F5F Stay Refreshed Hardware Desktop Possibly your power source got damaged from overheating.

Possibly your power source got damaged from overheating.

Possibly your power source got damaged from overheating.

E
entech
Member
210
10-27-2019, 03:55 PM
#1
My setup worked perfectly until I connected one more Molex cable to the SATA power adapter. Now my PC won’t turn on at all. Summary: Ryzen 5 3600, RX 2070, Super 2 M.2 NVMe, 4 SSDs, 2 HDDs, 4 fans, 4 DDR4 RAM, 500W quiet D tier gold PSU, non-modular, silent operation, no noise, no smoke, no odor. I removed everything I’d added and restarted, same outcome. After ten minutes of letting the PSU sit, it remained completely unresponsive. Could this be because the extra power draw caused it to fail? Without the final SSD and HDD, the drive setup worked for two days. When I reconnected the power cable into the PSU, I heard some unpleasant buzzing—something you might encounter if it wasn’t fully seated. Any ideas or advice? Recommend a high-end modular PSU rated at least 750W to handle all your components. Also, regarding power consumption: if your system required 550W, it would likely draw close to that value rather than the full 750W unless it was overloaded. In my experience, systems with heavy loads often draw near their maximum output instead of settling around a lower figure, especially when measured with a proper power meter.
E
entech
10-27-2019, 03:55 PM #1

My setup worked perfectly until I connected one more Molex cable to the SATA power adapter. Now my PC won’t turn on at all. Summary: Ryzen 5 3600, RX 2070, Super 2 M.2 NVMe, 4 SSDs, 2 HDDs, 4 fans, 4 DDR4 RAM, 500W quiet D tier gold PSU, non-modular, silent operation, no noise, no smoke, no odor. I removed everything I’d added and restarted, same outcome. After ten minutes of letting the PSU sit, it remained completely unresponsive. Could this be because the extra power draw caused it to fail? Without the final SSD and HDD, the drive setup worked for two days. When I reconnected the power cable into the PSU, I heard some unpleasant buzzing—something you might encounter if it wasn’t fully seated. Any ideas or advice? Recommend a high-end modular PSU rated at least 750W to handle all your components. Also, regarding power consumption: if your system required 550W, it would likely draw close to that value rather than the full 750W unless it was overloaded. In my experience, systems with heavy loads often draw near their maximum output instead of settling around a lower figure, especially when measured with a proper power meter.

R
140
10-27-2019, 07:07 PM
#2
Execute the PSU paper clip test.
R
RabbitKiller99
10-27-2019, 07:07 PM #2

Execute the PSU paper clip test.

D
dylosama
Member
150
10-28-2019, 01:37 PM
#3
a psu only supplies the power required; for example, a 1500 watt unit connected to a 400 watt device would draw just 400 watts plus losses and conversion overhead (assuming 110-120v with 20% loss). this applies to the rtx series which can experience power spikes that may damage the psu if you exceed its limits. choosing a psu close to its maximum capacity improves efficiency because using less of its maximum capacity yields better conversion performance.
D
dylosama
10-28-2019, 01:37 PM #3

a psu only supplies the power required; for example, a 1500 watt unit connected to a 400 watt device would draw just 400 watts plus losses and conversion overhead (assuming 110-120v with 20% loss). this applies to the rtx series which can experience power spikes that may damage the psu if you exceed its limits. choosing a psu close to its maximum capacity improves efficiency because using less of its maximum capacity yields better conversion performance.

A
Amicaty
Member
140
11-13-2019, 01:28 PM
#4
Did you change the PSU and power outlet connections? The 550W unit uses full capacity during peak use, but it draws about 100W when running quietly.
A
Amicaty
11-13-2019, 01:28 PM #4

Did you change the PSU and power outlet connections? The 550W unit uses full capacity during peak use, but it draws about 100W when running quietly.

X
xFqtal_
Senior Member
670
11-18-2019, 07:16 AM
#5
At startup, HDDs consume a significant amount of power, which might become problematic if the power supply is failing.
X
xFqtal_
11-18-2019, 07:16 AM #5

At startup, HDDs consume a significant amount of power, which might become problematic if the power supply is failing.

S
sirus13579
Junior Member
9
11-18-2019, 11:28 AM
#6
Thank you all. I realize the sudden change disrupted my thoughts. I verified everything carefully and decided to manually connect the power pins just in case the power button malfunctioned unexpectedly. By handling all the SATA cables, I accidentally dislodged the PWR_SW jumper. Uh, that’s a bit embarrassing... Please accept my humiliation. On a positive note, I’m relieved the problem is fixed now. After all, @lafrente raises a valid concern. Could the system operate in idle mode but behave differently under CPU and GPU stress? Would a lightweight stress-test program help identify issues with either component? I’m only thinking of simple software that could challenge the GPU or CPU slightly.
S
sirus13579
11-18-2019, 11:28 AM #6

Thank you all. I realize the sudden change disrupted my thoughts. I verified everything carefully and decided to manually connect the power pins just in case the power button malfunctioned unexpectedly. By handling all the SATA cables, I accidentally dislodged the PWR_SW jumper. Uh, that’s a bit embarrassing... Please accept my humiliation. On a positive note, I’m relieved the problem is fixed now. After all, @lafrente raises a valid concern. Could the system operate in idle mode but behave differently under CPU and GPU stress? Would a lightweight stress-test program help identify issues with either component? I’m only thinking of simple software that could challenge the GPU or CPU slightly.

Z
zFlame_br
Junior Member
2
11-22-2019, 03:59 PM
#7
Prime95 and Furmark are effective for stress-testing CPUs, offering strong performance but with some limitations. They might not fully reflect real-world usage, so check if your games include built-in benchmarks.
Z
zFlame_br
11-22-2019, 03:59 PM #7

Prime95 and Furmark are effective for stress-testing CPUs, offering strong performance but with some limitations. They might not fully reflect real-world usage, so check if your games include built-in benchmarks.

M
MrAlonsichuu
Junior Member
8
11-24-2019, 12:57 AM
#8
It's straightforward and often overlooked, which is nice to know.
M
MrAlonsichuu
11-24-2019, 12:57 AM #8

It's straightforward and often overlooked, which is nice to know.