Question Circuit Breaker trips when using new PC ?

[Nguyen2526]

I recently got a new PC with the following specs:

• CPU: AMD Ryzen 7 9800X3D
• CPU Cooler: NZXT Kraken Elite 360 RGB (2024) – 360mm AIO
• Fans replaced with Lian Li UNI Fan SL Infinity 120mm (3-pack, Fluid Dynamic Bearing)
• Motherboard: Gigabyte B850 AORUS ELITE WIFI7
• PSU is Asus ROG THOR P2 Gaming 850 W 80+ Platinum Certified Fully Modular ATX Power Supply
• RAM: Corsair Vengeance RGB DDR5-6000 CL36 – 64GB (4×16GB)
• GPU: Zotac GAMING AMP Extreme INFINITY GeForce RTX 5070 Ti
• Storage: WD_BLACK SN850X 1TB NVMe SSD (Model: WDS100T2X0E)
• Case: Lian Li O11D EVO RGB
• Power/RGB Accessories:
  • Lian Li Strimer Plus V2 (Wireless RGB w/ RF Controller):
    • 24-Pin Motherboard Strimer
    • Dual 16-Pin GPU Strimer
  • Lian Li UNI Fan SL Infinity 120mm (Standard + Reverse):
    • 2× 3-Packs (Standard)
    • 1× 3-Pack (Reverse)
    • 1× Single Fan Unit

My friend built and tested the PC before shipping it to me. It ran perfectly for two hours at his place. However, after setting it up in my room, it began tripping the circuit breaker—sometimes within 10–30 minutes, other times immediately—even during light tasks like watching YouTube or logging in.
I live in a small town where PC support is limited.
Local repair services charge around $190 just for diagnostics.
I checked the temps using HWMonitor. Most components are fine, but the CPU cooler idles at around 80°C, which seems high and might be due to a thermal paste issue (possibly an air bubble).
My brother's room and my room share the same circuit breaker. He only has a TV and a phone charger plugged in. I have the PC, monitor, and a charger.
I called my home warranty company to send someone out.



Technician 1:
Suspected the 20A AFCI/GFCI breaker was faulty—possibly heat-damaged (I’m in Texas).
Tested the wiring, ground, outlets, and for loose wires—no issues found. Recommended replacing the breaker, but couldn’t finish the job as he was leaving for vacation.



Technician 2:
When he came out, he tested and replaced the outlet (the issue persisted). Asked for wattage details. I told him the PSU is 850W and the monitor is 60W. I clarified that the PC likely only draws about 750W under load, not the full 850W. We tested the PC in the living room, and it didn’t trip the breaker or shut itself off. I mentioned the first tech’s theory about the breaker. He said he didn’t have time to check that day and would have to return. Which he never did he submitted a report claiming my system was overloading the circuit.
The warranty company denied the claim, saying they don’t cover overloads or equipment upgrades.
I tried to appeal with the third tech’s diagnosis, but they still denied it.



Technician 3 (Independent Electrician):
He opened and tested the breaker. Initially got 0.20V, then 120V after a reset. Also recommended replacing the breaker. Quoted $500 for the job.



I also had one last technician come out to be sure and he said the breaker is perfectly fine nothing wrong with it and it could be damaged wires in either mine or brother room he’s charging $280 for one bedroom just to check outlets and light switches if he finds anything he will charge extra to fix it.



Online Suggestions and Theories:

• Faulty PSU
• Inrush current triggering a sensitive AFCI breaker
• Breaker is bad and needs to be replaced
• Use a UPS to buffer startup spikes

I’m worried that the psu could be faulty or got damaged during shipment and could damage other components. I bought the psu off of amazon ship&sold by Amazon
Someone offered to replace the PSU for $175 (labor only), but I’d still need to purchase a new PSU—I’m open to recommendations.
Any advice or insights would be greatly appreciated. I also tested the outlet myself with klein tools receptacle. Result: Correct.

 

[Misgar]

GFCI is fine, with the following caveats:
These will greatly extend the lifetime of any attached surge protectors because they will cut the power as soon as any of those start redirecting surges to ground,

True, but the initial high current ATX PSU power on surge I'm considering flows from Line back to Neutral and (as far as I can work out) any current in the Class-Y capacitors in the PSU should be less than 2.5mA (estimate). This should not be enough to trip a 20mA or 30mA GFCI.

I'm theorising this from my work as an electrical and electronics engineer in power systems design, but since I don't live in the USA, I'm not familiar with all the regulations in force in home installations. My only concern is finding a solution for the OP.

All four bedrooms have AFCI, the kitchen and three bathrooms have GFCI, and even the garage has an AFCI.

I find it interesting that each bedroom is protected by its own AFCI. This is not standard practice where I live, probably because most house are constructed with brick walls and less timber than US homes, or the regulations are different.

When designing Milspec installations, I've always regarded GFCI's as a safety device to guard against death due to electric shock, not as surge protectors. Hence their use in the OP's bathrooms where water makes electric shock more likely (as happened to me in a Buenos Aires hotel bathroom on a wet floor - not a nice experience).
https://www.electricaltechnology.or...-fault-circuit-interrupter-types-working.html

We still don't know the nominal current rating of the OP's AFCI but I suspect it may simply be too low if it protects a single room. When I was living in a Hall of Residence at Uni, each room had a 1.35A circuit breaker outside. In the girls block, they continually tripped out their breakers when using hair dryers.

If, as I was guessing, the OP's bedroom breaker is rated as low at 10A, that makes it more likely to trip whenever the ATX PSU demands high currents. The biggest (short duration current spike) occurs when AC mains is connected to the PSU. Further high currents will occur during heavy gaming, rendering tasks or stress testing.

A PSU rated at 850W could pull over 900W from the mains (850W out plus 10% extra if the PSU is 90% efficient). If the mains is 120V, 900W = 7.5A. The PSU is never likely to draw 900W, but it needs to be considered. If the breaker is rated at 10A and it's a "fast trip" version, current surges on top of a continuous demand of, say, 5 to 6A could push the breaker up to its trip point.

I note the OP does not want to sacrifice the arc protection provided by an AFCI, but if the bedroom breaker is only 10A and US regs permit, it might be an idea to fit a higher current rating, e.g. 20A. Alternatively, do AFCI breakers come in a range of trip delay times (fast, medium, slow)? Is the OP's breaker trip time too sensitive (i.e. fast)? You need a competent qualified electrician with a good understanding of ATX PSUs to answer that question.

I'd try as many computers as possible in the room and see if any of them have PSUs which work without tripping the breaker. As I said earlier, some PSUs pull only 20A briefly at power on. Others pull 100A. Swapping to a PSU with a much lower current spike might be a solution. Read enough Tom's reviews (the full multi page versions) and you might find a "better" PSU with lower power on surge.

I doubt if the problem is anything to do with leakage currents from Line down to Earth, which in any case should trip a GFCI elsewhere if sufficiently high. Where I live, GFCI's are typically rated at 30mA and leakage currents through Class-Y components in an ATX PSU rarely exceed 2.5mA. You'd need at least 10 computers (25mA) plus a few more items with Class-Y filters to reach 30mA.

I've had two lightning strikes near my house. On both occasions, high voltages were conducted through underground wiring (powercables, phone lines, broadband coax) and possibly metal water and gas pipes. The result was dead routers, blown Ethernet switches and a cremated telephone answering machine. In a friend's house, high winds and storms (not lightning) trip their GFCI. It could be due to overhead power lines touching wet tree branches. I'm not sure.

Power systems are strange items and not always easy to trouble shoot, as they found out in Spain and Portugal recently with a major blackout.

 

[BFG-9000]

In the 60s and 70s there was a copper shortage due to Vietnam, so aluminum wiring became popular but unfortunately oxidizes, expands and contracts more so turned out to cause fires if used with terminals designed for copper, particularly those iffy "backstab" spring-contact types. It became almost impossible to insure houses with such wiring as they were ~55x more likely to catch fire, but eventually properly engineered terminals suited for aluminum wire became commonplace, though extra care was still required to use them.

Well nowadays the price of copper is sky-high enough to get rid of pennies, and construction build quality is noted for sloppiness and general lack of care in assembly, so the insurance companies require oversensitive AFCI devices whether or not aluminum wire is used. Otherwise no fire insurance for you. But they are well known to trip all of the time when no hazard is present.

OP did mention in the 1st post that it's a 20A breaker, which suggests 12AWG copper or 10AWG aluminum. And I didn't say GFCI may contribute any issue with their ATX PSU, but only that it may prove a nuisance if a surge protector is plugged into it. Because that diverts to ground, which will be detected.

 

[Nguyen2526]

I’m using a 20A breaker

https://imgur.com/a/DMjAN7F

View: https://imgur.com/a/DMjAN7F
We have a smart kettle that runs fine and doesn’t trip the breaker. @bill001g — the brand of switch we have isn’t even sold anymore. Also, we don’t use any window AC units or space heaters. For psu what would you guys recommend?
https://pcpartpicker.com/user/Poku/saved/YWtKcf sorry I just seen more responses after I posted this let me read real quick. My brother pc is around 550w? When we used it in my room it never trips the breaker. We plugged it in again this weekend and it works perfectly fine. When the electrician had replaced the outlet he told me to buy this one 15amp for outlet

https://imgur.com/a/7eVUj3s

View: https://imgur.com/a/7eVUj3s

 

[BFG-9000]

Well it's a 200A breaker panel being used as a sub-panel because the neutral is not bonded to ground. Maybe the main breakers are outside for easy access by firemen? The panel came with this bonding bar but you can clearly see it has been removed, as is required for a sub-panel:

It takes GE THQL and THQP size breakers.

Unless you can find an updated PSU tier list that you like, try one recommended here at Tom's.

 

[bill001g]

I still suspect the power supply so don't get too distracted by the stuff I post below.

I will assume the ones that always trip are the white breakers on the bottom?

What is very strange is how the breakers on the top are wired. Those are fairly standard 220/240 breakers and at first glance it looked correct. There are 2 hot wires used for 240 and red and black are the common color. When you look closely what they have done combined the red and black wires and only connected to 1/2 the breaker. This means this is going to run at 120 volts.

You would have to get a real electrician that know the details of the code if this is really allowed. I know there are simple things like only 1 wire is allowed to be put under a single screw. Running 1/2 a 240 breaker is more suspect. Why didn't they just pull the 240 breakers and put in 2 120. I wonder if the red and black wires go to different rooms on the far end. If it is really just a 120 circuit why didn't they just cut the red wire off and use just the black one.

 

[Misgar]

We have a smart kettle that runs fine and doesn’t trip the breaker.

A kettle element is pretty much a resistive load and might even have a small inductive element in its AC impedance which could reduce current surges.

An ATX PSU on the other hand has a dirty great big bulk capacitor (400V electrolytic) which gulps down current for a few milliseconds when power is applied. Hence the 20 to 100A initial surge.

Resistive loads and capacitive loads are two different things. I don't believe a kettle is a realistic substitute for an ATX PSU when testing breakers.

My brother pc is around 550w? When we used it in my room it never trips the breaker.

I'm betting your brother's PSU has a lower inrush current when power is first connected.

The inrush current is not necessarily lower because it's a 550W supply and yours is 850W. Remember that 550W unit I used as an example of high inrush currents? You can't always equate the power rating of an ATX PSU with its inrush current. A badly designed 550W PSU can be worse than a well designed 850W and vice versa.

Someone offered to replace the PSU for $175 (labor only), but I’d still need to purchase a new PSU—I’m open to recommendations.

Wow. $175 for a ten minute job! If you're reasonably competent and can wield a screwdriver, it's a DIY job. If you have any doubts, find a friend who builds PCs and buy them a beer (where I live, you can buy alcohol in a bar when you're 18).

Recently, Tom's PSU tests have stopped giving inrush current figures, which makes life difficult when shopping for PSUs with low inrush currents.

This test from 2020 shows a NZXT PSU with a low inrush current at 115V (but high at 230V for folk like me).
https://www.tomshardware.com/reviews/nzxt-c-series-850w-power-supply-review/2

I'm not saying this is a good PSU, but a 29A inrush current @115V is quite low. You won't be able to measure inrush currents yourself unless you have a storage scope and the relevant expertise.


This Corsair PSU from 2021 is even better. Only 24.440A at 115V.
https://www.tomshardware.com/reviews/corsair-rm850x-2021-power-supply-review/2

There's absolutely no guarantee that a modern 2025 RM850x will have the exactly the same design, but it's worth trying in place of your existing PSU.

 

[_D_D]

Inrush happens power is first connected to an unpowered circuit, not 10 minutes afterward.

However, after setting it up in my room, it began tripping the circuit breaker—sometimes within 10–30 minutes, other times immediately—even during light tasks like watching YouTube or logging in.

As a usually easy test maybe you could try a different power cord to the PSU to rule that out.

 

[Nguyen2526]

I still suspect the power supply so don't get too distracted by the stuff I post below.

I will assume the ones that always trip are the white breakers on the bottom?

What is very strange is how the breakers on the top are wired. Those are fairly standard 220/240 breakers and at first glance it looked correct. There are 2 hot wires used for 240 and red and black are the common color. When you look closely what they have done combined the red and black wires and only connected to 1/2 the breaker. This means this is going to run at 120 volts.

You would have to get a real electrician that know the details of the code if this is really allowed. I know there are simple things like only 1 wire is allowed to be put under a single screw. Running 1/2 a 240 breaker is more suspect. Why didn't they just pull the 240 breakers and put in 2 120. I wonder if the red and black wires go to different rooms on the far end. If it is really just a 120 circuit why didn't they just cut the red wire off and use just the black one.

Should I call an electrician to come out and ask them about it? I have a smart meter outside and there some other I think? I never really looked into it since it looked normal to me

 

[Nguyen2526]

Inrush happens power is first connected to an unpowered circuit, not 10 minutes afterward.

As a usually easy test maybe you could try a different power cord to the PSU to rule that out.

The only cord I have is the one my brother have which is 550w we had plugged it in before and it would still trip the breaker but the cord is universal

 

[Nguyen2526]

A kettle element is pretty much a resistive load and might even have a small inductive element in its AC impedance which could reduce current surges.

An ATX PSU on the other hand has a dirty great big bulk capacitor (400V electrolytic) which gulps down current for a few milliseconds when power is applied. Hence the 20 to 100A initial surge.

Resistive loads and capacitive loads are two different things. I don't believe a kettle is a realistic substitute for an ATX PSU when testing breakers.


I'm betting your brother's PSU has a lower inrush current when power is first connected.

The inrush current is not necessarily lower because it's a 550W supply and yours is 850W. Remember that 550W unit I used as an example of high inrush currents? You can't always equate the power rating of an ATX PSU with its inrush current. A badly designed 550W PSU can be worse than a well designed 850W and vice versa.


Wow. $175 for a ten minute job! If you're reasonably competent and can wield a screwdriver, it's a DIY job. If you have any doubts, find a friend who builds PCs and buy them a beer (where I live, you can buy alcohol in a bar when you're 18).

Recently, Tom's PSU tests have stopped giving inrush current figures, which makes life difficult when shopping for PSUs with low inrush currents.

This test from 2020 shows a NZXT PSU with a low inrush current at 115V (but high at 230V for folk like me).
https://www.tomshardware.com/reviews/nzxt-c-series-850w-power-supply-review/2

I'm not saying this is a good PSU, but a 29A inrush current @115V is quite low. You won't be able to measure inrush currents yourself unless you have a storage scope and the relevant expertise.


This Corsair PSU from 2021 is even better. Only 24.440A at 115V.
https://www.tomshardware.com/reviews/corsair-rm850x-2021-power-supply-review/2

There's absolutely no guarantee that a modern 2025 RM850x will have the exactly the same design, but it's worth trying in place of your existing PSU.

moved out here a few years ago and haven’t really made any tech-savvy friends or anyone who knows how to replace PC parts. I was thinking about doing it myself, but the cable management in the back looks kind of messy, so I’m a bit nervous about messing something up. With how much people are charging out here though, I might just go ahead and do it myself. Do you think I should stick with 850W or upgrade to 1000W? My friend back in Florida has the RM series and said it’s been great—she’s had it for years. So I might just go with that one. But does it have the features where it will shut down if anything happens? Like to protect itself or the other parts sorry if I’m not making sense.. I seen online some psu be exploding or catching on fire got me scared

 

[Nguyen2526]

sorry for the late response I really appreciate everyone helping me and suggesting fixes I really hope replacing the psu will solve my issue