No its not. Those are less than +/- 1% off for each RAIL. Considering that the caps/resistors/inductors/etc used in computer parts have a higher tolerance than that its DAMN good.
Marc, neither you or I know if those lines are under load or not. Quality on a PSU can only and SHOULD only be reviewed under load conditions. Overclock and load at the same time isn't the same thing as stock and idle/load. Throw a pelt on the 12V line on the psu and see what happens. (even if the pelt uses amps within the actual psu's specs)
I can grab a snapshot of fantastic voltagelines if i want or just stare in the bios if i want but load the stuff for a while and MOST Antec's will change a lot.
P4 2.6C @ 3.1 right now. 4 7200rpm drives, 1 5400rpm drive, a 10k raptor, a radeon 9800pro, 2 optical drives and 7 80mm case fans.
those voltages are at 100% usage with 2 instances of f@h running.
I'm surprised no one has asked this yet:
Those voltages off the bios, mbm5 (or similar), or did you use a multimeter?
Multimeter is better, yes, but within 1% should be not ACTUALLY out of spec unless the MBM specs for the mobo are off. One way to check MBM versus a decent other guess at reality is to see if when you reboot and go into BIOS the BIOS voltages shown are WAY different by 4-7% or more then you might want to multimeter under no load and then under full load and see what the differences are as a percentage of variance between the two measurements. If the voltages were 3-5% out in MBM, would definitely multimeter, or if other problems with box led me to suspect things were off, then would multimeter also. But so far do not see that here in this thread, so would provisionally high pass the box's power supply absent something that has not yet been discussed coming into the thread.
Quick way to say this-- if those voltages are present and no other issues, then accept MBM as good within a couple percent, and either way the possible values are well within spec still. If other weird issues, multimeter the legs where connected to what you suspect is maybe weird....
John D.
0
Straight_ManGeeky, in my own wayNaples, FLIcrontian
I am sure they do but they shouldnt. A good PSU should keep the Rails at their nominal values when under load or idle.
Well, most PSUs can also leg-to-leg balance and do when any one leg is overloaded by the nature of how a PSU works. AC to DC preconversion typically is done with one circuit, so there is one leg coming into the stepdown conversions for each leg. think of it as a big river of power flow from which each leg branches off as a smaller stream. Because of the way the branchng can be done, you can have three way split, a two way to 5 and 12 volt legs, OR additionally have the 3.3 leg coming as a branch off the 5 volt leg or the 12 volt leg.
If you have the latter, a secondary branching, and the first conversion leg's load is too high, it can toss the secondary legs low as the primary leg may go to devices before the branching for secondary legs. Limit for overall is by AC->DC conversion circuit's limit and and is reduced by efficiency factor. Multilegged PSUs need not be three monolithic single leg conversions off of one common source, they can subbranch in many ways.
Without being an engineer, it is easiest to load one leg then see the ampacity remaining for the other legs with something visual. Multimeter in right place is empirical, better than taking input from a sensor set that might not be in the right place(s) to show a defect (or defects). Multimeter from amny places under different loads, done carefully, lets you map the PSUs actuality reality and objectively becasue it cuts out the problems of sensor not in right place or sensor not reading right, or software not interpreting sensor right.
Macanz, a Pelt is a nice loader, it loads a lot, but unless the box is gonna have one normally, it can toss total load over what the box will normally need. So, add a Pelt and do not subtract enough load otherwise to compensate, you might have a PSU that is fine for existing box but has not enough wattage capacity for box plus the Pelt or Pelts. A BUNCH of fans that are high draw, a bunch of HDs that are high draw, could be used to load limit test a PSU also. Most of us do not use Pelts. I use a basic ballast tester based PSU tester plus 4-6 HDs plugged in at once to limit test rough, then multimeter. I am forcing a high 12 volt draw and seeing what is left on legs of lower voltage. If voltages suddenly toss low, I see what I am drawing out of 12 volt in total. When that happens, CD equiv amps in from AC under DC draw at fail tells me a rough efficiency ratio for the AC->DC conversion in cap in on AC versus DC cap out. THEN if need to I will leg load each leg independently, but if the first test fails to what is speced on label, it goes BACK if under warranty.
Comments
I can grab a snapshot of fantastic voltagelines if i want or just stare in the bios if i want but load the stuff for a while and MOST Antec's will change a lot.
those voltages are at 100% usage with 2 instances of f@h running.
Those voltages off the bios, mbm5 (or similar), or did you use a multimeter?
Multimeter is better, yes, but within 1% should be not ACTUALLY out of spec unless the MBM specs for the mobo are off. One way to check MBM versus a decent other guess at reality is to see if when you reboot and go into BIOS the BIOS voltages shown are WAY different by 4-7% or more then you might want to multimeter under no load and then under full load and see what the differences are as a percentage of variance between the two measurements. If the voltages were 3-5% out in MBM, would definitely multimeter, or if other problems with box led me to suspect things were off, then would multimeter also. But so far do not see that here in this thread, so would provisionally high pass the box's power supply absent something that has not yet been discussed coming into the thread.
Quick way to say this-- if those voltages are present and no other issues, then accept MBM as good within a couple percent, and either way the possible values are well within spec still. If other weird issues, multimeter the legs where connected to what you suspect is maybe weird....
John D.
Well, most PSUs can also leg-to-leg balance and do when any one leg is overloaded by the nature of how a PSU works. AC to DC preconversion typically is done with one circuit, so there is one leg coming into the stepdown conversions for each leg. think of it as a big river of power flow from which each leg branches off as a smaller stream. Because of the way the branchng can be done, you can have three way split, a two way to 5 and 12 volt legs, OR additionally have the 3.3 leg coming as a branch off the 5 volt leg or the 12 volt leg.
If you have the latter, a secondary branching, and the first conversion leg's load is too high, it can toss the secondary legs low as the primary leg may go to devices before the branching for secondary legs. Limit for overall is by AC->DC conversion circuit's limit and and is reduced by efficiency factor. Multilegged PSUs need not be three monolithic single leg conversions off of one common source, they can subbranch in many ways.
Without being an engineer, it is easiest to load one leg then see the ampacity remaining for the other legs with something visual. Multimeter in right place is empirical, better than taking input from a sensor set that might not be in the right place(s) to show a defect (or defects). Multimeter from amny places under different loads, done carefully, lets you map the PSUs actuality reality and objectively becasue it cuts out the problems of sensor not in right place or sensor not reading right, or software not interpreting sensor right.
Macanz, a Pelt is a nice loader, it loads a lot, but unless the box is gonna have one normally, it can toss total load over what the box will normally need. So, add a Pelt and do not subtract enough load otherwise to compensate, you might have a PSU that is fine for existing box but has not enough wattage capacity for box plus the Pelt or Pelts. A BUNCH of fans that are high draw, a bunch of HDs that are high draw, could be used to load limit test a PSU also. Most of us do not use Pelts. I use a basic ballast tester based PSU tester plus 4-6 HDs plugged in at once to limit test rough, then multimeter. I am forcing a high 12 volt draw and seeing what is left on legs of lower voltage. If voltages suddenly toss low, I see what I am drawing out of 12 volt in total. When that happens, CD equiv amps in from AC under DC draw at fail tells me a rough efficiency ratio for the AC->DC conversion in cap in on AC versus DC cap out. THEN if need to I will leg load each leg independently, but if the first test fails to what is speced on label, it goes BACK if under warranty.
John.