Question PC powers on sometimes

[coffeenfifa]

I have an interesting problem.

I have an Asus ROG Crosshair VIII x570 with an AMD 3900x in it and a Seasonic Focus 850W PSU. Other components are GSkill RAM, and a Gigabyte RTX 2070S. The issue is that the computer will power on sometimes. When it does power on it works great until I shut it down, then it is a gamble on wether it will start up again next time or not.

The only solution I've found is to uplug the the PSU and plug it in again. While the computer is powered off, there are still some motherboard LEDs on, which is expected behavior. I have to unplug the PSU, wait a few seconds until those LEDs turn off, plug it in again, and it will start again. After unplugging the computer will start 100% of the time on the next try.

The issue is also intermittent, it will not happen for a few power ons and then it'll happen.

This has nothing to do with the case switches as I'm able to replicate the issue with the motherboard/system completely removed from the case. It is also unrelated to the jumpers where the power switch goes as this motherboard has a "START" button that is on the opposite end of the motherboard.

I should also mention that this has always happened, ever since every component of the PC were new. I'm just now tackling the problem because I now need the hardware for a project and I need it to power on reliably.

Any ideas of what could be happening?

 

[PassMark]

Voltage levels might be marginal, or might be a power sequencing or ramp time issue. In the specs these are T1, T2, T2 & T6.

• Power-on Time (T1): The power-on time is defined as the time from when PS_ON# is pulled low to when the +12VDC, +5VDC, and +3.3VDC outputs are within the regulation ranges. The power-on time shall be less than 500 ms (T1 < 500 ms).
• Rise Time (T2): The output voltages shall rise from ≤10% of nominal to within the regulation ranges within 0.1 ms to 20 ms (0.1 ms ≤ T2 ≤ 20 ms).
• PWR_OK delay (T3): PWR_OK is a “power good” signal. It should be asserted high by the power supply to indicate that the +12VDC, +5VDC, and +3.3VDC outputs are above the under-voltage thresholds.
• Power-down warning (T6): During power-down, the PWR_OK signal should be asserted low at least 1ms before the +12VDC, +5VDC, and +3.3VDC outputs reach their under-voltage thresholds (T6 ≥ 1ms).

If the voltages or timings are out of range, the system might not start up.

You can get PSU testers to test this stuff, but the initial easy step is to borrow a power supply from a friend to test with.

Edit to add diagram:

 

[coffeenfifa]

Voltage levels might be marginal, or might be a power sequencing or ramp time issue. In the specs these are T1, T2, T2 & T6.

• Power-on Time (T1): The power-on time is defined as the time from when PS_ON# is pulled low to when the +12VDC, +5VDC, and +3.3VDC outputs are within the regulation ranges. The power-on time shall be less than 500 ms (T1 < 500 ms).
• Rise Time (T2): The output voltages shall rise from ≤10% of nominal to within the regulation ranges within 0.1 ms to 20 ms (0.1 ms ≤ T2 ≤ 20 ms).
• PWR_OK delay (T3): PWR_OK is a “power good” signal. It should be asserted high by the power supply to indicate that the +12VDC, +5VDC, and +3.3VDC outputs are above the under-voltage thresholds.
• Power-down warning (T6): During power-down, the PWR_OK signal should be asserted low at least 1ms before the +12VDC, +5VDC, and +3.3VDC outputs reach their under-voltage thresholds (T6 ≥ 1ms).

If the voltages or timings are out of range, the system might not start up.

You can get PSU testers to test this stuff, but the initial easy step is to borrow a power supply from a friend to test with.

Edit to add diagram:

So you think this is PSU-related and nothing to do with the motherboard?

 

[PassMark]

Could be motherboard or PSU. If you don't have any test equipment, then swapping out the PSU, then the motherboard is probably the best option.

 

[coffeenfifa]

Thanks. It will be easier to order a PSU to test from Amazon and see if that's the issue than testing or asking someone to lend me theirs.

 

[coffeenfifa]

Well, unfortunately, it was not the PSU. I swapped it out for a new one and I think the problem actually got worse. It is doing it way more than before. So now it has to be the motherboard?