• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

Power Supply Problem?

Elec

[H]ard|Gawd
Joined
Jul 20, 2000
Messages
1,944
Hey guys, I think my power supply is biting the dust, but just wanted some other opinions and info. It's an Antec TruePower330 that I've had maybe a year and a half. It's never given me any problems until a month or so ago. Anyway, when I turn on my computer, it POSTs fine but my hard drive clicks a couple times and then the computer freezes. If I wait a minute and then reset, it boots up fine and runs all day long without a hitch. At first I thought this was an HDD problem, but my drive passes SMART and WD's data lifeguard tools and somebody in another hard disk thread said it sounded like a similar PSU problem he'd had.

So I go and check my voltages, and my +5v looks way low. At POST, my BIOS is reporting a measly 4.19v on that rail. It seems to be around 4.5 when it successfully boots, and runs from 4.65-4.75v while the computer is in use according to MBM (haven't busted out the multimeter yet). It seems like the PSU is not providing enough voltage until "warmed up". Anyway, here's what I have in the box:

Epox 8RDA+
Barton 2500+ @ 3000+
2x256 Geil PC3200 Ultra
1x512 Geil PC3200 Ultra
Radeon 9700np
TDK 48x CD burner
Hitachi 2x DVD-ROM (circa 1998, woot!)
WD 2500JB drive (8meg, 7200rpm, 250gigs)
WD 400(not sure if it's a BB or JB, currently disconnected for testing purposes)
Audigy2
D-Link wireless card
crappy modem
Sunbeam Rheobus controlling 2x80mm LED fans, 2x80mm crappy case fans, 1x80mm tornado, 1x60mm exhaust fan
1x80mm Antec thermal controlled fan
12" blue cold cathode (on a switch and usually off)

So I guess I'm looking for two things. First, is my PSU borked? Rail sag, on it's last legs, overtaxed by my components, etc.? And second, if that's the case, what would be a good replacement? I've read the sticky, but I can't decide whether it'd be a better idea to just get something that works with what I have now or something that's a bit more future proof. I'm not planning on building a new box or any real significant upgrades for at least 6 months. My current mobo has a 20pin power connector and that's it (no extra 4 pin or 6 pin or any of that), got no SATA drives, no SLI, nothing fancy really. As an estimate from the takaman page in the sticky, I need 3amps on +3.3, 41.1amps on +5 (which seems ridiculous), and 8.9amps on +12 with 332W minimum. I'd also rather spend $100 tops if I do need a new PSU, with less being better. Thanks!
 
My computer will boot and log into windows then restart itself on this infinate loop. I'm thinking either Mobo or PSU. Is the PSU crapping out on me. I've had it for about 6 months. It's a no-name came with the case PSU...any thoughts?

Before you post telling me to post specs....

Specs:
P4 2.4A (Prescott)
ASUS p4p800e- deluxe
1 G Kinston Value Ram DDR 400 (Slot 0 and 2) ( I I I I )
120g WD (SATA)
80g Maxtor (SATA)
CD-Rom 50x (IDE)
3 x 80mm case fans
2 x 4in CC*
2 x LED's *

*currently not hooked up for testing.
 
ClawhammerUK said:
Only way to be sure would be to use a PSU that you know works and see it it still happens


While that is true, it's not exactly the happiest or most helpful thing in the world.

Me: "Doctor...am I going to need a heart transplant?!?"
Doc: "Well...I'm not really sure, but I think we'll slap one in to find out!"

I don't really have a lot of money to throw at this problem and I'm not about to run out for a $50-100 PSU unless that's the most likely problem/solution. The only other PSU I have lying around is an old 180 or 200W unit. I suppose I could unplug everything but the mobo and one hard drive and see if it'll power up properly.
 
Ok, got the multimeter out today. My BIOS is reading a little low, meter says the +5v rail was actually about 4.25v at POST on a cold boot. That's still 15% low. After about a minute or two, the voltage ticked up to around 4.5v at which point I was able to successfully boot into Windows.

So what's the deal? Nobody wants to weigh in on this? Is the supply going bad? Am I overtaxing it with my components? Is it possible I have a bad component which is screwing with the PSU? I'm not opposed to buying a new one, I just want a reasonable consensus that something is wrong with the one I have first. As I mentioned, after it "warms up" it'll run fine all day and night, so I'm not desperate for a solution, but I also don't really want to turn it on one day and have the PSU finally go (and take the rest of my gear with it).
 
I give up. All I've gotten is someone saying I'll know if I need to buy a new power supply by trying a power supply that works (which makes sense, but if I had a working power supply, I'd be USING it), and someone else trying to jack my thread.

So I guess I'll save myself a hundred bucks and continue using this PSU until it either stops working or suffers a catastrophic thermonuclear meltdown.
 
im going to assume that youve already gone through the testing phase where you remove all the non essential components from your pc and put them back in one by one to see where it starts to fail. if you havent, then thats where id start.


next, id call up a friend and have them bring their pc over so you could swap some parts to test the suspect psu... if thats not an option, i dont know what to tell you.

if you check the egg, i believe they have a couple of sparkle 350w psus for a reasonable price. i think $30 for one and $40 for the silent 12cm fan version.

to tell you the truth, im suprised you didnt get a "OMGOOSE YUO BNEEDS A PCPNC 510 ULTRA KICKY DELUEX PSU FOR TEH FUTUR PROFFING!!!11!!!"
 
I love you. Mainly because A) you bothered to reply, and B) you provided helpful information. So thank you!

The strange thing about the problem is it only seems to present itself when the PSU and/or system is cold. So I can either test changes first thing in the morning, or turn the box off and let it sit a while. Unplugging/removing several components didn't seem to make a difference in the voltage reading from the +5v rail, and it was actually going down as I tried a couple different things (from about 4.55 to 4.4 to 4.32). The box was mostly off as I was plugging and unplugging, so I assume things were cooling down.

I have also determined that for testing purposes at least, my old ACBel power supply will run the motherboard, RAM, CPU, vid card, and system HDD. I had no problems booting with this and the BIOS reported a +5v of about 4.75 and a +12v of 12.4. Will try it again tomorrow when the box is cold to see what happens. For giggles, this PSU is a hoss, check out these specs: 100W MAX on +5v and +3.3v combined, 15.0A on +5v, 12.0A on +3.3v, and a massive 3.0A on +12v :)

Anyway, will try both PSUs tomorrow with a minimal amount of gear connected (ideally the same for both units) to try to determine whether it really is a PSU issue and whether it does involve my Antec "warming up" as it seems to be.
 
RMA the PSU. 3 year warranty so you might as well use it. The +5V output should be within 3% of 5, so 4.85 to 5.15 (that's because it's a TruePower PSU, otherwise it's 4.75 and 5.25 for 5%).

1-800-222-6832 or customersupport@antec.com

AntecRep
 
Wow, figured the warranty had already lapsed, but thanks for the info! I'm not positive I can dig up proof of (date of) purchase though, as I bought it around a year and a half ago. I assume that'll be a problem?
 
personally Id ditch that Rheobus and see what develops

I'll run some numbers but fan controls have caused issues in the past

I kind of doubt the supply itself is toast unless its been leading a hard life
(elevated temperatures, high load, EMI feedback)

you also need to take software reading with a grain of salt
they are good relative indicators so if its dropped over a period of time thats one thing
or if observed or logged fluctuation exceeds the spec for the supply in its total range

but for definative values you really cant depend on software and need to take a direct reading with a Digital Multimeter

thus all your figures could be high or low by some unknown amount
that said when they are off by a great value its generally a glitch endemic to a given mobo and youd see other owners reporting that or you have a real problem
 
http://takaman.jp/D/?M=PaQKcG5HGdHfe@dSkTZiH3UcXaYbZAZavG5HBMZ&english

+3.3V @ 3A
+5V @ 39.3A
+12V @ 10.9

+3.3V + 5V combined max at 206.9W
total combined 3.3\5\12 @ 337.9W
total 347.9W

your overtaxing that supply your mobo is powering the CPU off the +5V rail as it lacks a 4 pin auxillary +12V mobo connector
while the totals represented above are theoretical maximums and are additive
thus never actually occur, the supply also has to be derated for temperature

yet its only rated at
30A +5V
28A +3.3V
22A +12V

with a combined capacity of 310W

its pretty safe to say your overtaxing the supply on the +5V rail
at least part of the time, and that while its been able to deal with that in the past you may have been degrading its ability to
alternately its simply a little hotter now say a nice layer of dust in the heatsinks to insulate against themal transfer and its not up to it any more

a different mobo that employs the +12V to power the CPU and it would likely be a different story
I think your right its time for an upgrade somewhere and maybe a better thermal solution than a single 80mm case fan

strangly with that board what you really want is a dinosaur a v1.2 supply
though a v1.3 would do with enough capacity,
but keeping it "lean" is tough because the capacity has shifted,
thus to get the capacity on the rails you need your going up in total wattage
but not using that capacity on the +12V rail

ATX.gif

Chart by Xbit \ Oleg Artamonov

Id recommend the FSP 530
http://www.newegg.com/app/ViewProductDesc.asp?description=17-104-968&depa=1
for the following reasons
1. its has a real hansom +3.3V\5V crossload that will easily handle that config
+3.3V@30/17A, +5V@44/24A, +12V@16/28A (for your config read just the first value X/Z)
2. its in the price range
$79
3. its a well known workhorse and has some upward potential
(in the event you get a board that powers off the +12V eventually)


the ranges given for the amps reflect crossloading (hit that link above and checkout Oleg's article at Xbit)
most supplies have a single +3.3V\+5V rail, your true power is a truely independently regulated supply and isnt like that, normally thats an advatage, its likely an advantage anyway even though its regulated independently its still likely able to tap amps from the +3.3V capacity which is how youve been running your config

AntecRep might shed some light on that ;)

you can go leaner than that, but if it was my money thats what Id do
(and add a few more case fans :p )
 
ColinR said:
Manufacturing date stamp not enough?

That's a good point. If it was manufactured in say, 2003, the three year warranty ain't up yet :)


Ice Czar: Sorry if I wasn't more clear, but a digital multimeter measurement was taken on a cold boot and my +5v rail was reading about 4.25v. That's higher than the 4.19v reported by BIOS but still way low. +12v read very close to that, 11.97 I think. As for the fans, I currently have 2x80mm intake and 1x80mm and 1x60mm fans exhausting, in addition to the fans in the PSU. It's pretty much the maximum possible with my case without cutting big holes in it. Also, every indication is that the power supply is providing MORE voltage on the +5v rail as it warms up. If it's been sitting overnight, BIOS reads 4.19v. Right now it's reading 4.68v. If I were to turn it off now and then back on again in a minute, it'd likely read in the 4.50v range at POST (as it did when I was doing some testing). Once I have turned the box on and let it sit 2 or 3 minutes (while failing to boot), I can reboot and it'll run stable all day and night. I assume this is not a permanent condition, but it is how it goes right now.

I don't really plan on any significant upgrades or additions to this box (maybe something like replacing the DVD-ROM with a burner), so I mainly just need it to work until I can afford to build a new one (which may be 6 months to a year). Unfortunately it seems that new systems are going to 12v uber alles so it's probably unlikely that any supply I'd buy now would see a future use beyond powering my current config. I'll check out that Fortron you linked. It seems kind of goofy in a way to be looking at a 530W supply for what's in my box - I'm not even sure supplies that big existed when I bought the Antec!
 
LOL your likely right
a quality oldschool 300 > 350W v1.2 would likely do it
just hard to find these days

Id look at some of the other FSPs available
but alot of capacity shifted off the +3.3\+5V on many of the supplies that used to be available in class and you do have alot of amps on the +5V
 
Elec said:
That's a good point. If it was manufactured in say, 2003, the three year warranty ain't up yet :)

Except the warranty is only good for the original customer, which a date stamp can't prove. :(
 
Yeah, I'm definitely the original owner, and my whole system isn't even 2 years old, but I don't think I can prove it. Most likely I ran across the receipt a while back and did a "Hmm...well I've had this power supply a year now and it works, so I probably don't need this anymore!" Besides which, if I really am overtaxing the PSU on the 5v rail, I don't think getting an exact-item replacement under warranty would be a great solution to the problem. It might fix me up for a while, but I'd likely have trouble again down the road.

Ice Czar: How do you feel this PSU http://www.newegg.com/app/viewproductdesc.asp?description=17-104-154&DEPA=0 compares to the Fortron you linked earlier? The rails appear to be pretty similar with a big +5v and I kind of like the 120mm fan and modular feature. It has a lower overall wattage rating though (450 vs. 530) and while I know FSP makes good stuff, I don't know much about OCZ. I see in the sticky there are a few Powerstreams recommended but no Modstreams. This particular PSU is only a buck more than the Fortron, which means it's also still in the price range :) I looked through the other supplies from FSP at the egg and the only cheaper one that might get it done is a 400W with 28A on the 3.3 and 40A on the 5v, which is probably cutting it too close.
 
LOL

also a Topower like the powerstreams
a good supply that exceeds spec in many regards
Id have no reservations recommending it,
jonnyGURU seems to think they are actually better than the Powerstreams

http://www.hardforum.com/showthread.php?t=889982

as far as total capacity of watts
it might be equal to the FSP, its rated at 40C
whereas the FSP's rating is unknown but certainly higher than what most of the industry employs
(often the Intel certification ambient of 21C)
example > http://www.powmax.com/PDF/LP-6100C_pass.PDF
 
LOL No worries, the Antec is still running things fine (aside from the new startup routine I've had to adjust to), so I'm not in a SUPER hurry to place an order, just kinda checking out my options :)
 
seems the Powerstream isnt all its supposed to be in the +5V rail
http://www.hardforum.com/showpost.php?p=1027530441&postcount=19

which is news to me as it did well at xbit
(edit DOH confused it with the 470)
http://www.xbitlabs.com/articles/other/display/powerful-psu_9.html
but that 520 does break up at the top end of the +5V\+3.3V load
ahh thats whats hes talking about
http://www.xbitlabs.com/articles/other/display/powerful-psu_11.html

hes tested a modstream recently
http://www.slcentral.com/raidmax-rx-520xp-psu/page4.php
(his test rig not the modstream)


here is the modstream
http://www.slcentral.com/ocz-modstream-520w-psu/
would be nice if hed give you an idea of what that heavy 34>40A +5V load might be like on it ;)
 
Ok... Ok... I'm here. :)

NEVER use the PowerStream in a PC that's got a load on the 5V (in Amps) greater than that of the 12V. All through the spectrum, if the load on the 5V was greater than 20A AND greater than the 12V (at lower than 20A on the 5V, the problem wasn't there) the voltages of ALL of the rails would vary wildly (watch my video. That was at 20A on the 5V, 10A on the 12V.)

It seems that X-bit experienced the same thing and graphed it quite nicely. When representatives of OCZ were asked about the issue, they simply said it was normal and that I should use a ModStream if I'm going to run a system that may load the 5V greater than the 12V.

Ironically, X-bit didn't exhibit the same problem when the 5V is loaded higher than the 12V, and since you're only getting 2A more on the 12V with the 520ADJ, I'd say a 420ADJ is actually a more "well rounded" power supply (good performance:cost ratio, for sure.)

Unfortunately, I have no experience with the 450W. It seems to be based on a different platform, though. Perhaps the same platform as the ModStream. If that's the case, I'd say that's a winner as well. ;)
 
547GG.gif

bakward compatible to ATX12V V1.2
red crossload profile below
(actual ATX12V 1.3 hybrid)

ocz2.gif

bakward compatible to ATX12V V1.3
green crossload profile below
(actual ATX12V 1.3 hybrid)

ATX.gif

Charts & Graphs by Xbit \ Oleg Artamonov

(assumed similar crossload profile for the higher wattage class)
 
LOL The one I'm interested in IS the ModStream though. I only mentioned the Powerstream because they're recommended in the sticky, but no mention of the ModStreams. The takaman calculator seems to indicate I've got a ginormous 5v load (in the neighborhood of 40A) but not much at all on the 3.3 and 12v. Newegg has the 450W modstream for 80 bucks and the 520W you reviewed for $96. SO I guess my questions are:

1) Are these good power supplies? (The review clearly shows you liked the performance of the 520W)

2) Would they meet my power requirements as discussed on the first page (in comparison to say, that 530W Fortron Ice Czar first linked)

3) Is there any real reason I should spend the extra 16 bucks on the 520W modstream over the 450W unit? It's only 16 bucks, but at the same time I don't have a lot of extra cash right now and I'm not sure if it'd be headroom/overkill :)
 
Im pretty confident of that Fortron and would defer to jonny in comparision to the modstream

the Fortron is an older design and Im sure its backward compatible to ATX12V v1.2

+3.3V@30/17A,+5V@44/24A,-5V@0.3A,+12V@16/28A
(v1.2 figures in bold)
 
The Modstream suffers the same marketing misconception as the PowerStream when it comes to the 520W. Just like the PowerStream, more than 70% of that extra "wattage" is on your 5V rail. There's only a 2A improvement on the 12V rail! And the 5V on the 450W is ALREADY at 45A... So I'd say IF you were considering an OCZ Modstream GET THE 450W!!!!!
 
Ok, well it looks like I'll probably pick up either the Fortron or the Modstream. Not entirely sure which one yet, gotta do a little more research and make sure there's not anything else I'm overlooking.

Thank you guys so much for your help! I'm doubting I would have gotten a clear idea of what my PSU issue was without it.
 
jonnyGURU said:
When representatives of OCZ were asked about the issue, they simply said it was normal and that I should use a ModStream if I'm going to run a system that may load the 5V greater than the 12V.

I don't think I totally buy that... by design, independant voltage regulation technology shouldn't exhibit that behaviour. And, as Oleg showed, the 470w doesn't, nor does the Antec Truepower he tested (except a couple odd blips when the 5v is low and 12v is between 125w and 150w). Still, the 520w is a good unit, and there are few circumstances that would have it run into this problem. My system might give it some trouble, but there aren't that many 5v oriented nForce2 IGP boards out there.
 
Oklahoma Wolf said:
I don't think I totally buy that... by design, independant voltage regulation technology shouldn't exhibit that behaviour. And, as Oleg showed, the 470w doesn't, nor does the Antec Truepower he tested (except a couple odd blips when the 5v is low and 12v is between 125w and 150w). Still, the 520w is a good unit, and there are few circumstances that would have it run into this problem. My system might give it some trouble, but there aren't that many 5v oriented nForce2 IGP boards out there.

I don't buy it either, but that's their excuse so if that's what they want to use, I'll make sure to apply that quote to the "review." Their claiming that the inability to load the 5V greater than the 12V regardless of what that 5V rail is actually regulated at completely negates the advantage of independent voltage rails. And people wonder why I like my Ultra 600W. :rolleyes:

As for not many 5V oriented nForce2 IGP boards: The Albatron, Biostar, Asus, Shuttle and Abit motherboards all regulate CPU voltage off of the 5V. The Epox, MSI and ECS regulate off of the 12V. I don't know all of the nForce2 boards out there, but I'm fairly certain that most actually do 5V CPU regulation.
 
jonnyGURU said:
As for not many 5V oriented nForce2 IGP boards: The Albatron, Biostar, Asus, Shuttle and Abit motherboards all regulate CPU voltage off of the 5V. The Epox, MSI and ECS regulate off of the 12V. I don't know all of the nForce2 boards out there, but I'm fairly certain that most actually do 5V CPU regulation.
Basically anything without the square 4-pin +12V connection, right? My Asus board doesn't have it. Most Abit NF7/AN7 boards do.
 
Yeah, that's a good question. My board is an Epox 8RDA+ (which is nforce2 ultra) and doesn't have the extra 4 pin connector. I've been assuming I need a new PSU with a really beefy +5v, based on the low +5v rail readings and the fact that the takaman calculator says so.
 
_Korruption_ said:
Basically anything without the square 4-pin +12V connection, right? My Asus board doesn't have it. Most Abit NF7/AN7 boards do.

Yeah. The 2x2 connector is the indicator. The single 12V lead on a typical 20-pin connector simply doesn't allow for enough juice to power a CPU, fan headers and any 12V any of your slots may need.


Elec said:
Yeah, that's a good question. My board is an Epox 8RDA+ (which is nforce2 ultra) and doesn't have the extra 4 pin connector. I've been assuming I need a new PSU with a really beefy +5v, based on the low +5v rail readings and the fact that the takaman calculator says so.

8RDA+ does not have the 2x2. The 8RGA+ does.
 
jonnyGURU said:
As for not many 5V oriented nForce2 IGP boards: The Albatron, Biostar, Asus, Shuttle and Abit motherboards all regulate CPU voltage off of the 5V. The Epox, MSI and ECS regulate off of the 12V. I don't know all of the nForce2 boards out there, but I'm fairly certain that most actually do 5V CPU regulation.

Didn't realize there were that many IGP boards still out there ;)

Odd thing is, the mATX version of Aopen's IGP board had the 12v connector, yet my full size Max version didn't. In fact, when mine came out, I think it was the only Aopen nForce2 board to run the CPU off 5v. It was also the only one to not support BIOS CPU multipliers, but my Barton's locked anyway so I do all my overclocking on FSB.
 
Oklahoma Wolf said:
Didn't realize there were that many IGP boards still out there ;)

Odd thing is, the mATX version of Aopen's IGP board had the 12v connector, yet my full size Max version didn't. In fact, when mine came out, I think it was the only Aopen nForce2 board to run the CPU off 5v. It was also the only one to not support BIOS CPU multipliers, but my Barton's locked anyway so I do all my overclocking on FSB.

AOpen is weird that way. It's like they have three versions of every board. I bet they had a version of both with and without

It's like I use those KM400 based AOpen boards for a bunch of builds. I forget the model, but it's got the VIA SATA RAID and RAID BIOS. One shipment, my supplier replaced that board with another one. Still had the KM400 chipset. Still had VIA RAID, but a different model number with no RAID BIOS. I was like, "WTF?!?!" Weird.

Speaking of AOpen and power supplies, for you system builders out there: I've been told that the mid-tower series has had their FSP 300W replaced with an FSP 350W with more juice on the 12V rail. Unfortunately, AOpen hasn't updated their website and I haven't gotten any new cases in yet so I don't have any exact specs on the PSU, but the cases are still around the same cost, so it's still a really good value. :D
 
Gonna be buying soon, so just to clarify, the two PSUs I'm looking at are the OCZ Modstream 450W and the Fortron-Source FSP530-60GNA.

To jonnyGURU, it looks like you reviewed the FSP and had it blow up! That's not particularly confidence inspiring, although all the reviews on Newegg talk about how great it is and it got Ice Czar's seal of approval. It sounds like you got a bad unit but still gave it a speculative 7/10. You also reviewed the 520W Modstream (slightly bigger than the 450 I'm considering) and gave it a 9.5. Any reason I should pick one over the other? Ice Czar calculated that I need something like 3A on +3.3, 40A on +5, and 11A on +12 and 350W min so that seems to be within spec for both supplies. I don't have any SATA or PCI-Express devices and likely won't for this system as the motherboard does not support them.
 
The FSP530-60GNA is probably the most failure prone unit I've ever seen from Fortron. However, this only amounts to about one case a month I hear of over about 5 message boards, and it's pretty popular still. Most of the time for the other models, it's rare to see a Fortron fail at all.

I have one of these units - it has been awesome ever since I bought it, and I won't be getting rid of it. I didn't like the way that intake fan pressed up against the heatsinks, it tended to make the unit run hotter than necessary, so I moved it to the other side of the PSU casing. Made it almost too long to easily fit in my case, but it made a big difference in operating temp and PSU fan speeds.
 
Elec said:
To jonnyGURU.. Any reason I should pick one over the other?


Nope. The research is there for you to digest. ;)
 
jonnyGURU said:
Nope. The research is there for you to digest. ;)

ROFL Can't quite decide whether that means "POOF vs. works fine, think about it dummy!" or "Both pretty good, just pick the one you like best!" But anyway, uh...thanks :D
 
Back
Top