a very high probability you have seen PSUs with several Kemet capacitors in them. You never noticed them simply because SMD capacitors are too small to carry logos, brand name or even value designations.

    The other brands are mostly found in specialty applications such as lab instruments, industrial machines and high-end audio.
  • Math Geek
    very interesting read. more in depth than i need to know yet for the most part understandable and with careful reading it did not leave me confused.

    nice article.
  • TallestJon96
    I only read 2/3 of it, but it's a good article.

    I basically have committed PC heresy with my cx600m. However I think that I'm in the clear with my 65w CPU and 145w CPU. I'd bet my total power draw is actually below 300w, the supposed highest efficiency point of a PSU.

    As a gamer, not a professional, I think it is better to get low power parts, and get a higher rating than you need, rather than get high power parts and high quality PSUs.

    Additionally, if you compare power consumption of a typical system from today to one from 5 years ago, power draw is considerably lower, with the exception of certain graphics cards. *cough* 390x *cough*
  • powernod
    I decided to sign up at Tom's forum, and the only reason was to state how excellent is Aris's article!!!
    Thanks Aris for this very useful article on behalf of us all who want to learn the basic knowledge for PSUs.
    Haven't finished it yet, but i'm very anxious for it !!!
  • GoZFast
    Very nice article!!! You made me remember my college physics courses lol
  • traumadisaster
    I'm glad there are people dedicated to this but I'm not. I can't even read all of the chapter titles in this article. I disagree with the importance you place on this and all of the references you made to this being crucial knowledge.

    PSU and MB are insignificant to me and I can blindly pick one by reviewing user comments from newegg in about 5 min, and it will last for years. For less than $100 each I'm set for nearly a decade.

    CPU and gfx card now that affects fps and is over $1000, actually the most important part to me.
  • Alexis Shaw
    Anonymous said:
    I'm glad there are people dedicated to this but I'm not. I can't even read all of the chapter titles in this article. I disagree with the importance you place on this and all of the references you made to this being crucial knowledge.

    PSU and MB are insignificant to me and I can blindly pick one by reviewing user comments from newegg in about 5 min, and it will last for years. For less than $100 each I'm set for nearly a decade.

    CPU and gfx card now that affects fps and is over $1000, actually the most important part to me.


    I heartily dissagree, user are not the best way to judge reliability, and a bad powersupply is at fult most of the time there is a hardware issue. Further a power supply should last more than one system build, and in general I keep mine for a decade at a time at least. So an investment in a good power supply is not a waste, and a bad one will kill that precious $1000 GPU or CPU. The demo dart power supply on the motherboard is a similar story, however in general they are of higher quality than a cheap mains supply.
  • Alexis Shaw
    Anonymous said:
    Anonymous said:
    Thank you very much for the list you provided. I am aware of almost all cap brands that you mentioned but unfortunately so far I found none of them inside a desktop/consumer grade PSU.

    There is a very high probability you have seen PSUs with several Kemet capacitors in them. You never noticed them simply because SMD capacitors are too small to carry logos, brand name or even value designations.

    The other brands are mostly found in specialty applications such as lab instruments, industrial machines and high-end audio.


    As well as SMT ceramic capacitors, Kemet makes through hole aluminium electrolytic capacitors. These are of high quality, though not as well known as their SMT capacitors. They also make high quality polymer SMT capacitors that are used as bulk capacitors on the power distribution circuitry on laptops and other devices.
  • ujaansona
    A concise and informative article, written with remarkable effort. As an instrumentation engineer I found it enchanting. Thank you Mr. Aris.
  • Brian Blair
    That's a monster of a PSU. But very very overpriced! I would much rather just buy a Seasonic at that price!
  • nukemaster
    Great writeup.
  • Aris_Mp
    Quote:
    In your list of top-tier capacitor manufacturers you missed out on some of the better american and european manufacturers, while these may not be used on many consumer-grade power supplies they are definitely top-tier and if you were to find them you would be happy. I suggest the addition of at least:
    Cornell Dubilier (USA)
    Illinois Capacitor (Now owned my Cornell Dubilier)
    Kemet Corporation (USA)
    ELNA (Japan)
    EPCOS (TDK company) (Germany)
    Vishay (USA)
    Würth Elektronik (Germany)


    Decide to include them in the article for reference and to show also that besides Japanese caps they are also some good US and German brands. Thanks again for your input!
  • hawkwindeb
    I was told there be no math questions...
    .
    LOL
    .
    Seriously though, very complete article
  • dstarr3
    Quote:
    I'm glad there are people dedicated to this but I'm not. I can't even read all of the chapter titles in this article. I disagree with the importance you place on this and all of the references you made to this being crucial knowledge.

    PSU and MB are insignificant to me and I can blindly pick one by reviewing user comments from newegg in about 5 min, and it will last for years. For less than $100 each I'm set for nearly a decade.

    CPU and gfx card now that affects fps and is over $1000, actually the most important part to me.



    You probably buy cheap hard drives, too.

    Enjoy your ticking time bomb. Because a cheap PSU and cheap motherboard is the quickest and surest way to a total system failure.
  • Quaddro
    Great article !!
    Toms should write more like this..
  • mctylr
    Quote:
    One or more bridge rectifiers fully correct the AC power stream after it passes the EMI/transient filter.


    You mean rectifies. A bridge rectifier converts the AC voltage into DC voltage.
    The name bridge, comes from being a bridge of four (power) diodes. ( image: Source: Play Hookey)
  • IWantDatHammer
    Finally a PSU article! :P
  • Aris_Mp
    I had it rectify and not "correct" but the proof editors changed it. Sorry will fix it ASAP. Of course they rectify and this is why they are called bridge rectifiers after all.