I have read this update on the PSU stress test on Tom's Hardware Guide, and i am pretty anxious about what they conclude: http://www.tomshardware.com/howto/20050 ... st-16.html
I know that a lot of the low-noise freaks on this site as myself have looked towards the Antec Phantom 500 as a serious PSU to handle the power deliveries without (or minimal) noise. But if the price to pay is instability, og even worse a wrecked Windows-installation, is it worth to pay that price?
Therefore my question is: How many of you guys have had the ripple-problems seen in practice, i.e. have any of you had any instability or unprovoked crashes / reboots?
Phantom 500 - ripple problems?!
Moderators: NeilBlanchard, Ralf Hutter, sthayashi, Lawrence Lee, Devonavar
80mv is such a little variation I doubt it's a problem in real use. Same goes for Antec's other power supply they tested. This is technically out of spec but is unlikely to cause instability except in the most overclocked on the very edge of being stable computers. Besides, most people don't buy the phantom to run it at 500W or even close to that amount. People buy it so it can be ran fanlessly and to do that you have to have a system which draws less than half of that amount.
If that ripple occurs only at full load, then I'd call it a non-issue. It would be nice if Tom's would give more details, like at what wattage the ripple crossed the 50mv threshold, or even if the ripple was continuous or transient, then we would know whether or not to even worry about it.
In "real" use, you're never likely to get loading above 300 total watts. Maybe 400 if you were really trying: OC'd CPU, a pair of high-end VGA cards, and a bunch of drives. But such a system would also never come close to loading 30A on the 3.3v line. (I can't imagine any system loading 30A on the 3.3 line, but I suppose it's theoretically possible)
And the problem with polling for user input is how would you know if the ripple is what caused the reboot? If the ripple only occurs at 500W load, then it's likely that only people like MikeC and Tom's will ever experience it.
In "real" use, you're never likely to get loading above 300 total watts. Maybe 400 if you were really trying: OC'd CPU, a pair of high-end VGA cards, and a bunch of drives. But such a system would also never come close to loading 30A on the 3.3v line. (I can't imagine any system loading 30A on the 3.3 line, but I suppose it's theoretically possible)
And the problem with polling for user input is how would you know if the ripple is what caused the reboot? If the ripple only occurs at 500W load, then it's likely that only people like MikeC and Tom's will ever experience it.
The problems i have seen mentioned in the "Phantom - have you had problems with it"-thread somehow have worried me, as i'm not so sure whether or not i should go for this PSU. I can clearly see your points, but i'm not sure how to relate to the enormeous information about specifically the Phantom-models, and the fact that i have rarely seen so much information that either shows good or bad behavior from a PSU. Normally i mostly see people telling they are happy about their well made PSU, but with this series it's not the case.Maybe I should just go for at try. I'm really counting on it be a important piece of hardware in my new rig based on minimal noise.
This is really interesting... I've outlined some of the issues with my system in this thread: http://forums.silentpcreview.com/viewto ... 335#201335
I really don't know if the PSU is to blame, which is why I was looking into the Seasonics... but my problems all occur during full load in hardware intensive games and during boot.
I really don't know if the PSU is to blame, which is why I was looking into the Seasonics... but my problems all occur during full load in hardware intensive games and during boot.
It's nice to see a review that examines things in more depth, and even though I can't see drawing that amount of current on the +3.3V rail, if (hypothetically) the ripple goes above 50mV at current draw of 25A, the Antec should have rated the supply for 25A current on the +3.3V rail. However, given the lack of information about what actually occured from tomshardware we're just guessing. I mean for all we know, the tested it with a load of 0.2A on the +3.3V rail, which is below the minimun load Antec state for that rail (and a quick check of the ATX specification shows a minimum load of 0.5A load on that rail)Rusty075 wrote:If that ripple occurs only at full load, then I'd call it a non-issue. It would be nice if Tom's would give more details, like at what wattage the ripple crossed the 50mv threshold, or even if the ripple was continuous or transient, then we would know whether or not to even worry about it.
In "real" use, you're never likely to get loading above 300 total watts. Maybe 400 if you were really trying: OC'd CPU, a pair of high-end VGA cards, and a bunch of drives. But such a system would also never come close to loading 30A on the 3.3v line. (I can't imagine any system loading 30A on the 3.3 line, but I suppose it's theoretically possible)
And the problem with polling for user input is how would you know if the ripple is what caused the reboot? If the ripple only occurs at 500W load, then it's likely that only people like MikeC and Tom's will ever experience it.
I guess we'll have to ask MikeC to improve the SPCR PSU test procedure to check for ripple as well.
As the other PSU's in the stress test have been set to draw a current of around 26 A it's likely to guess that the Phantom 500 has as well. It would be in the normal matter for Tom's to test under different scenarios for each PSU.|Romeo| wrote:It's nice to see a review that examines things in more depth, and even though I can't see drawing that amount of current on the +3.3V rail, if (hypothetically) the ripple goes above 50mV at current draw of 25A, the Antec should have rated the supply for 25A current on the +3.3V rail. However, given the lack of information about what actually occured from tomshardware we're just guessing. I mean for all we know, the tested it with a load of 0.2A on the +3.3V rail, which is below the minimun load Antec state for that rail (and a quick check of the ATX specification shows a minimum load of 0.5A load on that rail)Rusty075 wrote:If that ripple occurs only at full load, then I'd call it a non-issue. It would be nice if Tom's would give more details, like at what wattage the ripple crossed the 50mv threshold, or even if the ripple was continuous or transient, then we would know whether or not to even worry about it.
In "real" use, you're never likely to get loading above 300 total watts. Maybe 400 if you were really trying: OC'd CPU, a pair of high-end VGA cards, and a bunch of drives. But such a system would also never come close to loading 30A on the 3.3v line. (I can't imagine any system loading 30A on the 3.3 line, but I suppose it's theoretically possible)
And the problem with polling for user input is how would you know if the ripple is what caused the reboot? If the ripple only occurs at 500W load, then it's likely that only people like MikeC and Tom's will ever experience it.
I guess we'll have to ask MikeC to improve the SPCR PSU test procedure to check for ripple as well.
As for the poll i'm not asking for ripple problems as an isolated factor, but for instability, crashes or reboots in every way. If someone has bought this unit a are beginning to experience that they should of course take the poll.
They could have done something very very interesting. Instead, that is another "useless Tom's review".
I expect to read something "more solid" from xbit-labs.
My choice is between Phantom 500 and Seasonic 500. I'd have chosen Antec for the astonishing quality and superior voltage stability, but after reading all this I am a bit suspicious..
I expect to read something "more solid" from xbit-labs.
My choice is between Phantom 500 and Seasonic 500. I'd have chosen Antec for the astonishing quality and superior voltage stability, but after reading all this I am a bit suspicious..