Question I need help debugging the EZ Debug LEDs on an MSI Z790 Gaming Plus Wi-Fi motherboard ?

[NicDG]

I am trying to build a (mostly) new desktop PC with the following components:
i9-14900 KF
MSI Z790 Gaming Plus Wi-Fi Motherboard
CORSAIR Vengeance 64GB (2 x 32GB) DDR5 RAM CMK64GX5M2B5600C40
MSI Mech Radeon RX 6600 8GB GDDR6 PCI Express 4.0 Graphics Card
RAIDMAX RX-735AP 735W PSU
XPG GAMMIX S70 Blade: 1TB M.2 NVMe SSD

It is not getting through POST and the EZ Debug LEDs are not helping me isolate possible faulty components.

On the first power on all fans started spinning, but nothing further happened. I powered off and back on I had video and everything looked fine briefly. The MSI logo appeared on screen then the spinning circle froze and nothing further happened. This is the only time I saw a video signal. I powered off and then back on while closely monitoring the debug lights and saw:
All fans were spinning
red CPU light blinked on briefly (about a second) then switched off.
yellow DRAM light comes on for 20 seconds then goes out
Fans continue spinning, no video signal, and POST does not proceed any further.

I powered off and back on a few times with the same result. I tried the reset button instead of the power button after the yellow LED went out and the computer then restarted to a glowing red CPU light and went no further.
I powered off and on and pressed the reset button while the yellow RAM LED was on and nothing happened until the yellow LED went out, then the computer restarted and went through the cycle of red LED blinking on and off followed by the yellow LED on for 20 seconds then nothing further.

So I started trouble shooting.

I removed the GPU and nothing changed, still got the red LED on for a second then off followed by the yellow LED for 20 seconds.

Next I removed all the RAM and on attempting to boot, the red LED blinked on for a second then off and the yellow RAM LED came on it stayed on.

I tried each of the two RAM modules in slot A2, then each module in each wrong slot (i.e. A1, B1, B2), then with one module in A2 I tried the other module in the wrong slots (A1 and B1) and in each case after power on the red CPU light blinked on and off and the yellow DRAM LED comes on for 20 seconds then goes out and POST does not continue any further.

I next removed the CPU and when the I powered the computer on the red CPU light came on and stayed on. While the CPU was out I checked for bent pins and did not find any. I also unplugged and reinserted the CPU and motherboard cables more than once.

Support for 14th gen CPUs was added in Sept 2023, so I do not think a BIOS upgrade was necessary, but I updated to the most recent BIOS anyway, however it did not make a difference.,

Any thoughts are appreciated.

 

[Aeacus]

I removed the GPU and nothing changed

Well, yeah, since you have F-suffix CPU which doesn't have iGPU in it.

RAIDMAX RX-735AP 735W PSU

Why? 🙄

Do note that you have crap quality PSU.
This is easy to see by the actual wattage it outputs. +12V rail is rated 53A which makes 636W. And one can not magically combine rails to increase output wattage. In no way the PSU can output 735W. Your PSU, at best is 635W unit. Most likely far less, since the label on it lies.

Anyone who are able to afford Core i9, are able to afford good/great quality PSU as well. Like: Seasonic Focus/Vertex/PRIME, Corsair RMx/RMi/HXi/AXi, Super Flower Leadex Gold/Platinum/Titanium.

Since PSU powers everything, it is the most important component inside the PC.
Hence why NEVER cheap out on PSU! Also, never buy used PSU either.

Any thoughts are appreciated.

New good/great quality PSU is a MUST.

Symptoms show dead: CPU, MoBo, RAM or PSU. Or ALL of them dead.
Most likely dead MoBo, since MoBo is usually 1st to go when PSU acts up. CPU and RAM are more durable but not invulnerable either.

To test CPU and RAM, you need 2nd, compatible build, where to test them out.
Best to haul system to PC repair shop and pay for diagnostics.

Life lesson: do not let your desire for cheap PSU to overcome you. Since it currently costed you your expensive MoBo. Probably RAM and your Core i9 as well.

 

[NicDG]

Well, yeah, since you have F-suffix CPU which doesn't have iGPU in it.

I would expect the GPU LED to come on and stay on as no GPU was present

Do note that you have crap quality PSU.

True, but it has served me well for 10 years and even 636 watts should be sufficient to get the system running as long as the CPU is not stressed. None the less it was on the list to be replaced and a rep[lacement PSU in transit.

 

[Aeacus]

I would expect the GPU LED to come on and stay on as no GPU was present

GPU LED will only lit up when GPU is present but is detected as failure.

Since many Intel CPUs have iGPU in them, system can run without GPU and there's no need for GPU LED to light up in those instances.

True, but it has served me well for 10 years

A PSU that kills other components can not be considered as well serving, despite how long you were able to push your luck.

E.g my PSU is now 9 years old, but my PSU is Seasonic PRIME 650 80+ Titanium [SSR-650TD] and it was the best 650W PSU money could buy back then (confirmed by several reputable reviews). It has 12 year warranty, so, easy 3 more years lifespan for it. While Seasonic units are known for their durability, lasting FAR longer than their warranty period.
So, with great quality PSU, you can say that PSU has served well. But crap quality PSU that kills hardware - i don't think so.

and even 636 watts should be sufficient

Label already lies about total wattage. With this, there is 0 conformation if the PSU even is able to output 53A on +12V rail. Since no reviews (let alone reputable reviews) exist of your PSU, it can't be confirmed either.

Though, for such crap units, i've seen that they fail at much lower wattage levels. 100W to 200W less than what is said on the label.
Proof:

View: https://www.youtube.com/watch?v=f6snWfd1v7M


So, what you have, could very well be 350W to 450W unit.

and a rep[lacement PSU in transit.

Make and model (or part number) of the replacement PSU is?

 

[NicDG]

Make and model (or part number) of the replacement PSU is?

Thermaltake GF1 850W (with a 10 year warranty, so hopefully better quality) Sadly it has not changed the lack of booting. I am also working with MSI support who are also pointing to the PSU.

Thank you for your ongoing educational comments.

 

[NicDG]

Though, for such crap units, i've seen that they fail at much lower wattage levels. 100W to 200W less than what is said on the label.
Proof:

So, what you have, could very well be 350W to 450W unit.

Well those failures are pretty dramatic. Fairly obviously I never experienced anything quite so dramatic.

My previous build was a Ryzen 5 5600X, MSI MAG B550 TOMAHAWK Motherboard, 32 GB RAM and the Radeon 6600 GPU which Newegg estimates at 325 watts which is under your low estimate for real watts, so I guess that's why it "served me well"

I am now curious as to what its wattage actually is, but probably not curious enough to go to the time and expense to find out.

 

[Aeacus]

Thermaltake GF1 850W (with a 10 year warranty, so hopefully better quality)

Tier A+ PSU, so great choice. 👍
PSU tier list: https://docs.google.com/spreadsheet...JWkc/edit?pli=1&gid=1973454078#gid=1973454078

Sadly it has not changed the lack of booting. I am also working with MSI support who are also pointing to the PSU.

With PSU ruled out as one of the possible culprits, it leaves either: CPU, MoBo, RAM or GPU issue.

So, try testing with 2nd, known to work GPU and RAM. This would rule out these two.
(If you would not have F-suffix CPU, you wouldn't have to test GPU, instead you could use iGPU inside the CPU. Something to keep in mind when buying CPUs, don't get F-suffix chips.)

To test out CPU, you need to plop in any, 12th, 13th or 14th gen, known to work, Intel CPU. To see if MoBo POSTs.

If you do not have 2nd, compatible parts, you can haul your hardware to PC repair shop and pay for diagnostics.

Another option is going with new MoBo off the bat, since DOA MoBo is still a thing and i'd suspect it the most (after PSU was ruled out).

That is, unless your Raidmax PSU killed the new components you bought. In this possibility, CPU, MoBo, RAM and GPU can be all dead. Whereby you need to test them all out, individually. To see what survived and what didn't.

All-in-all, expensive repair. But that's the power of cheap, crap quality PSUs. They make you pay far more money afterwards, completely negating any savings made when buying them initially.

My previous build was a Ryzen 5 5600X, MSI MAG B550 TOMAHAWK Motherboard, 32 GB RAM and the Radeon 6600 GPU which Newegg estimates at 325 watts which is under your low estimate for real watts, so I guess that's why it "served me well"

325W is about right for that build.

I am now curious as to what its wattage actually is, but probably not curious enough to go to the time and expense to find out.

Well, with PSU testing equipment, you can find it out. Just increase the load on PSU in increments and look at what wattage output it goes "boom" with nice fireworks. (Just like in the video i linked.) Then you know how much it was able to output really.