Watts dont mean Jack

Status
Not open for further replies.
I

Ice Czar

Guest
Under Construction Hard Hat Area


Over the past few years a situation has developed that many are unaware of
while we have all enjoyed the benefits of ever increasing performance out of our computers,
several notable changes have crept into how those are powered.

Today many have experienced intermittent stability problems, and or outright damage to components because of poor or inadequate power. Power supply selection is rapidly becoming as important as selecting compatible RAM, and equally complex.
Basically what has occured is that while older PSUs (Power Supply Unit) output voltage on the +3.3 Volt and +5 Volt rails to power the Motherboard, Graphics Card and CPU, with the +12 Volt rail for the Hard Drives, Optical Drives and Fans. That has now changed.

Modern boards heavilly rely on the +12V rail to power more powerful processors and graphics cards, and with the introduction of the PCI-Express bus, this will only get worse.
In short an older PSU of yesteryear is likely woefully inadequate on the +12V rail irregardless of its overall watts rating, since a large portion of the power its producing simply isnt need on those rails anymore, but on the +12V rail instead.

This article is going to address the pitfalls and tradeoffs you need to be aware of when selecting a PSU so that you are able to make an informed decision as to what level of performance you need, so lets start with the basics.

Power Delivery is a chain, and your PSU is but a link in that chain
if you make it a strong link it can actually makeup for some on the shortcoming of the other links, unlike a real chain. Power quality from a utility varies greatly over the world, from utility to utility and even with the seasons, and while there really is no substitute for power conditioning or a UPS (Uninteruptable Power supply) a good PSU is able to effectively deal with a larger range of AC power than a cheaper PSU, and when the range its able to deal with is exceeded, safely shutdown without damaging components, something cheap generic supplies are infamous for not doing.

1.Source Power Brown outs, blackouts, spikes\surges ect.
In this category I would also place power issues due to pilot error, hard restarts and shorts, Shutdown properly and pay attention when mounting your motherboard and routing power cables.

2. Under Power: Basically too many components for the power supply,
dont be decieved by wattage figures, its the amount of amps per rail that is really important.

3. Voltage Stability how clean the power is



one of the problems right now is there are alot of PSUs out there built to the ATX12V v1.3 spec which calls for the dual rails, but they arent documented by the retailers or manufacturers

often youll find the appearantly very same supply with a different wiring harness
as an EPS12V (which has at least three rails) or ATX12V v2.0 and they might have the +12v rails documented

the same goes for the main if you veiw the connector on the mobo with the clasp on the right you would need roon to hang the extra 4 pins off the top

there are 24 to 20 pin cconverters readilly available
but there arent 8 to 4 pin coverters (so EPS12V to ATX12V v2.0 or v1.3 is problematic as well as the possibility that you wont have access to the power off the 3rd +12V rail, and certainly not if its a quad rail)

in addition there will be a new 6 pin video connector soon
and of course SATA

for a short course in the wiring harness difference
ATX12V v1.3....20 pin main 4 pin aux
ATX12V 2.0.....24 pin main 4 pin aux
EPS12V.........24 pin main 8 pin aux

the pins are backward compatible if there is clearance on the mobo so you can connect an 8 pin auxillary to a 4 pin header if there is room above the connector for basically double the pins when viewing the mobo header with the clasp on the right

in short there are supplies that meet the latest specs but are described by the wiring harness they have,
and they are mixed up in the retail marketplace ATX pot with ATX12V 1.1 supplies that dont have dual rails

confusing aint it? :p

add up the rails with this calculator
http://takaman.jp/D/index.html?english
and compare it to the specs listed on the PSU
then you build in a safety margin of from 1\2 to 1\3rd
by deducting 1\2 to 1\3 the value of the PSU's rated amps and see if it still fits
it actually varies with the distribution ratios your likely to need more +12V than +3.3V or +5V
(CPUs now being powered by the +12V primarily)
possibly more if it a long term infrastructure investment and there is growth built in
of if the veracity of the manufacturer is in question (generics tend to lie like dogs)
most 250 or 300 watt PSUs will actually run most configs, but stability has become an increasing concern with the tighter tolerances onboard (FSB)

There is a decrease in total capacity with the rise in temperature , which reduces your amps, the rated amp values where taken at 25C
while your likely operating temperature will be 40C (especially if the PSU is in the top of the case exhausting the CPU HSF) and that is roughly a 30% decrease

That is offset by the additive nature of the calculator, employing all tha maximum draw figures for the assorted components, something that will never occur

However it gets even more complicated if you have alot of drives and fans, those are typically given a "run time" draw value in a calculator, there "spinup" draw can be 4 to 5 times as much and they greatly contribute to transient response overshoot and undershoot in some supplies at startup if there isnt enough +12V

The way you torture a power supply is to give it a fluctuating AC feed to deal with (from surge to brownout), at the same time you ask it to deal with a really dynamic internal load change (like spinning up alot of drives) while still keeping the rails stable enough for the onboard voltage regulation components of unknown quality :p

Failure anywhere along the chain from too big a spike at the source to too long or high an overshoot or undershoot to the mobo, with too much ripple or noise for the onboard regulator to deal with and RAM or other components can go bye bye, ideally the power supply will trip off and protect your components, the operable range it has is largely what the difference in one PSU to another is about. And when it comes to a comparision of a flyweight generic, the whole protection scheme of shutting down in time really comes into question. And of course how stable the rails can be maintained, how low the AC Ripple and noise.

so if you elect to get a supply in the near future you have to ask yourself
if its going to be a true infrastructure investment
in the past that was typically true, now it a little tougher with more power hungry devices
PCI Express, video cards ect.

They recently added 4 more pins to the main connector from 20 to 24, and an additional 4 pin +12V auxillary power connector, (ATX12V 2.0) and the spec keeps jumping the total amps on the +12V rail (ATX12V 2.0 & 2.2), actually there are now two +12V rails and Ive seen power supplies that have Quad rails

ATX12V v1.1 (ATX 2.03 standard) is a 20 pin PSU with a 4 pin +12V connector
but if your considering a long term investment the most important thing for you to determine is the number of pins the mobo connector has, and get a ATX12V v2.2 compliant power supply (unless you need an SSI Compliant EPS12V), and if you have a 20 pin connector see if it can be attached directly with a 24 pin PSU or if an adapter is needed (cap clearence)

as far as what your rails are reading in a monitoring program, for starters,
you cant observe that during startup with software or the BIOS


here is a Codgen300X1 under a dyanmic load
Codegen300X1.jpg

from > http://terasan.okiraku-pc.net/dengen/tester/index.html
and > http://terasan.okiraku-pc.net/dengen/tester2/index.html
(but hosted independently)
note the instabiliy at spinup (an extreme example, but thats common if you review the links)

those are logged Digital Multimeter readouts off a Sanwa PC510 (0.08% accuracy) via RS232C or USB port to PC Link the logging software you see there

the test system is a Tyan Tiger 133 (S1834D).w\ Dual P3 650s overclocked to 865MHz (I think :p )
Dual sticks of 128MB PC133, a Matrox G400SH, Xwave 6000 sound card, NIC and a single HDD, FDD and CDROM
measured from startup, booting into Windows 2000, and Running 3DMark 2000 to conclusion, measured in 5 second intervals, which is definately less than ideal, but the best he could manage

not exactly a power hungry system that would be "pushing" those supplies too hard
I use those as illustrative of the instability that spinning up drives can produce
those supplies are hardly at the edge of their capacity, nor are they dealing with less than ideal AC power
Ripple & Noise, not being specifically addressed.
and while they arent "out of spec" they arent being taxed either


compared to a PC Power & Cooling 450ATX
PC Power450atxs.jpg

same source

so Ideally youd like to use a Digital Multimeter to read your rails directly, watch the spinup,
and then if you can find a realtively stable voltage state use it to calibrate the voltage your reading on the DMM to the software (easy to do in MBM)

a bit more cut and paste
-----------------------------------------------------------------------------------------------

Continuous Power vs. Peak Power at Spin-Up
12V power profile (current vs. time) of an IDE/ATA hard disk at startup. You can see that the peak power draw is over quadruple
the steady-state operating requirement. The graph appears "noisy"
due to frequent oscillations in current requirements

Peak vs. Continuous Power
Despite this extra capacity, it is still a good idea to not load up your system to the very limit of your power supply's stated power capacity. It is also wise, if possible to employ features that delay the startup of some disk drive motors when the PC is first turned on, so the +12 voltage is not overloaded by everything drawing maximum current at the same time.

"the majority of damaged RAM returned to memory manufacturers is destoryed by fluctuations in the voltage."
http://www.anandtech.com/showdoc.html?i=1774&p=8

Winbond Launches New Bus Termination Regulator April 4th 2003

"Winbond Electronics Corporation, a leading supplier of semiconductor solutions, today launched the W83310S, a new DDR SDRAM bus termination regulator. The solution, new to Winbond's ACPI product family, is aimed at desktop PC and embedded system applications with DDR SDRAM requirements.

Computer systems architectures continue to evolve and are becoming more complex; CPU and memory speeds continue to increase ever more rapidly with every technology turn. More and more high current/low voltage power sources are required for PC systems. This is particularly true for high-speed components such as CPU, memory, and system chipsets. The performance of these components is highly dependent upon stable power. Therefore, motherboard designers require accurate, stable, low-ripple and robust power solutions for these components.

Many system designs use discrete components to implement bus termination functions. This approach creates several problems including poorer quality load regulation; higher voltage-ripple, increased usage of board space and inconsistent designs when different discrete components are used.
"


the transient response is the critical internal measure, unfortunately its not a metric that is commonly supplied with the PSU specs
(this seems to be slowly changing, as some manufacturers are supplying the transient response now)

Transient Response: As shown in the diagram here, a switching power supply uses a closed feedback loop to allow measurements of the output of the supply to control the way the supply is operating. This is analogous to how a thermometer and thermostat work together to control the temperature of a house. As mentioned in the description of load regulation above, the output voltage of a signal varies as the load on it varies. In particular, when the load is drastically changed--either increased or decreased a great deal, suddenly--the voltage level may shift drastically. Such a sudden change is called a transient. If one of the voltages is under heavy load from several demanding components and suddenly all but one stops drawing current, the voltage to the remaining current may temporarily surge. This is called a voltage overshoot.

Transient response measures how quickly and effectively the power supply can adjust to these sudden changes. Here's an actual transient response specification that we can work together to decode: "+5V,+12V outputs return to within 5% in less than 1ms for 20% load change." What this means is the following: "for either the +5 V or +12 V outputs, if the output is at a certain level (call it V1) and the current load on that signal either increases or decreases by up to 20%, the voltage on that output will return to a value within 5% of V1 within 1 millisecond". Obviously, faster responses closer to the original voltage are best."



Clownboat said:
I can't seem to figure out if there's any significant difference between the "P series" the "AX model", i.e. 475AX or 475P.
 
Power Factor Correction
which is required in most countries, but not in the US
not a real big consideration, for the quick breakdown
> PFC Decoded @ dansdata

but both the EG475AX-VE & EG475P-VE still have all the Amps on the +12V1 rail
that changes in both lines if you go either down (425) or up (701)
I know that right now your not thinking that your going to be using this in another computer that will be 24 pin, but traditionally a PSU is considered an infrastruture investment, and there are a few other changes that where made in the ATX12V v2.01 Power Design Guide other than just the 20 to 24 pin
(Im going to type out this and cut and paste it later so all of them might not be applicable to this particular supply, like for instance the efficiency)


ATX12V Version 2.0

1.2. Key changes for ATX12V Version 2.0 and Later as Compared with ATX and Previous Versions on ATX12V Power Supply
This section briefly summarizes the major changes made to this document that now defines ATX12V power supply.
With the move to 12V voltage regulators for the processors, ATX guidelines for 5V as mains power are no longer provided

1.2.1. Increased +12 VDC output capability
System components that use 12V are continuing to increase in power.
In cases where expected current requirements is greater than 18A a second 12V rail should be made available.
ATX12V power supplies should be designed at accommodate these increased +12 VDC currents.

1.2.2. Minimum Efficiency
Minimum measured efficiency is required to be 70% at full and typical (-50%) load and 60% at light (-20%) load.
New recommended guidance has been added to provide direction for future requirements.

1.2.3. Main Power Connector
The 2 x 10 main power connector has been replaced by a 2 x 12 connector.
This was made to support 75 watt PCI Express requirements.
Pinout assignments are based on SSI recommendations.
With the added 12V, 5V, and 3.3V pins the need for an Aux Power Connector is no longer needed
and the guidance for this connector has been removed

1.2.4. Seperate current Limit for 12V2 on the 2x2 connector
The 12V rail on the 2 x 2 power connector should be a seperate current limited output to meet the requirements of UL and EN60950

3.2.3. Typical Power Distribution
DC output power requirements and distributions will vary based on specific system options and implementation.
Significant dependencies include the quantity and types of processors, memory, add-in card slots, and peripheral bays,
as well as support for advanced graphics or other features.
It is ultimately the responsibility of the designer to derive a power budget for a given target product and market

Table 3 through Table 5 and figure 1 through Figure 3 provide sample power distributions and graphical recommendations for cross loading.
It should not be inferred that all power supplies must conform to these tables,
nor that a power supply designed to meet the information in these tables will work in all system configurations.

.
ATX350.jpg


ATX400.jpg
-------------------------------------------------------------------------------------------------------------------------

and for comparision how the Server System Infrastructure (SSI) EPS12V spec calls out the +12V Rails (many supplies meeting both specs with a simple harness change)

6.1.1 12V Power Rail Configuration

There are two types of 12V rail configurations for systems: 'Common plane' and "Split plane' processor power delivery. The 'common plane' system has both processors powered from a single 12V rail (+12V1) from the power supply. The 'split plane' system has both processors powered by seperate 12V rails (+12V1 and +12V2) one dedicated to each processor. The system in both cases, has an additional 12V rail to power the rest of the baseboard +12V loads and dc/dc converters. +12V1, +12V2 and +12V3 should not be connected together on the baseboard to ensure that 240VA protection circuits in the power supply operate properly

Table 6: 12V Rail Summary
........................................................................................................................................................................................
Common Plane System........................................................Split Plane System
+12V1........Processors.........................................................+12V1........Processor 1
+12V2........Baseboard components other than processors.......+12V2........Processor 2
+12V3........Drives and peripherals..........................................+12V3........Baseboards and components other than processors
...........................................................................................+12V4........Drives and peripherals
---------------------------------------------------------------------------------------------------------------------------
http://www.enermax.com.tw/products_page.php?Tid=1&gon=236&Gid=26&Gid2=35
DC OUT....Rip&Noise...Tolerence...Rng1..Max/Min..Range 2.......Range 3........Range 4
+12V 1 ......120mV ......+5%,-4%..... 1.5A / 14A..... 0.5A / 5A.....1.5A / 14A... 0.5A / 9A
+12V 2 ......120mV ......+5%,-4%..... 1.5A / 16A..... 0.5A / 4A ....1.5A / 18A... 0.5A / 6A
+12V 3 ......120mV ......+5%,-4%..... 1.5A / 16A .....0.5A / 6A ....1.5A / 16A... 0.5A / 9A
+12V 4 ......120mV ......+5%,-4%..... 1.5A / 14A .....2A / 5A .......1.5A / 14A... 2A / 9A


so I checked the manual for that board, and there is no 6 pin auxillary power connector
just the 20 pin main and 4 pin 12V (JWR1 & JPW1) and there is plenty of clearance around the connectors,
so you could use any ATX12V v2 or even EPS12V supply youd like

that Enermax P just makes me nervous with all the Amps on the +12V1


well, that power supply will run your computer

and provided it has stable AC power to work with would likely do it without any issue,
(that was a hint invest in power conditioning )
the calculator just examines the current distribution required and as mentioned,
its the theoretical maximum, something that is unlikely to ever be approached offset by the temperature at which they measured that current,
being substantially lower than your likely operating temperature

It doesnt address the quality of the power (AC Ripple and noise) or how well it can deal with a fluctuating input AC load > Line Regulation
or a changing draw from the components > Transient Response
its not exactly a "tight" supply, having just the basic compliance with the ATX12V spec for noise and regulation

But you have sucessfully avoided the first pitfall of thinking watts mean much these days
a quailty 300watt out performing a generic 400 watt or even a 500 watt in many cases

That is a realitively good supply, and the veracity of the specs within the "normal" range,
they where likely measured with an actually useful threshold for unacceptable sag (brownout)
on the various rails (the point where they would drop below or shoot above the specification of the 3.3, 5, or 12 volt rails which is 5%
other generic supplies might just employ when a fuse blew

Im currently researching the draw of several different components including the 6800 Ultra
the tough part is they dont seem to publish that data but I do have this article from Spodes Abode
http://www.spodesabode.com/content/article/6800upower
It is almost impossible to measure how much current the cards draw through the AGP slot, but from the AGP specifications we know the absolute maximum the slot can supply is a total of 46 Watts, at various voltages. This means the overall power consumption of the GeForce 6800 Ultra must be somewhere between 77.5 Watts, and 123.5 Watts.
and that the auxillary power connector was drawing an additional 5 Amps (peak) on the +12V rail, and 3.5A (peak) on the +5V rail,
measured directly by the multimeter

the problem becomes that part of that figure is also included in your original calculation
so how much more to add to either rail is problematic, but in comparision a 9800 XT was drawing only 2.2Amps (+12V) and 3.5A (+5V) peak through the auxillary connector
so adding 3 to 5A to your +12V total for an Ultra, but Nvidia claims the GT are considerably lower powered.

ATX12V Version 1.3 Power Supply Design Guide underwhich this supply was built predates the current heavy employment of the +12V rails
and the upgrade path of any v1.3 supply is extremely short, the next generation (and some of the current gen) of mobos having 24 pin main power connectors
and real need of much fatter +12V rails, The PSUs in the v2.1 spec typically jumping from the 18A range up to 30>38A (broad generalization for a "typical" gaming rig sized PSU which is quite large in Watts) and split over two or more dedicated +12V rails, largely because of the additional requirements of PCI Express, and the change from using onboard +5V Voltage Regulation Modual(s) (that step down and power the CPU) to +12V powered VRMs

Id say that if the budget is firm, and your not getting the 6800GT, your safe, possibly even with the GT
provided the input power is stable, but that as little as $15 could make a big difference in peace of mind and likely let you run the GT with peace of mind as well, however your still investing in a very limited "patch" power solution that isnt going to see you into the next rig.

Good Luck ;) If youd like to reconsider the budget or a particular supply Id be happy to help
I do notice its on sale right now for $36, so its certainly an affordable patch, regarding the GT, I cant really say, sorry
http://www.newegg.com/app/ViewProductDesc.asp?description=17-153-006&depa=0
.

as a last note, if your where able to say force feed your computer AC air
or otherwise maintain a low ambient temperature, the capacity of the supply will be substantially better, and would likely be able to handle that "theoretical" extra load without issue, of course the opposite is true as well, if the room temperature is 100F you could easily develop power problems if loaded to close too your limit

In my rackmount I dont have the power supply exhausting the heat produced from the CPU HSF, a trend which accelerated with the AMD builders guide's for the Athlons (adding the second fan) solving that exhaust issue a different way and maintaining a cool air input for your PSU could be worth as much as a 10% to 20% increase in overall capacity, lowering it to the actual 25C it was tested at could be worth 30% increase over a typical 40C PSU enclosure ambient ;)

("typical" meaning exhausting a HSF directly below it in say a 70>80F room ambient with a pretty powerful and hot CPU, it could vary quite a bit)

still working on this so... expect editing think that enermax is OK
will doublecheck

IsAnybodyHome said:
Ice Czar:
hmm... I'm thinking that spending some extra money on a power supply would very much worth the peace of mind it'll bring me. But money will be tight, especially if i go with the 6800GT... Are there any decent v2.1 spec PSU's out there around $60? Or maybe even cheaper?? If i go with a lower wattage unit, as long as it can keep up with the +12V rail requirements. Like you said, the PSU will still be a patch, but that ok with me because i don't plan on adding any other components to my PC or seeing this PSU through to my next rig.


well, thats a little tough, for starters the "public" isnt really aware of the ATX12V v1.2 vs v1.3 vs v2.0 issue, and thus the supplies arent typically listed like that yet by retailers, they dont want to deal with people mistakenly buying a PSU that they cant attach to the mobo without a converter (if there isnt clearence on the connector) or why there are an extra 4 pins on the main connector

the mobo manufacturers are largely to blame since they didnt adopt those 24 pin connectors when the spec changed and have been playing a shell game since the spec came out cause they dont want to "force" their customer to buy a new PSU and have been dancing around the issue, but thats coming to an end soon. In the meantime this is a period of great confusion

so a recap
v1.2 your basic single +12V rail PSU with a 20 pin main connector
v1.3 a dual rail +12V w\ 20 pin main plus a 2x2 four pin +12V auxillary
v2.0 a dual rail +12V w\ 24 pin main plus a 2x2 four pin +12V auxillary

with each of those, more amps where added to the +12V rails so the supplies that are the fatest are the latest v2.0, only some of those are sold in triple wiring harness config as 20 pin + 4 and 24 pin + 4 and EPS12V 24 pin + 8

the easiest way to search for such an animal is to look for an EPS12V supply, those are a slightly different spec Intels Server System Infrastructure and the connectors are compatible for the v2.0 mains formfactor.org having adopted the mains from there, but the auxillary connector (Intel 2x2 +12V) is transformed in the EPS12V spec to an 2x4 eight pin, and again it can be used directly if there is clearence, but I know of no converters from 8 pin to 4 pin, as of yet (though I think I might start a company :p )
now most ATX12V v2.0 supplies are also EPS12V compliant and they just have different wiring harnesses, but when they are listed as ATX12V supplies, they often dont bother to list the +12V individual rails :rolleyes: but would in their EPS12V wiring harness trim

such an animal is typically $100 or so, but there is one at newegg for $65
AMS MERCURY EPS 12V 460W UL
with dual +12V rails > +12V1@18A, +12V2@15A

now just to make it more confusing there are those dual rails, one of the reasons they where implemented was to isolate the volatage instabilities induced by other components turning on and off (Transient Load) so in the EPS12V spec its listed like this

EPS12V ....6.1.1 12V Power Rail Configuration

There are two types of 12V rail configurations for systems: 'Common plane' and "Split plane' processor power delivery. The 'common plane' system has both processors powered from a single 12V rail (+12V1) from the power supply. The 'split plane' system has both processors powered by seperate 12V rails (+12V1 and +12V2) one dedicated to each processor. The system in both cases, has an additional 12V rail to power the rest of the baseboard +12V loads and dc/dc converters. +12V1, +12V2 and +12V3 should not be connected together on the baseboard to ensure that 240VA protection circuits in the power supply operate properly

Table 6: 12V Rail Summary
........................................................................................................................................................................................
Common Plane System........................................................Split Plane System
+12V1........Processors.........................................................+12V1........Processor 1
+12V2........Baseboard components other than processors.......+12V2........Processor 2
+12V3........Drives and peripherals..........................................+12V3........Baseboards and components other than processors
...........................................................................................+12V4........Drives and peripherals

the quad rail interation obviously being for a dual CPU board and you wouldnt want it for a single CPU

but what is strange is that though some supplies state they are dual rail they appear to not really be like the middle supply in the Enermax Noisetaker Line here
Enermax EG425-VE SFMA (ATX12V v1.3)
+12V1..... 0.5A / 15A (MIN/MAX)
+12V2.......0.5A / 14A

Enermax EG475P-VE SFMA (ATX12V v1.3)
+12V1..... 1.5A / 33A
+12V2.......0A / 1A


Enermax EG701P-VE SFMA (ATX12V v1.3)
+12V1..... 0.5A / 18A
+12V2......0.5A / 17A

so the later spec supplies have isolated rails for the 12V power (all the way up to quad)
what your starting to see now is strange VRM schemes like Gigabytes DPS Dual Power System which is a 6 phase converter on the mobo, and OCZ is releasing Power Clean Technology, all this in a effort to get cleaner more stable power to very sensitive components, those are just a few examples, the VRM moduals on my mobo run hotter than my chipsets

I think alot of this has to do with the older power supplies, you dont really see the bending over backwards on server boards where the EPS12V spec calls out triple rails and quad rails, again the obvious reason is to isolate the transient loads

if there isnt an overshoot (to much ) or undershoot (too little) voltage on a given rail the Voltage regulation modual onboard is better able to deal with it, step it down and feed it to the CPU and RAM so isolating the rails is a really helpful thing

Regarding Generics
A Different Perspective on Power Supplies
Power supplies become increasingly expensive
 
twajetmech said:
and leave yourself enough headroom for a few future upgrades.
twajetmech thanx for the insight ;)

and the above statement directly touches on another point
in the past investment in a PSU could be considered an infrastructure investment
and as such the ATX standard has alot of inertia behind it, so much so that when the extra +12V amps was required, that almost all mobo manufacters adopted to ignore the new ATX12V v2 standard with the 24 pin ATX main power connector in favor of not making their customers shell out more for a new PSU and added some form of 4 pin connector supplemental to the board (some normal molex connectors) the same goes with supplemental power of with video cards, instead of widespread adoption of AGP Pro
All this to retain the "typical" 20 pin main power connector on legacy ATX PSUs and compatibility

leading the unsophisticated consumer to ask, well I got a 400Watt isnt that good enough?
with the invariable answer, hell Im running a 350 watt and have more components powered than that

but different amp distribution
a good 350 watt is far better than a crap 500 watt

In the workstation and server market, the standard has "normalized" with the adoption of EPS12V compliant SSI PSUs

but in the desktop market its still in flux, and will be even more so with the adoption of BTX

so its no longer a infrastructure investment, and that little bit extra, might be for your antique collection of monster ATX form factor mobos, compared to your nice petite picoBTX


Myrdhinn, Im pretty sure computerpro3 reference to "REAL" 510 watts means that that figure was derived at 40C as a test temperature, the power supply when compared to the Antec test parameters of 25C would actually be a 600 Watt PSU, and have a max output of 650 Watts
PCP&Cs offer the fattest amps per rail of any desktop PSU I know,
(excluding redundant shared bus N+1s which go all the way up to +800watts possibly more)
and the Fortron 530 right up there as well, and every time I see them reviewed, come in well over their ratings on amps

I suspect that FSP Group (Fortron Source \ Sparkle Power) is again manufacturing for PCP&P
but likely "to spec" of a PCP&C design, and there are likely several manufacturers rounding out their various models
(Zippy Emacs for instance, a big player in server PSUs)

ATX = 20 pin plug, your Pentium III's ATX connector and PS

ATX12V v1.1 = 20 pin atx connector + 4-pin plug for "Additional 12V" (same pdf as above)

ATX12V v2.0 = 24 pin atx connector + 4-pin plug for "Additional 12V"

ATX12V v2.2 = 24 pin atx connector + 4-pin plug for Aux +12V to support PCI Express

SSI Compliant PSUs

EPS12V v1.6 = Power supply with 24-pin EPS12V connector, plus one 8-pin additional 12v connector

EPS12V v2.1 = Power supply with 24-pin EPS12V connector, plus one 8-pin additional 12v connector
added higher power levels for 650 watt PSUs and updated 12v peak requirements for 450 & 550 watt PSUs

EPS1U v2.1 = Power supply with 24-pin EPS12V connector, plus one 8-pin additional 12v connector fitting into a 1U height

EPS2U v2.1 = Power supply with 24-pin EPS12V connector, plus one 8-pin additional 12v connector fitting into a 2U height

ERP2U v2.0 = Redundant Power Supply with 24-pin EPS12V connector, plus one 8-pin additional 12v connector fitting into a 2U height



Organizations

SSI = Server System Infrastructure, an Intel spec, which, among other things, defines the "EPS" "enhanced" ATX specification

FormFactors.org

BTX PSU Form Factors

Balanced Technology Extended Interface Specifications v 1.0a Describes basic 24 pin main power connector and Aux +12V 4 pin connector

CFX12V v1.0 = 24 pin + Aux 4 pin +12V Power connector Compact Form Factor with 12V connector

LFX12V v1.0 = Low Profile Form Factor with 12 Volt Connector

Special PSU Form Factors

TFX12V v1.2 = 20 pin connector Thin Form Factor with 12V Connector

TFX12V v2.0 = 24 pin connector Thin Form Factor with 12V Connector

SFX12V v2.3 = 20 pin connector Small Form Factor with 12V Connector

SFX12V v3.0 = 24 pin connector Small Form Factor with 12V Connector

"Updates for these design guides include an increased +12 VDC output capability to support system components that are continuing to increase in power, minimum efficiency for both standby and active modes of operation, details on an optional S-ATA power connector to support devices such as S-ATA drives, and acoustic guidance to support low noise systems. PS3 mechanical guidance has also been introduced into the SFX12V Power Supply Design Guide."
-Intels Desktop Form Factors for ATX12V, SFX12V and TFX12V PSU Design Guides
 
gee said:
On "multiphase" power...

CPU core voltage regulators (aka, VRM's) are what are called synchronous buck regulators. In a simplest form (a "one phase" regulator) a synchronous buck converter looks like this:

Code:
  5V/12V
    |
  |-+
--|   (fet)
  |-+
    |
    +--[inductor]--+----- OUT
    |              |+
  |-+             ---
--|   (fet)       --- (capacitor)
  |-+              |
    |              |
    V              V
During operation, the top FET turns on and the left side of the inductor goes to 12V. Current starts flowing through the inductor and the capacitor is charged. When the desired voltage is met on the capacitor, the top fet turns off and the bottom one turns on. Another way to describe operation is that the two FETs create a square wave with an "average" voltage, and the inductor and capacitor filter that square wave into the actual core voltage.

When the top FET is off, the CPU gets its current from the charge in the capacitor, combined with a (usually much smaller) freewheeling current in the inductor. This causes the voltage on the capacitor to drop. Then when the top FET turns back on, the CPU is powered through the top FET and the capacitor charges up again. So what happens is that you get ripple on the capacitor - this is a problem if the ripple is too high.

Another problem is that if the CPU is sucking close to 100A, then the FETs have to handle this current, and making a FET that handles that kind of current *and* switches at several hundred KHz is more or less impossible. So what multiphase converters do is take two or more of these circuits and connect the outputs together. Since the CPU current is spread across several FETs, the amount of current that individual FETs have to conduct is less and the circuits run very efficiently. Also not only are these 100A currents hard on the FETs, they're also hard on the input and output capacitors that have to handle these currents.

But the real beauty of multiphase power supplies is that many of the above circuits are "phased" so that they "take turns" providing power to the CPU.

Suppose you have a 12V powered VRM and a 1.5V CPU, then the top FET in a single circuit will only be on for 12.5% of the time. This means that for the remaining 87.5% of the time, the output capacitor has to power the CPU, and it has to be large enough so that the CPU voltage doesn't drift beyond the ripple spec. But hook two converters up 180 degrees out of phase, and the output capacitor will only have to power the load for 50-12.5% = 37.5% of a cycle - this drops the ripple by more than half. Hook up 3 or 4 phases, and things get far better. The effect is that much cleaner power is delivered more reliably to the CPU. The only expense is that you need more FETs and inductors, but they're only pennies anyway.

This is a half-asleep EE writing this, so this may make no sense at all... but I hope it clears things up.

something or other so it will let me post the damn qoute

Ice Czar said:
You know anything about Gigabytes proprietary Dual Power System? (6 phase)

Haven't even heard of it. What do they mean by "6 phase" here? They don't mean signal phases, do they?

while we are at it, what is a few typical implementations of Voltage Regulation Moduals\Power Distribution?

Intel publishes schematics for their reference motherboards. It looks like the docs for this 875P motheboard is the newest one they've posted. Drawing #74 in that file starst the documentation for the power distribution; Drawing #75 shows the motherboard's power supply connector, and starts describing how lines fan-out from there.

Section 15 of the 875P Chipset Platform Design Guide describes the scheme Intel likes. Figure 15-1 in that document gives a big block diagram of their distribution map, which is a little more approachable than the schematic. (The schematic uses lots of signal names -- and I can't find a table that gives definitions for them. It's not too easy to interpret the schematics because of that.)

.B ekiM

gee said:
In a simplest form (a "one phase" regulator) a synchronous buck converter looks like this:

Sure. But there's a lot you've left out: there's feedback from the output to some circuit that compares to the desired regulated voltage, then decides which transistor to turn on.

So what multiphase converters do is take two or more of these circuits and connect the outputs together.

Do you mean that they're ganging the transistors to get more current through them? Or that they're bridinging multiple buck converters together? That is, for a multiphase implementation, is there a single feedback loop and controller, or many? As many as there are phases?

It seems like there couldn't possibly be multiple controllers. Or, at least, that it must be very difficult to have multiple controllers. If there were, what prevents one controller from thinking the voltage is too high, turning on the low-side FET, while the other converter thinks the voltage is too low, and turning on its high-side FET? Maybe I'm not saying it the right way, but how do you manage cross-phase shoot-through? That would be bad, but certainly possible because of variances in component tolerance.

But the real beauty of multiphase power supplies is that many of the above circuits are "phased" so that they "take turns" providing power to the CPU.

Suppose you have a 12V powered VRM and a 1.5V CPU, then the top FET in a single circuit will only be on for 12.5% of the time. This means that for the remaining 87.5% of the time, the output capacitor has to power the CPU, and it has to be large enough so that the CPU voltage doesn't drift beyond the ripple spec. But hook two converters up 180 degrees out of phase, and the output capacitor will only have to power the load for 50-12.5% = 37.5% of a cycle - this drops the ripple by more than half. Hook up 3 or 4 phases, and things get far better. The effect is that much cleaner power is delivered more reliably to the CPU. The only expense is that you need more FETs and inductors, but they're only pennies anyway.

OK, so that helps alot. Then, there's multiple controllers, multiple FETs and inductors, but only one capacitor? Is that right? How does the implemnetation of the controller change for this? This is really just like pulse-width modulation motor control, isn't it?

Is there a book or website (or even just a manufacturer's design guide) you'd recommend? That would keep me out of your hair!

Thanks for the informative post, gee.

.B ekiM
 
Originally posted by gee, just as some of the above was by mikeblas
just gathering everything in one place here ;)


mikeblas said:
Sure. But there's a lot you've left out: there's feedback from the output to some circuit that compares to the desired regulated voltage, then decides which transistor to turn on.
I've also left out the input capacitor, the input inductor, the controller, the current sensing scheme, the VID lines from the processor and many other things. I'm trying to keep it simple. :D
Do you mean that they're ganging the transistors to get more current through them? Or that they're bridinging multiple buck converters together? That is, for a multiphase implementation, is there a single feedback loop and controller, or many? As many as there are phases?

It seems like there couldn't possibly be multiple controllers. Or, at least, that it must be very difficult to have multiple controllers. If there were, what prevents one controller from thinking the voltage is too high, turning on the low-side FET, while the other converter thinks the voltage is too low, and turning on its high-side FET? Maybe I'm not saying it the right way, but how do you manage cross-phase shoot-through? That would be bad, but certainly possible because of variances in component tolerance.
They're bringing multiple buck converters together. Each "phase" has two FETs and an inductor (and its own current sensing method, usually) and they combine together at the output capacitor. There's usually a big row of output capacitors instead of just one... these caps see an ungodly amount of ripple current, and a single capacitor wouldn't have the current handling capability to not explode, and a low enough ESR (inline resistance) to keep the controller stable.

In every computer motherboard that I've ever seen, excluding SMP motherboards with independant VRMs per CPU, there's only one controller chip. But this isn't necessarily impossible; you can get controller chips that are designed to be chained together and allow a ridiculous number of phases to be used for ultra-high-current applications.

With regards to the "phases interfering with each other" issue, these controllers don't directly use a voltage sample to determine when to run, but instead they measure their output current and shut off their top FETs when a set current is reached - and feedback from the output voltage is used to set this "trip point". It sounds confusing - google for "current mode control" and you might find a better explanation than I can give.

OK, so that helps alot. Then, there's multiple controllers, multiple FETs and inductors, but only one capacitor? Is that right? How does the implemnetation of the controller change for this? This is really just like pulse-width modulation motor control, isn't it?

Is there a book or website (or even just a manufacturer's design guide) you'd recommend? That would keep me out of your hair!

Thanks for the informative post, gee.

.B ekiM
At work, I use quite a few parts from Linear Technology. Here's some you might find interesting... read through some of these datasheets and their appnotes.

http://www.linear.com/prod/datasheet.html?datasheet=1145 - LTC3738, 3-phase Intel VRM9/10 Vcore regulator.
http://www.linear.com/prod/datasheet?datasheet=513 - LTC1629, 2-phase switching regulator. You can chain 8 of these together to create a *12 phase* regulator!

ON Semi and National Semiconductor also have similar offerings. And TI probably makes more individual switching regulator chips than all three of these combined...
 
Ice Czar said:
answering my own question

Field Effect Transistor. A solid-state device in which current is controlled between source and drain terminals by voltage applied to a non-conducting gate terminal. See also channel, drain, gate, and source.

Id assume
Correct.

Motherboards use N-channel MOSFET transistors in their Vcore regulators. The more hefty ones used on motherboards can turn on with a lower internal resistance than the traces that actually hook them up, and they can switch on and off at many MHz.

Regular NPN transistors can't switch anywhere near fast enough, and that aside, their on-voltage combined with the current they face in a Vcore regulator would blow them up anyway!
 
http://support.intel.com/support/processors/pentium4/sb/CS-008619-prd483.htm

Intel® Pentium® 4 Processors
Processor only boots at 2.80 GHz



Symptom(s):


Processor boots at 2.80 GHz
Processor running slow
Slower speed than expected

Solution:
Many board manufacturers have designed their initial Intel® 865 and 875 chipset-based motherboards to previous platform guidelines. These motherboards may not meet the currently required processor electrical, mechanical, and thermal specifications. Some processors 3.20 GHz and above based on 90nm technology may only boot to 2.80 GHz. When this invalid configuration is detected, the processor will run at this slower frequency to reduce the chance of damage to the processor and/or motherboard.

Today, however, most vendors are now shipping boards that comply with the new revised guidelines that do address the electrical, mechanical, and thermal specifications of these processors. The revised guidelines can be located at http://developer.intel.com/design/Pentium4/guides/

There is no simple method to visually determine whether a motherboard meets the new platform requirements other than to reference or verify the motherboard AA# with the board manufacture.Check carefully with the board vendor to ensure board compatibility with the Intel® Pentium® 4 processors at 3.20 GHz and above or with the 'E' letter designator.

Intel 845 chipset-based boards are not compliant with the new CPU requirements and may not have the necessary updates to recognize the new processor speeds, resulting in only booting to 2.80 GHz. Some systems may display a message that outlines this invalid configuration.

See also: Information on Intel® Desktop boards

For information on Third Party boards- check with the board manufacturer
 
ATX12V v2.0
1.2 Key Changes for ATX12V Version 2.0 as Compared with ATX power Supply

This section briefly summarizes the major chaanges made to this document that now defines ATX12V power supply. With the move to 12V voltage regulators for the processor, ATX guidlines for 5V as main power are no longer provided.

1.2.1 Increased +12 VDC output capability
System components that use 12V are continuing to increase in power. In cases where expected current requirements is greater than 18A a second 12V rail should be made available. ATX12V power supplies should be designated to accommodate these increased +12VDC currents.

1.2.3 Main Power Connector
The 2 x 10 main power connector has been replaced by a 2 x 12 connector. This was made to support 75 watt PCI Express requirements. Pinout asignments are based on the SSI recommendation.
With the added 12V, 5V, and 3.3V pins the need for an Aux Power connector is no longer needed and the guidance for this connector removed.

1.2.4 Seperate current limit for 12V on the 2x2 connector:
the 12V rail on the 2x2 power connector should be a seperate current limited output to meet the requirements of UL and EN 60950

4.5.1. ATX Main Power Connector
Connector Molex housing: 24 pin Molex Mini-Fit Jr. PN# 39-01-2240 or equivalent
(Mating motherboard connector is Molex 44206-0007 or equivalent)
18 AWG is suggested for all wires except for the +3.3V sense return wire pin 11 (22 AWG)
(For 300 W configurations, 16 AWG is recommended for all +12VDC, +5VDC, +3.3VDC and COM
Pin.....Signal.......................Color.........Key
1........+3.3VDC...................Orange......Square
2........+3.3VDC...................Orange......Dome
3........COM.........................Black........Dome
4........+5VDC......................Red...........Square
5........COM.........................Black........Square
6........+5VDC......................Red...........Dome
7........COM.........................Black........Dome
8........PWR_OK..................Grey..........Square
9........+5VSB......................Purple.......Square
10.......+12V1DC..................Yellow.......Dome
11.......+12V1DC..................Yellow.......Dome
12.......+3.3VDC...................Orange.....Square

13.......+3.3VDC...................Orange.....Dome
[13].....[+3.3V default sense...Brown......Square
14.......-12VDC......................Blue.........Square
15........COM.........................Black........Square
16........PS_ON#....................Green.......Dome
17........COM.........................Black........Dome
18........COM.........................Black........Square
19........COM.........................Black........Square
20........Reserved...................N/C...........Dome
21........+5VDC......................Red..........Dome
22........+5VDC......................Red..........Square
23........+5VDC......................Red..........Square
24........COM.........................Black........Dome

4.5.2 +12V Power Connector
Connector Molex 39-01-2040 or equivalent
(Mating motherboard connector is Molex 39-29-9024 or equivalent)
Pin.....Signal.........18 AWG Wire................Key
1........COM..........Black............................Square
2........COM..........Black............................Dome
3........+12V2DC....Yellow / Black Stripe......Dome
4........+12V2DC....Yellow / Black Stripe......Square

EPS12V v2.02

Revision History

6.1.1 12V Power Rail Configuration

There are two types of 12V rail configurations for systems: 'Common plane' and "Split plane' processor power delivery. The 'commob plane' system has both processors powered from a single 12V rail (+12V1) from the power supply. The 'split plane' system has both processors powered by seperate 12V rails (+12V1 and +12V2) one dedicated to each processor. The system in both cases, has an additional 12V rail to power the rest of the baseboard +12V loads and dc/dc converters. +12V1, +12V2 and +12V3 should not be connected together on the baseboard to ensure that 240VA protection circuits in the power supply operate properly

Table 6: 12V Rail Summary
........................................................................................................................................................................................
Common Plane System........................................................Split Plane System
+12V1........Processors.........................................................+12V1........Processor 1
+12V2........Baseboard components other than processors.......+12V2........Processor 2
+12V3........Drives and peripherals..........................................+12V3........Baseboards and components other than processors
...........................................................................................+12V4........Drives and peripherals

Table 7: P1 Baseboard Power Connector - Common Plane

Pin.....Signal.......................Color.........Key
1........+3.3VDC...................Orange......Square
..........3.3RS.......................Orange w\ white Stripe
2........+3.3VDC...................Orange......Dome
3........COM.........................Black........Dome
4........+5VDC......................Red...........Square
5........COM.........................Black........Square
6........+5VDC......................Red...........Dome
7........COM.........................Black........Dome
8........PWR_OK..................Grey..........Square
9........+5VSB......................Purple.......Square
10.......+12V2..................Yellow.......Dome
11.......+12V2..................Yellow.......Dome
12.......+3.3VDC...................Orange.....Square

13.......+3.3VDC...................Orange.....Dome
14.......-12VDC......................Blue.........Square
15........COM.........................Black........Square
16........PS_ON#....................Green.......Dome
17........COM.........................Black........Dome
18........COM.........................Black........Square
19........COM.........................Black........Square
20........Reserved (-5V in ATX)..N/C...........Dome
21........+5VDC......................Red..........Dome
22........+5VDC......................Red..........Square
23........+5VDC......................Red..........Square
24........COM.........................Black........Dome

Table 7: P1 Baseboard Power Connector - Common Plane

Pin.....Signal.......................Color.........Key
1........+3.3VDC...................Orange......Square
..........3.3RS.......................Orange w\ white Stripe
2........+3.3VDC...................Orange......Dome
3........COM.........................Black........Dome
4........+5VDC......................Red...........Square
5........COM.........................Black........Square
6........+5VDC......................Red...........Dome
7........COM.........................Black........Dome
8........PWR_OK..................Grey..........Square
9........+5VSB......................Purple.......Square
10.......+12V3..................Yellow.......Dome
11.......+12V3..................Yellow.......Dome
12.......+3.3VDC...................Orange.....Square

13.......+3.3VDC...................Orange.....Dome
14.......-12VDC......................Blue.........Square
15........COM.........................Black........Square
16........PS_ON#....................Green.......Dome
17........COM.........................Black........Dome
18........COM.........................Black........Square
19........COM.........................Black........Square
20........Reserved (-5V in ATX)..N/C...........Dome
21........+5VDC......................Red..........Dome
22........+5VDC......................Red..........Square
23........+5VDC......................Red..........Square
24........COM.........................Black........Dome


6.1.3 Required Processor Power Connector
Connector Housing: 8-pin Molex 39-01-2080 or equivalent
Contact: Molex 44476-1111 or equivalent

Table 9: Processor Power Connector-Common Plane
Pin.....Signal.........18 AWG Wire................Key
1........COM..........Black............................Square
2........COM..........Black............................Dome
3........COM..........Black............................Dome
4........COM..........Black............................Square

5........+12V1........Yellow / Black Stripe......Dome
6........+12V1........Yellow / Black Stripe......Square
7........+12V1........Yellow / Black Stripe......Square
8........+12V1........Yellow / Black Stripe......Dome

Table 10: Processor Power Connector-Split Plane
Pin.....Signal.........18 AWG Wire................Key
1........COM..........Black............................Square
2........COM..........Black............................Dome
3........COM..........Black............................Dome
4........COM..........Black............................Square

5........+12V1........Yellow / Black Stripe......Dome
6........+12V1........Yellow / Black Stripe......Square
7........+12V2........Yellow...........................Square
8........+12V2........Yellow ..........................Dome
 
Status
Not open for further replies.
Back
Top