• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

PSU is failing... how much of a monster do I need?

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

ninegoodthings

Registered
Joined
Feb 7, 2005
I have started OCing my new rig and it's not going well, probably because of the PSU. It turns off like I flipped the PSU switch after I run it for a while at speed... anyway...

Here's what I have hooked up:

AMD 3200+ @ 220x12 = 2640 (2.1v)
PQI Turbo 3200 @ 220 (2.6v)
NF7-S Motherboard
ATI AIW 9600
2xLTD-163 16x DVD Drives
1xHP CD-RW 8000 4x-2x-4x
3xSeagate 160GB 8MB ATA100 (2 on PCI Card)
1xMaxtor 160GB 8MB ATA133
Floppy
SoundBlaster PCI Card
Ethernet Card
2xDelta fans pulling between 5 and 10w
Fan controller

So, essentially, four large hdds, three cd-drives, and an AIW 9600.

My 12v rail was around 11.6v when it was working. I used CPU Burn-in for a couple of hours, then when I came back the computer was turned off.

Right now I'm running 100x12=1200 (1.1v) just to be safe... nice to have a 30c idle though :)

What do you guys think? Do I need to go to 400? 450?
 
here is how I do it (still under construction)

-------------------------------------------------------------------------
Here is your basic tool
takaman's Power Supply Calculator

this PSU calculator will break out the amps per rail
in other words how much current each of the 3 main voltage rails needs to power a given configuration,
these change quite a bit, and Watts alone mean next to nothing these days
especially if there are amps on under employed rails something that is quite common

Figure the Theoretical Maximum

1. Set the Utilization
at the top set the Utilization to 90% and apply 100% utilization to drives and fans

2. Select the Basic Power Scheme
select the basic mobo, this is very important as an AMD board without a +12V mobo connector (little 4 pin 2x2 connector or an inline 6 pin) or a P3 mobo will be powering the CPU off the +5V rail, whereas a modern board will be powering the CPU off the +12V rail
(P4 mobos and AMD boards with an Auxillary +12V mobo connector)

3. Select additional components
make as close a match as possible, if you have figures you can select other device not listed and add the amp values manually
this is especially usefull in the case of Processors not listed or AGP Cards
here are some realworld figures

Power Consumption of Contemporary Graphics Accelerators: ATi
Power Consumption of Contemporary Graphics Accelerators: Nvidia

Example
click on the charts for a given card for instance the ATi X800XT
here you have both a stock speed and an overclocked set of values
employ either but pull the values from the load current column, and you need to add both the power supplied from the AGP slot and the auxillary connector
in this case
+3.3V (AGP only) 2.71A
+5V Aux 3.22A + AGP +5V 0.11A = 3.33A
+12V Aux 3.11 + AGP 0.14A = 3.25A
now leave this window up as we will be using it later
enter the values in the PSU calculator you will have to round to the nearest tenth
so if your overclocking round up, probably a good idea in any event
then double check the combined watts from the calculator to the load summary power column of the chart, in this case
63.8 Watts in the PSU calculator to 63.23 Watts on the chart
(double check like this where possible)

The Processors are pretty up to date, however if you have something older, or a CPU not listed, refer to > Processor Electrical Specifications
Example
you figure and then add the amps based on which rail would be powering the CPU
as mentioned above, that is going to be either the +5V rail on older boards or the +12V rail on newer ones
you will be working from either the Maximum Power Distribution values in Watts,
Total Thermal Power values in Watts, or the
Thermal Design Power values in Watts

as an example we take an AMD Athlon64 3400+
which has a Thermal Design Power of 89 Watts
since AMPS = Watts / Voltage
so 89 Watts divided by 12 Volts equals 7.41Amps
and double checking back to the calculator (in this case)
we have 7.4Amps and 88.8 Watts

wherever possible use amp values supplied by your components manufacturer
this is generally well documented in the case of HDDs and fans, less so with many other copmponents, when in doubt look through the options on the calculator and select a worse case load
following this post will be additional tables of odd components as Im able to compile them, such as baybuses, lights, pumps ect.

4.Note Down the Theoretical Maximum
The calculator will supply you with not only the totals for us to refer to but also a url link to look at your figures, post both your full config, the link and PSU Calculator figures, noting that they are the theoretical maximum, also make sure the link works before you close that window
Example
http://takaman.jp/D/?M=PbQBQbd@dSAgG5DGHTTekG5@0DaH9vHCMZ&english
3.3V rail MAX 5.1A or above
+5V rail MAX 9.9A or above
+12V rail MAX 17.1A or above
+3.3V + +5V rails combined 66.7 Watts or above
+3.3V + +5V + +12V rails combined MAX 272.3 or above
Total Output 282.3 Watts or above
a few notes on what those mean
first many but not all supplies have a combined caopacity for the +3.3V & +5V rail
some are truely independent so that may or maynot apply
second the difference between the 3 main rails combined output
and the total is accounted for by the signal lines, the +5V standby power,
and the largely unused negative power rails
in the latest ATX12V v2.01 spec the -5V rail has actually been eliminated,
but the -12V rail is still there

and finally that Watts figures are deceptive,
on the one hand that theoretical maximum will never happen
all the components drawing full power at the same time simply isnt real
we will be addressing that next
and on the other hand, that total watts figure would be for a supply running at the temperature it was rated at,
something your unlikely to be able to replicate, we will also address that below

Figure a Startup Draw
Any mechanical component in your computer that spins requires more power to overcome inertia and spinup, than it does once its spinning, generally the runtime draw is roughly a quarter of the spinup draw

we will be using the figures we develop here for our realworld worse case senerio that follows as the static draw, but this spinup draw can be critical in and of itself if your building a box with alot of storage
DAS (Direct Attached Storage) NAS (Network Attached Storage) and SAN (Storage Attached Network) appliances has typically required very healthy +12V rails, often also employing redundant power supplies or N+1 suuplies (a modual system where ty[pically 2 moduals are capable of powering the computer and a third hotswap, though more moduals are also common)

many storage controllers allow a staggered spinup of an array so as not to incur the full draw at the same time, this is of course a mitigating factor, and can make the difference between having to buy a much more expensive supply if your close to the capacity of a given supply

so open a calculator and notepad
and add the amp draws for

1. CPU Fan(s) if applicable
2. Case Fans
3. HDDs
4. Pumps if applicable

in the PSU calculator amp draws displayed are for a single device
for instance a 80mm 4500rpm fan is listed as .37A on the +12V rail
youd times that by the number of fans, same for the HDDs
once we have a spinup draw, we are going to quarter it for a static draw
Example
2 x 80mm 6800rpm Delta CPU fans @ 0.5A per on the +12V rail = 1A
3 x 120mm 2600rpm midcase rackmount hotswap fans @ 0.52A per = 1.04A
8 x 40GB 7200rpm HDDs @ 2A per = 16A
18.04A on the +12V rail total
and a static draw of 4A

so for a realworld spinup draw we add the above 16Amps
plus 100% utilization of the mobo, say a RAID controller card
25% to 50% utilization for the CPU and GPU and come up with a figure
we also add in the full spinup draw of an optical drive, but dont include it a a static draw


Realworld Worse Case Senerio

OK employing our static Draw above
we add 100% utilization of the CPU, GPU mobo, and an optical drive spinning up
Ive quized the gaming and video card forums, and its pretty evident that 100% usage of the CPU and GPU can occur in a gaming rig, if you have a different type of box, you might consider exactly what might be hitting full draw at the same time, like for instance a few Gigabit NICs, and the worse case CPU log youve seen on a server

on a modern board the +12V rail will be the primary focus
but on an older board all the rails need to looked at carefully
hell thats true in either case :p

Figure the Individaul Draw on each +12V Rail
If you have a modern board, or if you have a server\workstation
its very likely that you will be looking at PSUs with more than one +12V rail
so we need to look at how much draw there is on each of the rails
it breaks out like this for an ATX12V v2.0(1) supply
1. Add the +12V1 up
CPU, mobo, +12V power for PCI cards
2. Add the +12V2 up
everything else, including the Auxillary Power to the Video Card(s)
(this is where you pull up the Video Card Chart we used above)

for an EPS12V supply its even more fun
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 '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

now that you have a more realistic power draw its time to assess PSU
employing the Theoretical Maximum, Spinup Draw, Realworld Worse Case Senerio, and the Individaul Draw on each +12V Rails[/B] baselines we will adjust and weight PSUs ratings to try to match up applicable supplies


Deciphering PSU Ratings
this is wear we pass through the gates of Moria and into the netherworld of PSU marketing :rolleyes:

Rated Amps and Watts
unless specifically stated otherwise, these figures are likely taken at a temperature of 22 to 25C, the problem is that your far more likely to have an internal PSU enclosure temperature of 35 to 50C, especially if its the primary exhaust for the CPU Heatsink and case.

What we dont get to typically see is what is called a derating curve
Clearing The Fog That Surrounds Derating Curves @ powerdesign365 (registration required)

Derating is the negative slope of the power-versus-temperature graph. Specifically, it shows that as the operating ambient temperature increases, the converter's maximum output power drops to ensure reliable system operation. Derating curves provide a quick way to estimate the maximum output power of a converter at a given temperature
or simply given as
0ºC ~25ºC for full rating of load, decrease to zero Watts of power at 70ºC
(they even spin these figures as youll find if you follow the link)

so, as a rough estimate, deduct 1\3rd the rated amps per rail and start to compare that to your baselines, the surviors proceed to the next round
(following this tutorial there are some suggestions regarding lowering the temperature of the supply, both increasing its capacity and its longevity)

AC Input Voltage
another way manufacturers make their supplies look better is to rate them with a narrow range of acceptable AC Voltage, it much harder for the supply to maintain a stable reference voltage on the mail rails if the AC source power is fluctuating
if the VAC drops from 110V to 90V the supply must draw more amps from the socket to compensate, and doing that an maintaing a stable DC voltage isnt easy, thus its a real test of the voltage regulation, by limiting the range in the spec of the power supply they are fudging the figures on you, so its a sign that you might be dealing with a less reputable manufacturer, some examples
PCP&C 510.................... 90 to 264 VAC, 47 to 63Hz
Enermax EG565P............90 to 135V or 180 to 265V, 47 to 63Hz
Neopower........................100 to 240 VAC. 47 to 63Hz
POWmax Ultra-X.............103 to 132V or 206 to 264V, 47 to 63Hz
Raidmax 470...................100 to 120V or 200 to 240V, 50 to 60Hz
Codegen 500...................115 to 230V 50 to 60Hz

to be continued
next Load Regulation, Transient Response and Baseline reference voltages
 
:eek:

Well there's a hell of an answer :)

I'll get around to figuring that out a little later today... thanks!
 
Actually, was just messing around getting some ratings...

Since there are only spots for two HDDs on takaman's PSU Calculator, I found the seagate I have, took the amps rating for the 12v and multiplied it by three, since I have three drives:

2.8*3=8.4 and 100.8W ??? That's a TON on watts! Each drive seperately is rated at 12.5W max during seeking... what gives?
 
WELCOME TO THE FORUMS!!!

____

Personally I'd trust the calculator. As for a powersupply it would appear you need something heavy duty. Anything from Fortron or OCZ should do the job nicely.

What kinda price range are you in?
 
Well at least now I have someone else pushing the calculator :D

Nice post BTW
 
Sentential said:
WELCOME TO THE FORUMS!!!

____

Personally I'd trust the calculator. As for a powersupply it would appear you need something heavy duty. Anything from Fortron or OCZ should do the job nicely.

What kinda price range are you in?

I'd trust the calculator too ;) just wondered if I was putting the wrong number in the wrong place somehow.

My price range is dictated by what will make my computer work I suppose... I'm not sure what I'm getting into as far as price, mostly cause I don't really know what I need yet.

I suppose I'll know by tonight.

ps> Thanks for the welcome :)
 
Apparently my needs go something like this:

Total: 376.2W
3/5: 108.8W
3/5/12: 366.2W
3.3: 3.1A
5: 19.6A
12: 21.4A

Although I guess I could get a 400W PSU, not many of them seem to have 3, 5, and 12v rails with enough amperage to work. Looks like I'm stuck around 450 or 500W, especially after OCing.

Here's the link: http://takaman.jp/D/?M=PbQEcG5HLd@de@dSBgW5H3hG5DIHGT6UOUcn@rXcYaZBuG50aMZ&english

Can anyone recommend a good PSU? I understand that they're more efficient if they're running closer to max capacity, also the prices are not super-low... I guess I'm looking at around $100. Oh well, more money down the hole.

I try to look on the bright side by thinking I can always reuse something like a 500W PSU. It sorta works... sometimes :eh?:

Edit:

Do I need to make sure that I get a PSU with dual 12v rails? I've read around that it's the new ATX spec, but not all the PSUs I've looked at support it.
 
Last edited:
ninegoodthings said:
Actually, was just messing around getting some ratings...

Since there are only spots for two HDDs on takaman's PSU Calculator, I found the seagate I have, took the amps rating for the 12v and multiplied it by three, since I have three drives:

2.8*3=8.4 and 100.8W ??? That's a TON on watts! Each drive seperately is rated at 12.5W max during seeking... what gives?

spinup draw, its roughly 4 time the seek draw listed by seagate
they should also list a real spinup draw there
the nice thing about this calculator is you can use other feilds for figures, just select say the modem, other device not listed, and add in the value to the appropriate rail(s), it does do some rounding of course :p
(Ive added parts of values to different fields to get it more accurate, or you can just figure it yourself AxV=W, W/V=A, W/A=V)


Ive been pushing this calculator quite alot ummm...elsewhere :p

efficiency varies as pointed out, the sweet spot changes from supply to supply
Mike Chin at SPCR does a real good job of mapping the efficiency sweet spot ;)

for a serious overclocking supply Id look for exceptional load regulation and AC Ripple
top honors to the PC Power & Cooling Turbo Cool 510 (Express\SLi\Deluxe\Standard) the first 2 have the 24 pin main, the last 2 the 20 pin in stiock trim, you can cutom order whatever you like though

then the Topower P6 Series Supplies are highly regarded, such as the OCZ Powerstreams, which have a nice Topower Options package, OCZ claims 1% load regulation, but the P6 supplies are listed at 3%, think that might be more realistic given an average crossload (three pages in from there, is an animated gif of the OCZ-470, which is no longer made and might not be as "tight" as the current crop)

both of those are ATX12V v1.3 supplies (an OCZ 600 watt excepted)
ATX12V v2.0 supplies are the ones with dual rails
dual rails where adopted to isolate dynamic loads from sensetive components like the CPU, following in the footsteps of the EPS12V spec, the disadvantage of that is since the rails are isolated, amps not used on one are unemployable on the other, a well regulated single rail supply is thus more flexible

the Antec Neopower is a v2.0 as is the OCZ 600 watt, both exceed spec in load regulation
 
Ice Czar said:
Figure the Individaul Draw on each +12V Rail
If you have a modern board, or if you have a server\workstation
its very likely that you will be looking at PSUs with more than one +12V rail
so we need to look at how much draw there is on each of the rails
it breaks out like this for an ATX12V v2.0(1) supply
1. Add the +12V1 up
CPU, mobo, +12V power for PCI cards
2. Add the +12V2 up
everything else, including the Auxillary Power to the Video Card(s)
(this is where you pull up the Video Card Chart we used above)

Nice guide :)

I kinda skimmed over it, but wanted to point out that on some PSUs like the Fortron Blue Storm series (aka Sparkle THA/THN line) it is the 12v2 line being used for the CPU. I have confirmed this with Fortron themselves.

Another correction - the "Powmax Ultra-X" is not made by Leadman, the usual junky OEM supplier of Powmax. It and the equivalent Ultra units are made by Wintech, a company with far better build quality (though still not what I'd call top of the heap). Powmax is just reselling units from Ultra I believe.
 
I was looking at the OCZ Powerstream 420, which seems like a pretty forward looking PSU - has plenty of room to grow into on all rails as well as overall wattage. Plus it's built for BTX, which I guess is the next thing coming... or has arrived... or something. And it's got the exposed pots. I also read that it has no PFC... is this a major consideration?

Also I checked out the PC&C TurboCool 510, but the main differences between the two seem to be mainly based on opinion.
 
Oklahoma Wolf said:
Nice guide :)
"Powmax Ultra-X" is not made by Leadman,

DOH :p

sorry an old cut and paste
actually the 500 watt X-connect at least is made by Young Year Electronics


the Ultra X-connect model X-ULT500P (pdf)
is listed at Underwriters Laboratories as manufacturered by
Taiwan Young Year Electronics CO, Ltd

ninegoodthings said:
Also I checked out the PC&C TurboCool 510, but the main differences between the two seem to be mainly based on opinion.

actually I wouldnt say that
the PCP&C 510 is the superior supply
OCZ has been fairly elusive in backing up their claimed specs, and reputible testing hasnt born out their claims of 1% load regulation, their claimed AC Ripple as well might just be on the leads with the ferrite cores, not for the supply in whole
and finally in the flawed but indicative Maxiimum PC PSU review, they did some interesting torture tests, including dropping the VAC to 60 (which is one serious brownout) the 510 didnt even blink maintaining a 12.08V throughout
the OCZ shutdown

they are however both rated at 40C so your looking at apples to apples amp ratings ;)
 
Last edited:
Aspire is also using Young Year. The fact that some X-Connects are turning up YY also doesn't surprise me too much - both OEMs produce rather cheaply made stuff IMO. Now that I think about it, I think I remember seeing some early YY X-Connects. However, some are definitely Wintech based on visual clues from online pics (Wintech part numbers on main transformers, and the odd UL file number trace).

The PC P&C 510w and the OCZ units have more different than just opinion - build quality at PC P&C is better also (though I am not sure yet who the OEM is for the 510 - Zippy has been ruled out, but Win-Tact and Fortron are also possibilities). PC P&C units rarely fail - they know how to pick a good OEM with an excellent reputation. I'll put the Topower made OCZ units at CWT/Antec levels of build quality, but one can do better at least in that regard.
 
Now that I look at prices I can see why the PC P&C is the superior supply... unfortunately it's WELL outside of my price range.

I suppose I will be going with the PowerStream... now the question comes up, unfortunately about 420vs520. Is the extra 100w worth the extra $30? I almost think I should just go for it, so that I don't have to spend more in the future if I need more wattage. Plus it's got the five year warranty.
 
well I just read something that might have some bearing on everyones purchasing decisions

the boys over at trustedreviews got to go over to Germany and use a Fast Auto FA-828ATE load tester (very similar to what Intel employs in there cert program, but they use Chroma

and well in passing the mentioned....

Other additions to newer PSUs includes a 24-pin EPS style power connector, instead of the 20-in ATX power connector. This has been designed for the new Intel Pentium 4 processors, but has been a server and workstation standard for some time. Added to this is a new eight pin 12V connector, which will eventually replace the four pin 12V connector, but this is unlikely to happen for a few months yet. A six pin connector for PCI Express graphics cards should start appearing shortly as well. Some of the power supplies on test came supplied with this six pin connector as a small converter, which would work with pretty much any power supply.

the part in bold, I rushed right over to formfactor.org but there is no sign of the new spec, but it sounds very plausible to me, basically ATX12V is going to conform to Server System Infrastructure (SSI) EPS12V
so that might bear on everyones upgrade path (unless you have an EPS12V that is :p )
 
ninegoodthings said:
I also read that it has no PFC... is this a major consideration?
While I think PFC is an important feature it is not a major one to base a decision on. I think many people confuse PFC and effciency. What PFC does in a nutshell is make a complex (inductive) AC load appear as a simple (resistive) load. There is a lot of math and stuff that could be put here that will explain it more. The major benifit of PFC is to your utility company. It keeps the power companies mains from becoming polluted (voltage and phase fluctuations). Many people think PFC PSU's are cheaper to operate, not true. If anything it actually consumes more electricity but it is a negligible amount. It is possible that in the long run it could save you money. This would occur because for every power factored device we use we actually ease the stresses on the main grid. This in turn would lower the power companies cost and presumably they would pass the savings on as lower rates. Fat chance. Corrections welcomed.

Efficiency refers to how well a PSU converts AC voltage into those DC voltages. PSU efficiency is expressed in percentage. 68% means, for example, that 68% of the AC power is converted into DC power. So if a PSU draws 100W in AC power to produce a total of 68W in DC power, its efficiency is 68%. The remaining power, 32%, is lost as heat in the conversion process. This heat must be dissipated so that the PSU does not overheat.

High efficiency therfore means the PSU will run cooler and possibly save you some money. Running cooler as we know also extends the life of electronics and means less noise (less fans). It also means you will have to turn your space heater back on in the winter 'cause your folding rig is just runnig too cool now :)
 
Last edited:
BeerCan said:
High efficiency therfore means the PSU will run cooler and possibly save you some money. Running cooler as we know also extends the life of electronics and means less noise (less fans). It also means you will have to turn your space heater back on in the winter 'cause your folding rig is just runnig too cool now :)

Damn, so I guess the power gained by efficiency will be converted into power I need to use to heat my apartment ;)
 
BeerCan said:
While I think PFC is an important feature it is not a major one to base a decision on. I think many people confuse PFC and effciency. What PFC does in a nutshell is make a complex (inductive) AC load appear as a simple (resistive) load. There is a lot of math and stuff that could be put here that will explain it more. The major benifit of PFC is to your utility company. It keeps the power companies mains from becoming polluted (voltage and phase fluctuations). Many people think PFC PSU's are cheaper to operate, not true. If anything it actually consumes more electricity but it is a negligible amount. It is possible that in the long run it could save you money. This would occur because for every power factored device we use we actually ease the stresses on the main grid. This in turn would lower the power companies cost and presumably they would pass the savings on as lower rates. Fat chance. Corrections welcomed.

Efficiency refers to how well a PSU converts AC voltage into those DC voltages. PSU efficiency is expressed in percentage. 68% means, for example, that 68% of the AC power is converted into DC power. So if a PSU draws 100W in AC power to produce a total of 68W in DC power, its efficiency is 68%. The remaining power, 32%, is lost as heat in the conversion process. This heat must be dissipated so that the PSU does not overheat.

High efficiency therfore means the PSU will run cooler and possibly save you some money. Running cooler as we know also extends the life of electronics and means less noise (less fans). It also means you will have to turn your space heater back on in the winter 'cause your folding rig is just runnig too cool now :)


very true however it can impact your UPS scheme (sometimes greatly)


Power Supply System Integration Part 2: Battery Backup @ Power Electronics Technology

UPS Nonlinear Loads

Most end users are oblivious to the problems encountered when an online UPS interacts with nonlinear load, such as a switchmode power supply. Fig. 3, on page 42, illustrates the interface between an online UPS and a switchmode power supply. The switchmode supply within the host system consumes its current in high amplitude pulses, and as a result overloads the UPS, distorts its output and degrades in performance. This isn't the case if the SMPS has power factor correction (PFC); and in the United States most SMPS don't have PFC.

Repeated measurements under practical conditions show that in non-PFC SMPS, the current peak to rms ratio reaches a crest factor in the range of 2.5 to 3.5. This high crest factor imposes a difficult load for the UPS.

Manufacturers specify UPS capacity in volt-amperes, and set the UPS current limit with resistive loading. This affects the user in two ways: first, the actual power (in watts) that the online UPS can deliver is 75% to 80% of its VA rating. Second, the UPS cannot support a nonlinear load like a SMPS to full power capacity. In fact, it can support an SMPS only rated to 30% to 40% of its capacity. A 1kVA online UPS will only be suitable for backing up a system energized by a 300W to 400W (non-PFC) switchmode power supply. Attempts to draw higher power by the SMPS will severely distort the output of the UPS and may result in shutdown. Some UPSs support nonlinear loads to various degrees, but the above rule is safe to use as a guideline.

The whole idea of generating a high-purity sinewave within the UPS, and then rectifying it within the SMPS is wasteful. A UPS may be deployed to run non-electronic loads, such as lights or motors. However, this is a minor portion of the market for online UPS. Many applications back up computers and telecom devices, as well as instrumentation systems used for critical applications. In all of these applications, the immediate load for the UPS is the SMPS within the systems. Peak current and inrush current are associated with this load at startup. This current of 40A to 80A per SMPS will cause the UPS to shutdown due to overloading, or the UPS must transfer the load to utility to draw its inrush.

Im "upgrading" the importance of active PFC in purchasing, as it does seem to have a great deal to do with not abusing and effectively employing the capacity of an online UPS (and by extention Id assume a line interactive UPS \ SPS when its on batteries, a far more common UPS scheme)

and lastly, I salvaged a Ferrups, and Ferroresonant Transformers, dont play well at all with non-PFC power supplies, it was oscilating and crashing :eek:

a bit more on PFC
PFC Decoded
and on UPSs
Power Conditioning 101
DIY UPS all @ dansdata ;)
 
Last edited:
Well actually that makes great sense. Because with a UPS if you are using a non-pfc PSU then you are in effect polluting its "mains" instead of the power companies. I did not really take that into account thanks for pointing it out. As for the ferrups, here is what I think you should do with that ;) http://igor.chudov.com/tmp/ferrups-meat-smoker/ Not really a big fan of those for home use anyway. I know about that because my other great love is good bbq
 
LOL

think Ill stick with Mesquite :p
Yes i have an old 1.4 KVA Best Power Ferrups, but the price was right
free
not sure I qualify as a home user, even if I am at home
run a 20U rackmount 2x4U 2x2U 3xDot Hill SANnets

last time I had the Ferrups online it was crashing an Antec PP403X
so quite awhile ago, I researched it and found the basic issue and took it offline, had plans to employ it elsewhere but after discussing the issue with an EE on my home forum Ive decided to trial it again with 4 of the PFC supplies I have, starting with an Ablecom EPS12V 2U 400 watt running an old Celeron as an IDS (overkill I know, but its a 2U and I didnt have anything else :p ), and working my way up to my NAS and Workstation if all goes well
 
Last edited:
Back