PC shuts off with 2 sticks of ram after installing Ryzen 5 5600x

[chrcoluk]

Do you have any spare components to hand, cpu, board or other sticks of ram?

Something is faulty, potentially poor pin contact on CPU, so reseating might be worth trying, try one slot at a time, one stick at a time. That sort of thing.

 

[eidairaman1]

Well listen, my b550-a pro arrived but im waiting for an atx case cause this new one is an atx, so its gonna be around a week till i put it all together, ill let you know if thats what solved my problem, but considering you have a better psu and ram kit, yet you have the same symptoms, it’s probably that shitty asrock mobo thats causing the issue for the both of us. Also thanks for letting me know about the fact that 1866 doesn’t cause shutdowns, cause i still haven’t tried it yet, i haven’t tried anything below default speeds, maybe 2133 only but it used to make it shutdown regardless. i’ll give it a shot and see if it runs like that without shutting down at least for the time being. If that doesn’t solve it, i guess we’re both looking at a faulty cpu, but i honestly hope that’s not it.

You can test the parts out of the case for now using a non conductive surface such as a table

 

[stinnergfx]

Do you have any spare components to hand, cpu, board or other sticks of ram?

Something is faulty, potentially poor pin contact on CPU, so reseating might be worth trying, try one slot at a time, one stick at a time. That sort of thing.

well i do have the old cpu but its useless considering i've flashed my bios to the newer version which practically killed the support for it, and i cant even roll back to it, asrock doesnt let u flash old bios versions on this board if u upgrade

You can test the parts out of the case for now using a non conductive surface such as a table

i'll get some thermal paste tomorrow so i can test everything out for the time being i guess

 

[eidairaman1]

well i do have the old cpu but its useless considering i've flashed my bios to the newer version which practically killed the support for it, and i cant even roll back to it, asrock doesnt let u flash old bios versions on this board if u upgrade


i'll get some thermal paste tomorrow so i can test everything out for the time being i guess

Breadboarding takes alot of headache out

 

[senex]

I decided to see if there was a thermal component to this issue. So I opened the side of my computer, pointed a 2ft box fan at it, changed the setting to 2666 mhz. Normally this would shutdown at the 1:17 mark. But with the box fan it lasted 1:27... So it had an effect. I then tried 2400mhz with the fan, it got up to 2:39!. So there seems to be a thermal component.

I then started reviewing the board with my thermal camera. In the 1866 mhz situation, the VRMs to the left of the cpu where HOTTER (probably due to running longer). However there is a grey part to the right of the ram that got hotter at higher ram frequency. The part is PL15, which is an inductor (choke/coil). Under stable 1866mhz operation it got to 226f. But in the 2400mhz situation, right before the power off it hit 235f.


So I decided to re-run the 2400mhz test with various thermal changes:

• CPU fan at 100%: no change
• Pointed fan at VRMs to the left of CPU: No change
• Pointed fan above the CPU at VRMs: No change
• Pointed fan at top-right of motherboard: 3:40 (extra 2 minutes 20 seconds)
• Pointed fan at bottom-right of the motherboard 2:15

I am thinking the issue may be that PL15 gets too hot and causes the poweroff. It is a grey component that I dont recognize and is above the large motherboard connector. Maybe it didn't power off with 1 stick since there was a larger gap between it and the fan allowing air flow.

 

[eidairaman1]

I decided to see if there was a thermal component to this issue. So I opened the side of my computer, pointed a 2ft box fan at it, changed the setting to 2666 mhz. Normally this would shutdown at the 1:17 mark. But with the box fan it lasted 1:27... So it had an effect. I then tried 2400mhz with the fan, it got up to 2:39!. So there seems to be a thermal component.

I then started reviewing the board with my thermal camera. In the 1866 mhz situation, the VRMs to the left of the cpu where HOTTER (probably due to running longer). However there is a grey part to the right of the ram that got hotter at higher ram frequency. The part is PL15, which is an inductor (choke/coil). Under stable 1866mhz operation it got to 226f. But in the 2400mhz situation, right before the power off it hit 235f.


So I decided to re-run the 2400mhz test with various thermal changes:

• CPU fan at 100%: no change
• Pointed fan at VRMs to the left of CPU: No change
• Pointed fan above the CPU at VRMs: No change
• Pointed fan at top-right of motherboard: 3:40 (extra 2 minutes 20 seconds)
• Pointed fan at bottom-right of the motherboard 2:15

I am thinking the issue may be that PL15 gets too hot and causes the poweroff. It is a grey component that I dont recognize and is above the large motherboard connector. Maybe it didn't power off with 1 stick since there was a larger gap between it and the fan allowing air flow.

Add heatsinks to vrms and maybe have the coil replaced

 

[Princess Garnet]

I can say that this has been an interesting thread.

I had stability issues that started a year ago after I changed the graphics card (loss if display, and PC restarted), and while this was a common issue with that graphics card, I did have temperatures in mind as a cause somewhere. Everything I could monitor in software seemed alright, but I don't have the equipment or knowledge to check things to the extent you did. The reason I suspected RAM was I run four sticks of RAM and they are all dual rank and higher speeds, and oddly, I also saw stability change (get worse) if I ran at lower speeds. I did test with the side panel off, but not with a large fan. I stopped giving it thought because... an RMA on the graphics card solved the issue in all but one instance.

I'll be very interested to see if the new motherboard solves this for you (or, I don't know if you're still going down that route and getting a new one?). That CPU's IMC should more than handle those speeds, and the RAM itself is rated for it, and you ruled out bad RAM. I figured the motherboard or its BIOS and the settings it was trying were just not capable enough, but it seems you may have found the deeper problem.

A lot of people say AM4 motherboards are overbuilt on VRMs and got better compared to the late AM3 era where boards were burning up/throttling, but stuff like this is why I'm always skeptical on the very cheap motherboards. I mean, that's a 5600X and even at low RAM speed, it can't do it? It's not like it's a demanding ask of it or anything. I wonder if your individual board is faulty, or if that particular board is consistently poorly made in this regard.

 

[Zach_01]

A lot of people say AM4 motherboards are overbuilt on VRMs and got better compared to the late AM3 era where boards were burning up/throttling, but stuff like this is why I'm always skeptical on the very cheap motherboards. I mean, that's a 5600X and even at low RAM speed, it can't do it? It's not like it's a demanding ask of it or anything. I wonder if your individual board is faulty, or if that particular board is consistently poorly made in this regard.

I can say that only for B550 series maybe.
Up until X570 (included) a lot of boards have crappy VRM that cannot run properly 95~105W TDP (125~142W PPT) CPUs
And definitely some 300/400series boards are struggling even with lower tier CPUs

 

[stinnergfx]

Update : Problem is solved, I hooked everything up just enough to post the system and run a memory test, i have yet to deal with the cable management. The motherboard was the culprit, i’m no longer getting shutdowns on my tests, granted, i haven’t tried it with xmp profiles but the default ones work just fine

 

[Princess Garnet]

I can say that only for B550 series maybe.
Up until X570 (included) a lot of boards have crappy VRM that cannot run properly 95~105W TDP (125~142W PPT) CPUs
And definitely some 300/400series boards are struggling even with lower tier CPUs

Oh, yeah, I should have specified I was more referring to the 500 series chipsets, but I thought the older ones were good enough too.

 

[Speedyblupi]

Update : Problem is solved, I hooked everything up just enough to post the system and run a memory test, i have yet to deal with the cable management. The motherboard was the culprit, i’m no longer getting shutdowns on my tests, granted, i haven’t tried it with xmp profiles but the default ones work just fine

Good job on finding a fix.
Your old A320M-DVS R3.0 motherboard should have supported faster RAM in theory, but it has a weak memory topology (in plain English: the connection between the CPU and RAM slots is badly designed), which means that it doesn't always work reliably, because it's susceptible to electrical interference which can cause errors when the RAM is set to high speeds. B550 motherboards, especially ATX models like your new B550-A Pro, usually have much better memory topologies, which makes them much more likely to work with faster RAM than your old motherboard could handle. Your XMP profiles are very likely to work now (though not guaranteed - there's always some randomness in the quality of RAM modules and your CPU's memory controller. This is what people sometimes call the "silicon lottery").

 

[senex]

I'm happy you got it fixed with the motherboard swap. After all my testing and seeing your solution, I am pretty confident my memory speed stability issue is caused by my motherboard.

And as Speedyblupi said, it is likely due to memory topology. Maybe memory traces running to close, un-balanced lengths, near noisy components, components overheating... something like that that would become visible with better CPU.

This computer system that I have requiring memory to be at 1866 is really my kids, so he can live with it or buy a new motherboard and ebay his old one. May not be worth the $ for the boost in FPS for him. His choice.

Whats funny, is I upgraded the CPU thinking I could get an extra 3+ years of life while re-using everything... If I knew the motherboard would do this, I might have considered a different path... Those micro-center deals look pretty good...

 

[stinnergfx]

Good job on finding a fix.
Your old A320M-DVS R3.0 motherboard should have supported faster RAM in theory, but it has a weak memory topology (in plain English: the connection between the CPU and RAM slots is badly designed), which means that it doesn't always work reliably, because it's susceptible to electrical interference which can cause errors when the RAM is set to high speeds. B550 motherboards, especially ATX models like your new B550-A Pro, usually have much better memory topologies, which makes them much more likely to work with faster RAM than your old motherboard could handle. Your XMP profiles are very likely to work now (though not guaranteed - there's always some randomness in the quality of RAM modules and your CPU's memory controller. This is what people sometimes call the "silicon lottery").

Thank you, I'm running the xmp profile as we speak, its been like that for days now and there's no instability, so that works fine as well

I'm happy you got it fixed with the motherboard swap. After all my testing and seeing your solution, I am pretty confident my memory speed stability issue is caused by my motherboard.

And as Speedyblupi said, it is likely due to memory topology. Maybe memory traces running to close, un-balanced lengths, near noisy components, components overheating... something like that that would become visible with better CPU.

This computer system that I have requiring memory to be at 1866 is really my kids, so he can live with it or buy a new motherboard and ebay his old one. May not be worth the $ for the boost in FPS for him. His choice.

Whats funny, is I upgraded the CPU thinking I could get an extra 3+ years of life while re-using everything... If I knew the motherboard would do this, I might have considered a different path... Those micro-center deals look pretty good...

Appreciate it.

Well regardless of the fps boost, it will make future upgrades way easier, so even if everything were to work on that old one, a mobo is still the staple of all the components so its always better to not have to stress when you're doing any sort of upgrading in the future. Im not the type to care too much about some quirks and features a motherboard can have, as long as everything is stable. And now i have double the dimm slots incase i wanna run 32 gbs of ram or more (up to 128 i believe), and 2 m.2 slots that i didnt even have on my previous board, so its definitely something to consider if you're looking to upgrade any components in the future.

But I was in the same boat, i thought that the cpu change would be the end of my build and i wouldnt have to think about upgrading anything in the near future until this happened, and along with the motherboard i also had to buy a case which was just an extra expense but it looks damn good. Not to mention, I had no case fans on my previous one and the cpu is running 45-46 on idle atm, so it would probably cause overheating issues at one point.

 

[RJARRRPCGP]

MSI B550 A-Pro is another solid and cheap option

View attachment 363606
View attachment 363607

Far better than Pro VDH and not so much more expensive.

Hope that it doesn't develop the "literally-won't-turn-on" fault that seems to plague B450 Tomahawk.

Do you have any spare components to hand, cpu, board or other sticks of ram?

Something is faulty, potentially poor pin contact on CPU, so reseating might be worth trying, try one slot at a time, one stick at a time. That sort of thing.

If it's literally shutting off like with core temps being too high, then I suspect a short. Sounds like overcurrent.

Through MemTestX86 tests I see the following:
2x16gb @2400mhz+ shuts down in less than a minute
2x16gb @2133mhz shuts down in about 10 minutes
2x16gb @1866mhz doesn't shut down
1x16gb @3200mhz doesn't shut down

Looks like a rare circuitry fault. May be the result of a broken off SMD component. That's not a classic RAM fault.

Looks obvious that the motherboard is bad, because of having to run it at DDR3 speeds! It's like having a V7 instead of a V8, LOL.

 

[senex]

Since I found the component overheating (inductor next to the ram), I rigged a mini fan to cool it off. Now I get DDR4 2400 MHZ. I survived a MemTestX86 for 30+ minutes. Very stable