System won't power on at all after rebuild into new case

P

philb2

[H]ard|Gawd
Joined
May 26, 2021
Messages
1,846
Advice needed on troubleshooting completely dead system, after a rebuild.

I rebuilt my system using all the same components but in a new case, Phanteks Enthoo Pro. System uses an ASUS X570 Strix-E motherboard and contains 3 HDDs plus SSDs. It was working fine before I removed all the components from the old case.

When I press the power button, NOTHING happens. No drive noise, no motherboard LED progress lights, no Ryzen CPU fan, nothing. I even tried to short the Power buttons on the twin-row header with a screwdriver. Nothing. Zilch.

Is it possible that the CPU isn't seated properly and that explains the situation? When I inserted the CPU back into the AM4 socket, the socket would not lock the CPU into place. Doing a search, that's a common issue. So I inserted the CPU into the socket, and then reattached the heatsink, assuming that the heatsink would keep the CPU in place. I know that I'm on the hook for a socket replacement, eventually, according to various threads I read on this issue.

I have to say this all makes me look very bad on front of She Who Must Be Obeyed. :eek:
 
Check the processor orientation in the socket. You got to get that right for the socket to lock (Read the friendly manual). It should lock fairly easily. The CPU cooler is not used to hold the CPU in place.

Other likely cause would be a short circuit - motherboard to case.
Check the mounting screws carefully, and be sure the correct ones are used. Look for any rogue contact.

The common diagnostic in these instances is to assemble the motherboard out of the case - on a static free workbench - using just core components: power supply, motherboard, CPU, CPU cooler, one stick of RAM, graphics (if the CPU has graphics, use that), monitor, keyboard.
See if it boots. If it does then transfer the same setup to the case and test again. If it boots, add the rest of your components back one at a time and test again.
 
pendragon1 said:
if the socket arm isnt locked down the chip isnt making proper contact.
Click to expand...
The socket arm is locked down BUT the CPU is not tight in the socket. I did some searching and it seems that the AM4 socket is a bad design. You need to twist loose the HSF before trying to remove it from the CPU. I didn't know that when I removed the HSF.

So now I think I may be up s--- creek without a paddle. Not sure what to do next.
 
TV2 said:
Check the processor orientation in the socket. You got to get that right for the socket to lock (Read the friendly manual). It should lock fairly easily. The CPU cooler is not used to hold the CPU in place.

Other likely cause would be a short circuit - motherboard to case.
Check the mounting screws carefully, and be sure the correct ones are used. Look for any rogue contact.

The common diagnostic in these instances is to assemble the motherboard out of the case - on a static free workbench - using just core components: power supply, motherboard, CPU, CPU cooler, one stick of RAM, graphics (if the CPU has graphics, use that), monitor, keyboard.
See if it boots. If it does then transfer the same setup to the case and test again. If it boots, add the rest of your components back one at a time and test again.
Click to expand...
Yeah I'm going to do that tomorrow, using a sheet of cardboard to hold those components. I won't even use the NVidia GPU. I just want to see if the motherboard goes through the bootup steps and the "boot LEDs" light up. If that works, then I plug in the GPU and retry. If all that works, then I transfer the motherboard back into the case, and repeat.
 
philb2 said:
The socket arm is locked down BUT the CPU is not tight in the socket. I did some searching and it seems that the AM4 socket is a bad design. You need to twist loose the HSF before trying to remove it from the CPU. I didn't know that when I removed the HSF.

So now I think I may be up s--- creek without a paddle. Not sure what to do next.
Click to expand...
You should really pull the cpu and have a look at it and the socket before you go any further. Something is jacked up pretty bad if
A. Your CPU isn't seating properly.
And
B. You have a completely dead rig as a result.
Trying to run the system will more likely than not, make things worse.
 
lopoetve said:
Yeah that arm needs to be locked - why wouldn't it close?
Click to expand...
That's what I got stuck on.
Unless there is debris in/under the socket retaining lever, the pins are directional.
I thought AM4 and the older versions the retaining arm had to be down for the pins to make proper contact.
 
reseat the cpu and put pressure on the center of the cpu while lowering the arm. repaste and try again.
 
somebrains said:
That's what I got stuck on.
Unless there is debris in/under the socket retaining lever, the pins are directional.
I thought AM4 and the older versions the retaining arm had to be down for the pins to make proper contact.
Click to expand...
somebrains

You are right. I inspected the AM4 socket and saw how retaining clips (?) moved into position when the retaining arm was down. That said, because of the issue with the AM4 socket, even after the retaining arm was done, the CPU was not held in place.
 
philb2 said:
somebrains

You are right. I inspected the AM4 socket and saw how retaining clips (?) moved into position when the retaining arm was down. That said, because of the issue with the AM4 socket, even after the retaining arm was done, the CPU was not held in place.
Click to expand...

I would stop messing with it if the cpu is properly seated.
I would move past this idea that the AM4 cpu pins must have more retention relative to a cpu cooler + paste pulling it.

I have heated and rotated enough AM4 cpu coolers off the cpu with no damage to any component.
That's enough retention as far as I'm concerned.
 
philb2 said:
somebrains

You are right. I inspected the AM4 socket and saw how retaining clips (?) moved into position when the retaining arm was down. That said, because of the issue with the AM4 socket, even after the retaining arm was done, the CPU was not held in place.
Click to expand...

Then you are doing something wrong, there are not issues with the AM4 socket if your cpu is put in properly, You seat the CPU, properly, and put down the arms and it locks into place and the CPU can not move at all, if it is moving, you either broke something or have something wrong.
 
... juuust to be sure mate:

Would you kindly confirm, that you have checked the psu? Like the on/off, (and very carefully the) 120/240watt button?
It simply hasn`t been mentioned yet, so...

Rebuilding a minimal rig outside is the way, like other wise members already have mentioned.


And If you need a real life story to get inspired:

My mate Ben did have wrapping paper of a chewing gum fall between his new case and the motherboard.
THAT was not on anyones list of things that could go wrong whilst building a new rig.
So keep an open mind ;)

Good luck!
 
Last edited:
Im going to make a wild guess here but I would bet he put the cpu in the socket the wrong way. Likely smashed a bunch of pins flat so it wont go into the socket completely. Might not have noticed there is an arrow on the cpu and socket that you line up and it goes right in the right way.
 
Yeah, we aren't getting the whole story here. My guess is he ripped out the CPU trying to get the cooler off on the last go-round. He would have responded to have inspected the CPU for bent/broken pins by now.
 
mothman said:
Yeah, we aren't getting the whole story here. My guess is he ripped out the CPU trying to get the cooler off on the last go-round. He would have responded to have inspected the CPU for bent/broken pins by now.
Click to expand...
Well, when I first removed the HSF, there were no bent CPU pins. Then I replaced the HSF so I could test the motherboard. (Still dead.) Only this time, I heeded all the advice and twisted the HSF carefully before removing it. Damn, would you know it, now the CPU does have bent pins.

Can I, who has no special skills and not so great eyesight, fix those pins? Or is this a sign from heaven telling me that I should really replace my 3900X with a 5900X CPU?
 
philb2 said:
Well, when I first removed the HSF, there were no bent CPU pins. Then I replaced the HSF so I could test the motherboard. (Still dead.) Only this time, I heeded all the advice and twisted the HSF carefully before removing it. Damn, would you know it, now the CPU does have bent pins.

Can I, who has no special skills and not so great eyesight, fix those pins? Or is this a sign from heaven telling me that I should really replace my 3900X with a 5900X CPU?
Click to expand...
If it wasn't locked down, yeah- twisting will bend the pins . You can straighten them - generally a mechanical pencil with the lead out will do it, or a ballpoint pen (been YEARS since I've had to). Or just upgrade. But figure out why that socket wouldn't close!
 
philb2 said:
Well, when I first removed the HSF, there were no bent CPU pins. Then I replaced the HSF so I could test the motherboard. (Still dead.) Only this time, I heeded all the advice and twisted the HSF carefully before removing it. Damn, would you know it, now the CPU does have bent pins.

Can I, who has no special skills and not so great eyesight, fix those pins? Or is this a sign from heaven telling me that I should really replace my 3900X with a 5900X CPU?
Click to expand...
You can try. Sometimes it's not too hard, other times it can be really difficult. I've done many, if you need help just message me.
 
You must log in or register to reply here.
Back
Top