Sign in with
Sign up | Sign in
Your question

PSU trips breaker

Tags:
  • Power Supplies
  • Computer
  • Monitors
  • Components
  • Product
Last response: in Components
Share
January 11, 2011 1:50:13 AM

I've got an 1100W power supply (this one, to be exact: http://www.newegg.com/Product/Product.aspx?Item=N82E168...).

I recently got a 30" monitor, and when I tried plugging it in and turning on the computer, I would always trip the breaker for the room. I thought maybe I was plugging too many things in, so I tried unplugging EVERYTHING except for the computer and the 30". Still same issue.

It does seem to work if I turn the computer on and then plug in and turn the monitor on, but this seems disconcerting. I checked the basement and the breaker for the room is 15A. It seems like something is definitely wrong (or my math is wrong), since 1100W should be about 9A. The monitor manual says it draws 2.4A. 11.4A < 15A.

Does this make sense? My other question is if either the monitor or PSU being faulty could cause this, and how would I check? I'm about ready to call my landlord and have her send an electrician, but I don't want them to get here and tell me I'm plugging in junk.

More about : psu trips breaker

a b ) Power supply
January 11, 2011 2:02:56 AM

lousy lousy psu :( 
m
0
l
a c 1218 ) Power supply
a b C Monitor
January 11, 2011 3:48:54 AM

You say the 15 Amp breaker is for the room.

Are the room lights also on this circuit?

Are there other power outlets on this circuit, and if there is, are there other items plugged in and powered on at the same time?

It may be the power on current surge that the power supply draws when it is first turned on that is tripping the breaker when other items are powered on at the same time.

That 11.4 Amps is the nominal (steady) current draw when the power supply is under full load. Initial power on current surge (spike) is higher than nominal.
m
0
l
Related resources
January 11, 2011 4:08:21 AM

That seems reasonable. I did try unplugging everything other than the computer and the monitor (with the monitor off, even) but I still had the same issue. And yes, the room lights are also on this circuit, but they were off.

I'm getting around it now by turning the computer on, and then flipping the switch on the surge protector that runs the monitor. If there was a better way, that'd be great .. maybe not.
m
0
l
a b ) Power supply
January 11, 2011 4:35:48 AM

what are your system specs?
for most things an 850w is sufficient
m
0
l
January 11, 2011 12:56:32 PM

It is an arc fault breaker (I think this is what you're describing as an RCD). When I get home I will double check everything in the case to make sure nothing is out of place, but I did not modify the computer (except to plug the monitor in), so I would think if that was the issue, it would have been tripping previously.

And yes, 1100W might be a bit overkill (maybe not?). I have a core i7 + 3 hard drives + 1 BDROM drive + 2 GTX460s.
m
0
l
a b ) Power supply
January 11, 2011 1:02:12 PM

an 850w HIGH QUALITY psu would pe plenty for you...xion, while not being the total worst, isnt exaxty the best of psu makers. how long have you had the psu?
m
0
l
January 11, 2011 3:50:40 PM

It is nearly new (3 months). If that's a likely culprit I can run down to TigerDirect and pick up a better one and see if I still have the same issue.
m
0
l
a c 1218 ) Power supply
a b C Monitor
January 11, 2011 5:15:30 PM

After calculating the +12 Volt DC current draw for your configuration and the four PCI-E Connectors required for your graphics cards a high quality 750 Watt power supply like the CORSAIR HX Series CMPSU-750HX is more than sufficient to meet your system's power requirements.
m
0
l
a b ) Power supply
January 12, 2011 3:41:19 PM

Surely the wattage rating of the PSU is not the total picture. Whatever rating it has, the actual power it pulls from the wall is determined by the load of all the computer components. So if the computer uses 600W, and the PSU is 80% Efficient, the PSU will only pull 750W, not 1375 W (that's 1100W at 80% Eff, works out to 11.4A at 120 VAC). This indicates that the computer actually does consume a lot of power and a smaller PSU may NOT be a good idea.

Now, I agree that the initial start-up rush current will be higher, and maybe that's part of the issue. But it does seem that the computer is actually using a lot of power. Some breaker designs have a bit of a time delay in their response to allow for short-term surges of current, and some are less tolerant. The new designs that combine plain overcurrent protection with Ground Fault detection (the GFCI system as it's called in North America) probably have much less time delay for the GCFI function (although a heavy temporary current that is balanced on both Hot and Neutral lines if not a Ground Fault indication).

Bottom line here appears to be that OP is trying to use a lot of power on a circuit that is already loaded near max. Maybe the start-up condition is sufficient to exceed the max load, even temporarily. In that case there are two potential solutions. One is to move some of the load to another circuit. The other is to stagger the start-up of some devices, thus reducing the temporary max load.
m
0
l
a c 131 ) Power supply
a b C Monitor
January 12, 2011 7:50:27 PM

They are also called "GFCI," for "Ground Fault Circuit Interrupter." Outlets (but not necessarily breakers) with that designation are REQUIRED for new construction in bathrooms, and I believe outside outlets. They are good safety features in those applications, where moisture getting where it doesn't belong can cause electrocution.
Still, Xion isn't known for PSU quality, although SMPS designs do have high inrush current as their primary capacitors charge. A better PSU may not address this issue if what you have is a breaker problem. Only 15A seems low; new construction almost always uses 20A breakers on branch circuits. You might simply have a defective breaker.
Getting a quality PSU (~750W) is certainly a good idea, but you may want to call your landlord about that circuit even if that "fixes" the immediate problem.
m
0
l
!