XRTRazor

Honorable
Jun 7, 2012
16
0
10,510
Hi, so I built my first pc today, got all the hardware put together except for a doa optical drive, started up fine all the fans spinning etc got to bios looked around didn't actually change anything then I turned it off.

Next time I go to start it up as soon as I press the power button it trips the 15amp breaker that my 2 bedrooms are on. Hrmm ok so what changed? Well the first time I powered it up the only things on that breaker were a 60w light, the monitor which should only be drawing 40w, and the pc itself(i5-2500k,sapphire 6850,500gb 7200hdd,2x4gb gskill 1600DDR3, with an Antec 620w NEO ECO) while the second time it was 60w light,monitor,13inch tv drawing 50w, and another tv drawing 105w.

So started to play around with stuff seeing what combo would make the breaker trip and after some trial and error it seems that as long as the pc itself is turned on with little else on the circuit after it was to the bios screen(couldn't try anything with real load since doa optical drive and no usb sticks laying around to install windows) you could turn just about anything you wanted on, at one point had the pc,monitor,2 ceiling fans,2 window fans,60w light,40w light,and both tvs on. But, if you started the pc with a decent load on the circuit already it would instantly trip the breaker.

So my question does it sound like my PSU is the culprit and for some reason is just drawing way too much at startup? Unless I'm reading it wrong the max amps that psu should draw would be 9 but I'd think it wouldn't even be drawing anywhere near that considering the rest of my hardware.
 
Solution
Interesting thing about electronics, we only tend to report average and effective values as those are all that tend to matter. The power in your wall actually fluctuates from +170 V to -170 V, the current flowing through that breaker is going in both directions, and the instant your PSU tries to turn on it draws in excess of 50 A from the wall, this is called inrush current and it is a very brief but very large current spike that occurs when all of the electronics in the PSU try to charge up simultaneously. Breakers and fuses take some time to trip, it is not long by a human scale but it is by an electronics time scale. What is happening for you is that for some reason your breaker is tripping much faster than it should, the short...
Interesting thing about electronics, we only tend to report average and effective values as those are all that tend to matter. The power in your wall actually fluctuates from +170 V to -170 V, the current flowing through that breaker is going in both directions, and the instant your PSU tries to turn on it draws in excess of 50 A from the wall, this is called inrush current and it is a very brief but very large current spike that occurs when all of the electronics in the PSU try to charge up simultaneously. Breakers and fuses take some time to trip, it is not long by a human scale but it is by an electronics time scale. What is happening for you is that for some reason your breaker is tripping much faster than it should, the short inrush current shouldn't last long enough to impart enough energy to trip it, but in your case once the breaker has anything else on it and then the inrush spike comes it trips.

You have a couple of options, you can either try to get the computer on its own circuit, find it a 20 A circuit to share(bigger breakers blow slower), or have an electrician or other adequately qualified person take a look at your wiring and breakers and try to figure out why it doesn't work in your scenario.
 
Solution
Hmmm. That's an odd one. It does sound like it's the inrush current when turning the system on that's causing the breaker to trip, but it shouldn't be doing that, obviously.

My house is old and VERY poorly wired, with literally EVERY wall socket in the house on the same breaker, and I've never had that problem myself, even with my current system (with a 750W Rosewill Capstone PSU).

It COULD be the PSU pulling more than it's supposed to, I suppose, but there's really no way to troubleshoot that without a watt meter/multimeter. But even if it was pulling twice what it was supposed to at startup, it shouldn't cause the breaker to trip, I don't think.

Strange indeed, and I'm gonna follow this to see what it ultimately IS.
 

XRTRazor

Honorable
Jun 7, 2012
16
0
10,510
So a friend of mine is gonna swap out the 15a breaker for a new 15a breaker in the next few days first to see if that helps, if not I might get an electrician to put in a 20a breaker instead. Hopefully won't need to rewire anything as the 2 bedrooms were an addition to the house like 8-10 years ago and were wired by a electrician that came very highly recommended, been awhile but believe at the time he said could use 20a breaker on the same wiring but it would be overkill for 2 bedrooms. Also the town's electric inspector who was kinda a stickler for code back then came out and gave it the green light.

Still kinda confused/annoyed how an increase of 100w ish over my old system would create such havoc :/
 

InvalidError

Titan
Moderator

I doubt inrush current in typical PSUs is anywhere near that high since most 15A breakers I have seen have a 25-28A inrush trip threshold. Even crappy PSUs usually have an NTC resistor to limit inrush while higher quality PSUs have line filters and APFC with soft turn-on which also help reduce inrush.
 
It actually has very little to do with an increase in power draw, it has more to do with the internal design of the psu. Tom's PSU reviews aren't good, but they are the only ones who test inrush current(first time that has been useful), if you look at the chart below you will notice that even though they are all approximately the same power level the inrush current levels are all across the board.
power_on_current.png

It is all about the layout of the primary side of the power supply, the initial current spike is to charge up the capacitors from zero to get them up to the incoming voltage, inductors and coils can help to reduce this spike and keep it to more reasonable levels and each PSU has a different layout on its primary side so your old one might have had a big coil before the input side(if it was passive PFC it did) which kept the inrush current low and didn't trip the breaker.


The worst part about computer technologies is that "features" mean something got changed which can lead to unexpected bugs, in this case massive inrush current actually being a problem.

Switching to a new 15A breaker should fix the issue, you aren't anywhere near the current limit for the breaker your current one just trips too damn fast.
 
my first thought was that something is shorted in your case which causes the breaker trip, but since you say it gets to BIOS without any appliances turned on then I guess it's the inrush.

breakers do age, but if you saying it's only 10 year 2 bedrooms breaker it shouldn't have had that much of a heavy use.
 

XRTRazor

Honorable
Jun 7, 2012
16
0
10,510
So finally got a chance to test my pc at a neighbors house and was able to power it up about 15 times in a row and it didn't trip his, granted all his breakers are 20a though. What I think might have made a bigger difference though is that all his breakers were just regular breakers, while it appears mine were AFCI ones, was told that some of the older AFCI breakers had problems with detecting when it was an actual arc and when it was just like an appliance drawing a heavy load at startup.
 

InvalidError

Titan
Moderator

AFCIs both old and new can also trip on noisy loads such as rectifier/thyristor-driven DC motors, insufficiently filtered switching PSUs, non-linear loads like fancy neon lightbulbs, old-fashion fluorescent tube ballasts, etc.

You do not want AFCIs to be so sensitive that they will trip with every little arc but you also want them to be sensitive enough to the right things to trip on non-intentional arcs, which may be a tough balancing act between false positives and false negatives.