Question New AM5 build just died… now up and running

  • Thread starter Deleted member 2838871
  • Start date
D

Deleted member 2838871

Guest
Ryzen 9 7950x3D
64GB Trident Neo DDR5 6000
Gigabyte AERO 4090
EVGA G+1000W
MSI Mag X670E Tomahawk WiFi

System has been fine for 3 days. Just now I was on steam playing card games… I heard a faint popping sound coming from PC… Orange debug LED came on and lost display signal.

I powered off and now when I power on I see all the lights and fans come on for 1 second with red/orange debug LEDs and system shuts off.

For some reason I’m not even upset. 😂😂 Someone mentioned growing pains with a new platform and here they are.

I actually have a EVGA 1300W PSU arriving today. This 1000W unit is from previous build. It’s 2 years old and have had no problems. PCPP build sheet shows about 900W so I decided yesterday to bump it up to 1300W and then this happens. Weird.

Suggestions? This is the first build I’ve ever had a problem with and I have no idea where to begin.

Thanks!
 
Odd thing is this is the third post this morning I've seen where a hardware component has failed. It's almost as if something in the atmosphere swept over the earth last night and caused hardware to fail.

What does the LED debug info say in the mobo manual? (Although, it isn't always accurate.)
 
D

Deleted member 2838871

Guest
Odd thing is this is the third post this morning I've seen where a hardware component has failed. It's almost as if something in the atmosphere swept over the earth last night and caused hardware to fail.

What does the LED debug info say in the mobo manual? (Although, it isn't always accurate.)

Red is CPU and yellow is ram. It looked orange but maybe it was yellow. Definitely wasn’t red.

I went back to bed with it unplugged for the last 2 hours and got up and checked and it’s still doing the same thing.

I press power button… fans and lights come on for a fast second and it turns off. I see the red and yellow leds on mobo.

My first thought is dead mobo but I could be totally wrong. Never had to diagnose a problem like this. As said a new PSU is arriving today and I could throw a new mobo at it too if needed.
 
D

Deleted member 2838871

Guest
Popping sound is probably the PSU.

‘That would really be weird then considering I ordered a new one 2 days ago.

it was a very faint pop noise and it sounded like it came from the mobo and not from the PSU. The mobo led immediately came on and display signal was lost. So I powered down and am now caught in a won’t power on loop… but I am seeing the lights and fans spin for 1 second when I press the button.

Anyway… since new PSU will be here today I went ahead and ordered the same mobo that will also arrive today.

Will swap PSU first and if that doesn’t work will reseat ram… cpu… etc and replace mobo if all else fails.

Will follow up.
 
D

Deleted member 2838871

Guest
Seems like I’ve read about a couple of failures on am5 in the last couple of days. Seems like both used msi boards. Haven’t used am5 but it seems like I remember seeing that msi boards that generation were having issues but I could be mistaken.

I‘ve always used MSI boards and this is the first issue I’ve ever had. As said I ordered another with Amazon same day delivery so will work on the PC this evening and see if I can resolve it.

Strange that it would work fine for 3 days and then go bad.

I just removed the GPU and that didn’t help. Power button I am hearing a click coming from PSU and the lights fans come on for a fast second then shuts off.

Will swap PSU later… and if that doesn’t work… the mobo.
 
Might reset the cmos. Not to say that msi makes bad products but this series could be having issues. MSI has a very good rep.

The fact that you say the fans and lights come on for about 3 seconds and you hear the psu click, sounds to me that when you hit the power button that power is going through the board and giving the psu a signal to turn on and that it starts to. So definitely could be power related as well. If your psu is 5 years old, the psu is the place to start.
 
D

Deleted member 2838871

Guest
Might reset the cmos. Not to say that msi makes bad products but this series could be having issues. MSI has a very good rep.

The fact that you say the fans and lights come on for about 3 seconds and you hear the psu click, sounds to me that when you hit the power button that power is going through the board and giving the psu a signal to turn on and that it starts to. So definitely could be power related as well. If your psu is 5 years old, the psu is the place to start.

Yeah I’m definitely not giving up on MSI yet. Always been happy with their boards.

Hitting the power button turns on the lights and fans for one second with the click from PSU and then shutdown.

The PSU is a EVGA G+ 1000W Gold… new with my Jan 2021 build. I moved it to this new build thinking it would suffice being relatively new. My PCPP parts list shows 900W.

The other day after building PC I thought better of it… wanting a bit more room so I ordered a 1300W unit.

Ironic that these issues happen after I did that. At any rate… will do the PSU swap first and go from there… with a motherboard swap being the last option.

Thanks… will follow up.
 
D

Deleted member 2838871

Guest
Very strange issue:

Well the good news is that doesn’t appear to be the case here…. But who knows.

The 1300W PSU arrived and I did a quick swap with the mobo and cpu cables and nothing changed.

So I disassembled everything and am now waiting on the new mobo to arrive later today. Going to work for a few hours and will tackle it later this evening.

All parts and the socket look completely normal… I see no indications of anything being burnt up like seen in the link pics. Thermal paste on CPU and Noctua cooler looked good.

I was going to have to do some disassembly this weekend anyway to install the PCIE extender and NVME card so I can run 8 SSDs but wasn’t expecting to have to rebuild the entire PC. It is what it is.

Since I am basically starting from scratch if anyone has any tips to rule out any hardware like booting with 1 ram stick… etc… please chime in. Never had to troubleshoot before.

GPU has been ruled out and so has the 1000W PSU I believe.

So… CPU? Mobo? Ram? Couldn’t be anything else could it?
 
  • Like
Reactions: Dark Lord of Tech
So… CPU? Mobo? Ram? Couldn’t be anything else could it?
Since you heard a pop there will be visual evidence of what popped, it's just going to take a long time to look everything over, it's most probably going to be a very small SMD, cap or diode, on the mobo that will have a small chip in it from material that chipped off from the pop if not have a visible scorch mark.
Sometimes the part can completely disintegrate leaving an empty space where a component should be, this can be harder to identify if you don't have much experience with what should be where.
 

laxman10100

Distinguished
Oct 27, 2012
11
3
18,515
Hopefully it's the PSU!

900W seems like a crap ton of power to need for that build. My 4090 & 12900K build never went above C. 600W of power usage under load. I am not familiar with AMD CPU power usage though; so perhaps it would use more.

Anyway, give us an update on what happens once you plop in that new PSU. Hopefully it works again and your CPU/motherboard didn't fail like other have recently mentioned regarding their X3D chips. Last week I swapped out my 12900K & motherboard for a 7800X3D for better gaming performance and less power usage....and now I'm starting to sweat a little lol.
 
D

Deleted member 2838871

Guest
Hopefully it's the PSU!

900W seems like a crap ton of power to need for that build. My 4090 & 12900K build never went above C. 600W of power usage under load. I am not familiar with AMD CPU power usage though; so perhaps it would use more.

Anyway, give us an update on what happens once you plop in that new PSU. Hopefully it works again and your CPU/motherboard didn't fail like other have recently mentioned regarding their X3D chips. Last week I swapped out my 12900K & motherboard for a 7800X3D for better gaming performance and less power usage....and now I'm starting to sweat a little lol.

If only you saw the parts list… 😂😂

8 SSDs 2 HDDs 4090 and a partridge in a pear tree.

Results are in…

old mobo and cpu = dead

System powers on fine with new mobo and no cpu and remains dead with old mobo and no cpu.

That being said… Anyone have any thoughts on which CPU I should go with? Stay with 7950x3D? Go with 7950x?

70/30 gaming productivity which is mainly video/editing and encoding of 6k drone footage and blu ray discs.
 
  • Like
Reactions: MrN1ce9uy

laxman10100

Distinguished
Oct 27, 2012
11
3
18,515
If only you saw the parts list… 😂😂

8 SSDs 2 HDDs 4090 and a partridge in a pear tree.

Results are in…

old mobo and cpu = dead
Hmmm. That still seems like it is a bit too high for the power draw. Eight SSDs and two mechanical hard drives would only pull a total power draw of around 60-70W. And my max power draw - also with a 4090, 12900K, one SATA SSD and three NVME drives (around 40-50W total draw) - was only around 600-650W maximum.

Regardless, I'm sorry that your CPU and motherboard are the culprits here =(.
Have you looked at the CPU and motherboard's CPU socket to see if there is any bulging or deformation on either of those two parts?

Also, have you tried clearing the motherboard's CMOS, rebooting it and then waiting 2-3 minutes to see if it still allows you to enter the BIOS? I have noticed on my recent switch to AMD (after being with Intel for so long) that when the memory timings and voltages are not optimal for that motherboard, it can take a LOOOONG time to boot whilst training the memory configuration. And if it then fails to boot, the "Dr. Debug" screen fails too and the system never posts and never allows me to enter the BIOS. But once I clear the CMOS and reboot, after about 2-3 minutes I can re-enter the BIOS and re-configure everything again in order to get it working. But if I just power down, turn off the PSU, unplug the PSU and wait 10-15 minutes before hooking it all back up; it still doesn't allow me to enter the BIOS. I have to clear the CMOS for it to allow me to access the BIOS again.

I never had to do anything like that with my Intel builds. But I guess there was just a bit more AMD-related motherboard configuration stuff to learn before I could get it all up and running reliably.

I've been building, updating, maintaining and optimizing computers (and computer builds) for around 12-13 years now - so the fact that it took me 3-4 hours to figure it all out had me a bit worried. So, if you're not super familiar with all of the available hardware, then I can see how stuff like this could cause problematic issues for folks.

*Edit* - Sorry for all of the edits. I kept remembering more information to jot down after I posted the comment, haha. My apologies.
 
Last edited:
If only you saw the parts list… 😂😂

8 SSDs 2 HDDs 4090 and a partridge in a pear tree.

Results are in…

old mobo and cpu = dead

System powers on fine with new mobo and no cpu and remains dead with old mobo and no cpu.

That being said… Anyone have any thoughts on which CPU I should go with? Stay with 7950x3D? Go with 7950x?

70/30 gaming productivity which is mainly video/editing and encoding of 6k drone footage and blu ray discs.
Looking at your specs at the bottom of your signature are these all SATA SSD's?

4x 2TB Samsung 990 Pro / 4x 2TB Samsung 970 Evo Plus
 
D

Deleted member 2838871

Guest
Hmmm. That still seems like it is a bit too high for the power draw. Eight SSDs and two mechanical hard drives would only pull a total power draw of around 60-70W. And my max power draw - also with a 4090, 12900K, one SATA SSD and three NVME drives (around 40-50W total draw) - was only around 600-650W maximum.

I dunno… just going by what PCPP says.

Anyway… I do think my 2 year old 1000W PSU is probably cutting it a bit close so I went 1300W.

Regardless, I'm sorry that your CPU and motherboard are the culprits here =(.
Have you looked at the CPU and motherboard's CPU socket to see if there is any bulging or deformation on either of those two parts?

Yes. The mobo and cpu and cpu socket look completely normal. Can post pics if you want.

Also, have you tried clearing the motherboard's CMOS, rebooting it and then waiting 2-3 minutes to see if it still allows you to enter the BIOS? I have noticed on my recent switch to AMD (after being with Intel for so long) that when the memory timings and voltages are not optimal for that motherboard, it can take a LOOOONG time to boot whilst training the memory configuration. And if it then fails to boot, the "Dr. Debug" screen fails too and the system never posts and never allows me to enter the BIOS. But once I clear the CMOS and reboot, after about 2-3 minutes I can re-enter the BIOS and re-configure everything again in order to get it working. But if I just power down, turn off the PSU, unplug the PSU and wait 10-15 minutes before hooking it all back up; it still doesn't allow me to enter the BIOS. I have to clear the CMOS for it to allow me to access the BIOS again.

No I didn’t. After hearing a noise coming from mobo the display went dead and PC shut off… further attempts to power on resulted in a quick spin of fans and lights coming on for one entire second then immediate shut off.

Only instance where PC powered up was the new mobo without the CPU installed.

It did take 3-4 minutes on the initial boot the other day and then ran fine for 3 days prior to death.

I never had to do anything like that with my Intel builds. But I guess there was just a bit more AMD-related motherboard configuration stuff to learn before I could get it all up and running reliably.

I've been building, updating, maintaining and optimizing computers (and computer builds) for around 12-13 years now - so the fact that it took me 3-4 hours to figure it all out had me a bit worried. So, if you're not super familiar with all of the available hardware, then I can see how stuff like this could cause problematic issues for folks.

*Edit* - Sorry for all of the edits. I kept remembering more information to jot down after I posted the comment, haha. My apologies.

This is the 5th PC I’ve built… but the first one I’ve ever had a issue with afterwards.

Thanks for the assistance!
 
Last edited by a moderator:
  • Like
Reactions: laxman10100
D

Deleted member 2838871

Guest
Anyway, give us an update on what happens

I'm up and running. New mobo and CPU... system informed me of the CPU change and then booted into Windows normally. Initial post took about 30 seconds.
 
D

Deleted member 2838871

Guest
Looking at your specs at the bottom of your signature are these all SATA SSD's?

4x 2TB Samsung 990 Pro / 4x 2TB Samsung 970 Evo Plus

Yes. The Gen 4 990’s on the mobo and the Gen 3 970’s in a PCIE NVME enclosure.


I am also using a PCIE extender cable to avoid having to use the slot that shares bandwidth with the 4th m.2 on mobo.



I made a post about it the other day in the storage forum asking what my options were because I wanted to keep the storage space rather than sell the 970s for $50 each if I even got that. They retail for $129 new so it made more sense to keep them.

Anyway, give us an update on what happens

I'm up and running. New mobo and CPU... system informed me of the CPU change and then booted into Windows normally. Initial post took about 30 seconds. Subsequent reboot from restart to Windows desktop took 1 minute.
 
I'm up and running. New mobo and CPU... system informed me of the CPU change and then booted into Windows normally. Initial post took about 30 seconds. Subsequent reboot from restart to Windows desktop took 1 minute.
Did you go into bios to lock down the vcore to a sufficiently low level or are you rolling with it?!
Also RMA or did you pay for the new stuff again?
 
D

Deleted member 2838871

Guest
Did you go into bios to lock down the vcore to a sufficiently low level or are you rolling with it?!
Also RMA or did you pay for the new stuff again?

I didn’t make any changes other than NOT updating to the beta BIOS version like I did the first time. I went with the BIOS the board came with.

Still running default settings… only change was enabling the ram speed. If it’s a good idea to drop the vcore I can definitely do that.

Yes the cpu and mobo have been RMA’ed.
 
  • Like
Reactions: Why_Me

laxman10100

Distinguished
Oct 27, 2012
11
3
18,515
I didn’t make any changes other than NOT updating to the beta BIOS version like I did the first time. I went with the BIOS the board came with.

Still running default settings… only change was enabling the ram speed. If it’s a good idea to drop the vcore I can definitely do that.
You've probably seen this by now, but AMD has stated to ensure that your motherboard does not allow the SoC_Voltage to go over 1.3V.

It seems that the ASRock motherboards (as I have the Taichi Carrara) never pushed the SoC_Voltage over 1.3V with or without EXPO enabled. My X670E has always just sat at 1.3V after enabling EXPO.

Granted, I also use the -30uv/75° setting and so my SoC_Voltage has never really gone over 1.15V lol. But now I am wondering if ASRock users may have to adjust their Memory_Controller_Voltages, as ASRock boards set that to either 1.35V or 1.4V when EXPO is enabled...

I suppose I can try setting those to 1.3V tonight and seeing if it still posts and stays stable.
 
D

Deleted member 2838871

Guest
Did you go into bios to lock down the vcore to a sufficiently low level or are you rolling with it?!
Also RMA or did you pay for the new stuff again?
Just got the email from NewEgg that my dead parts were received and RMA approved.

(y)
 
  • Like
Reactions: laxman10100
D

Deleted member 2838871

Guest
Well I am officially good to go. After building on 4/19 with parts ordered from Newegg... after the implosion on 4/22 I ordered a new cpu and mobo from Amazon... received the next day and avoided having to wait on the RMA.

PC has been running fine since... updated to new BIOS on 4/28 and am waiting on more updates.

I received the RMA replacement parts from Newegg yesterday... and just returned them to Amazon and got my $1000 back that has been in limbo for a week.

Game on. (y)