Question No POST with 2 RAM Modules, 1 RAM Module Memtest error

[Kyhz]

Hey! First of all, sorry for my grammar, I'm not a native English speaker
I want to revive a dead computer that died because a PSU failure.
I bought new PSU, RAM and SSD.
Then I found a boot loop problem (I'd say the same that Im having right now) so I bought a new motherboard because I thought it could be the problem. I bought the motherboard secondhand and It worked!!!
I installed everything and it worked the first day! I managed to install Windows 10 and all the stuff and I kept the computer for 2 months, then I went to start it and ran into the same problem than before.

Things that I tried:

• Replacing CPU and nothing changed
• Replacing PSU and nothing changed
• Disconnect all peripheral and SATA and nothing changed
• Update BIOS and nothing change
• Hard reset BIOS to default values and nothing changes (CMOS reset)

Possible solution:
The RAM is brand new but is this amazon model: TECMIYO 8GB Kit (2x4GB) PC3 12800 DDR3 1600 CL11 1.5V 240Pin Non-ECC Unbuffered UDIMM Desktop Memory Ram Module
I managed to boot correctly the computer by only booting with 1 RAM stick in DIMM2. If i insert both RAM stick it doesn't boot. Even If I insert 1 stick in DIMM1 It doesn't boot.
It's very weird because It worked the first day with both sticks!! And I haven't used the PC since then.
I also checked If the DIMM1 slot has any pin burnt but I can't see anything.

I attach a video of the boot loop and my computer components:
Motherboard: MSI H61-P32/W8
No GPU.
CPU: Intel i5-3450
PSU: Mars Gaming MPII550
RAM: TECMIYO 8GB Kit (2x4GB) PC3 12800 DDR3 1600 CL11 1.5V 240Pin Non-ECC Unbuffered UDIMM Desktop Memory Ram Module

As I said, I can boot with only 1 stick. So I decided to run a memtest of that signle stick, and I run into this problem in the middle of the test:

 

[Ralston18]

Your memtest image is not viewable.

When you replaced the CPU how did you apply the thermal paste and what brand thermal paste did you use?

Do you have another known working PSU that you can install and test with?

 

[Kyhz]

Your memtest image is not viewable.

I think I fixed the img in the first post. Anyway here's the link:

https://imgur.com/BgvosvC

View: https://imgur.com/BgvosvC

When you replaced the CPU how did you apply the thermal paste and what brand thermal paste did you use?

I applied it drawing a "X". This thermal paste: https://www.amazon.es/gp/product/B0187EWNFU/ref=ppx_yo_dt_b_asin_title_o02_s00?ie=UTF8&psc=1

Do you have another known working PSU that you can install and test with?

Yes! I already did it and It didnt work either.

 

[Ralston18]

Not familiar with that thermal paste but an "X" application may not be correct.

May have worked at first but degraded and is now failing to properly transfer heat.

Hopefully some other reader will post a comment accordingly.

 

[tecmo34]

The "X" method is a highly recommended method, outside of spreading a thin layer over the CPU. You get better coverage with the "X" over the a single dot that most recommend, so the user should be good there.

I don't know if I would trust that RAM and does sound like a possible issue. Have you tried a GSkill or Corsair RAM in the PC? I trust those two brands the most when it comes to RAM.

Have you updated the BIOS?

 

[80-watt Hamster]

What, specifically, are you wondering? Your initial post doesn't really ask a question.

Any component can fail at any time, even when new. Memory is the most common failure after hard disks, followed by motherboards, then probably power supplies. CPU failure is pretty rare, though having gone through a blown PSU, all bets are off. All you can really do here is try different combinations of components. I'd honestly start with the old board and RAM, one stick and one slot at a time. If all of that works, repeat with new memory. It's then up to you whether to bother testing on the new board.

Regarding thermal paste, for that processor/heatsink combination, the vertical line method is pretty solid (IMO) due to the shape of the Ivy Bridge die. Make sure all the mounting pins are secure, that can bite even experienced builders on occasion. If you think CPU overheating may be a problem, boot into Windows if you can and monitor the temperature for awhile (unless you can find a monitor in BIOS, the manual for your board makes it look as though there isn't one), ideally while running a mild benchmark or stress test. If you stay under 60C at idle or 75C under load, you should be fine.

 

[Kyhz]

The "X" method is a highly recommended method, outside of spreading a thin layer over the CPU. You get better coverage with the "X" over the a single dot that most recommend, so the user should be good there.

I don't know if I would trust that RAM and does sound like a possible issue. Have you tried a GSkill or Corsair RAM in the PC? I trust those two brands the most when it comes to RAM.

Have you updated the BIOS?

Yes I tried updating BIOS but didn't work either...
I bought HyperX RAM. Lets see If it works. If not then i guess is not RAM problem

 

[Kyhz]

What, specifically, are you wondering? Your initial post doesn't really ask a question.

Any component can fail at any time, even when new. Memory is the most common failure after hard disks, followed by motherboards, then probably power supplies. CPU failure is pretty rare, though having gone through a blown PSU, all bets are off. All you can really do here is try different combinations of components. I'd honestly start with the old board and RAM, one stick and one slot at a time. If all of that works, repeat with new memory. It's then up to you whether to bother testing on the new board.

Regarding thermal paste, for that processor/heatsink combination, the vertical line method is pretty solid (IMO) due to the shape of the Ivy Bridge die. Make sure all the mounting pins are secure, that can bite even experienced builders on occasion. If you think CPU overheating may be a problem, boot into Windows if you can and monitor the temperature for awhile (unless you can find a monitor in BIOS, the manual for your board makes it look as though there isn't one), ideally while running a mild benchmark or stress test. If you stay under 60C at idle or 75C under load, you should be fine.

I'm wondering what component could be causing the boot loop.
I checked 2 different RAM models, both of questionable reliability, the tests results are confusing:

TECMIYO BRAND

• 2 RAM sticks: Boot loop
• 1 RAM stick in DIMM2 : Can boot to windows but memtest fails with the image provided in 1st post
• 1 RAM stick in DIMM1: Boot loop

Komputerbay BRAND:

• 2 RAM Sticks: boot loop
• 1 RAM stick in DIMM2: Not boot loop, but I got image signal crash as image provided in 1st post before booting to Windows
• 1 RAM stick in DIMM1: Boot loop

I'll wait for tomorrow for the HyperX to test them out.
The CPU temp seems to be ok (28ºC at idle)

 

[Kyhz]

I checked with brand new HyperX RAM and I keep having the same problem...
I guess is the secondhand motherboard that I boguht that maybe worked the first day but then stop working...
I don't have the old motherboard so I can't replace it to do more testing... And I can't afford to buy a new one.

That's a defeat I guess 🙁

 

[80-watt Hamster]

I checked with brand new HyperX RAM and I keep having the same problem...
I guess is the secondhand motherboard that I boguht that maybe worked the first day but then stop working...
I don't have the old motherboard so I can't replace it to do more testing... And I can't afford to buy a new one.

That's a defeat I guess 🙁

That seems the most likely conclusion; sorry to hear that, mate. Is there much of a secondhand market where you live? Selling two sets of RAM (keep the HyperX) and the MB as for parts/not working could get you a good portion of the way to another board.

 

[Kyhz]

I think I found something wierd in the CPU pins... It looks like burnt pins:

In the mobo cpu coket I can't see any pint bent

Is it possible that from the start It was the CPU the problem and It damaged the old and the new motherboard?

I got any problem during Prime95 test for 2h but I got a crash during a Cinebench benchmark.

 

[80-watt Hamster]

I've never heard of a processor causing damage to a motherboard, but I'm by no means an expert, and anything's possible as they say. I don't see any bent pins in your pic, either, though it looks like there may be a couple bits of crud in there. Don't know what to say about the discoloration on the processor. Maybe it's the lighting, but is there corrosion on the pins? The retention screws look rusty.