One (Hopefully) Last Problem

[Brandon Shinobu]

Hello! 

 

Some of you may remember me from This Post where I was tracking down a system freeze. I'm happy to say that after correcting the memory speed as mentioned, the freezing problem is completely gone. Unfortunately, a new wrinkle has appeared that hasn't happened at all until after that saga, and once more I'm not sure how to track it down. I've re-included my build notes below.

 

Build

• Asus Zenith II Extreme Alpha (motherboard)
• 3960x AMD Threadripper
• IceGiant Elite Prosiphon CPU (cooler)
• 128Gb G.Skill RAM, 3200Mhz 14-15-15-35
• 1600w Leadex SuperFlower PSU 80+ Ti
• Lian-Li 24-pin power Strimer cable extension (power struggled to reach without an extension)
• CoolerMaster 90-degree 24-pin power adapter with built-in capacitor (bought to reduce strain on the 24-pin mobo connector
• 3 M.2 NVMEs. Samsung 980 Pro 2Tb x2 and 250Gb x1

 

Issue

 

Basically, the system just suddenly reboots. No blue screen, no windows errors in the Event log, no errors when the system boots back up, it just suddenly shuts off and then boots back into Windows. It;s happened in games, it's happened when running chrome, it's happened when it was just sitting idle doing nothing. It doesn't happen overly frequently. So far it's happened one to three times a day. I can have long sessions of using heavier programs and it'll be fine, then reboot. I haven't been able to figure out any specific trigger.

 

I should emphasize that this is the very first time this issue has cropped up.. While yes, my computer was having freezing issues before, this problem only appeared after I reinstalled my GPU. I had RMA'd my GPU thinking it was the problem, found out it was the RAM, fixed the RAM, then reinstalled the GPU when the manufacturer sent it back. In worse packaging than I sent it in.

 

Anyway, thank you all for the help; hopefully this is just a setting or something somewhere I can correct and be done with, but I appreciate any advice/thoughts to help me figure it out.

[TheStyne]

Sounds like a power delivery issue, given that it suddenly reboots without warning. Usually if it was some other component at fault you would get a blue screen or freeze, or something. I've personally never heard of Super Flower as a power supply brand, but that one looks like it's in the A Tier on the forum so it should be a good model. I don't suppose you have another power supply to try?

[Brandon Shinobu]

5 minutes ago, TheStyne said:

Sounds like a power delivery issue, given that it suddenly reboots without warning. Usually if it was some other component at fault you would get a blue screen or freeze, or something. I've personally never heard of Super Flower as a power supply brand, but that one looks like it's in the A Tier on the forum so it should be a good model. I don't suppose you have another power supply to try?

Only a very old 1000W PSU from my last computer. It's about 11-12 years old now. I'm not even sure I have the connectors to hook it all up. Could it possibly be the 90 degree CoolerMaster adapter that's causing the problem somehow? Normally I'd just pull it out to test, but the Prosiphon is difficult to get underneath, so I'd probably end up re-mounting it on the CPU just to get to it.

 

I likewise hadn't heard of SuperFlower before, but most places I looked at swear by them, and apparently they also provide some of the PSUs for brands like EVGA, etc.

[TheStyne]

1 minute ago, Brandon Shinobu said:

Only a very old 1000W PSU from my last computer. It's about 11-12 years old now. I'm not even sure I have the connectors to hook it all up. Could it possibly be the 90 degree CoolerMaster adapter that's causing the problem somehow? Normally I'd just pull it out to test, but the Prosiphon is difficult to get underneath, so I'd probably end up re-mounting it on the CPU just to get to it.

It certainly could be, but no way to be 100% sure without checking. Could also be the Strimer, no idea what the quality of that thing is like. Best bet would honestly be to hook all power supply cables directly into the components and give it a try, just to rule out any variables. I know it's a hassle, but that would be my personal next step  

[Brandon Shinobu]

Just now, TheStyne said:

It certainly could be, but no way to be 100% sure without checking. Could also be the Strimer, no idea what the quality of that thing is like. Best bet would honestly be to hook all power supply cables directly into the components and give it a try, just to rule out any variables. I know it's a hassle, but that would be my personal next step  

I'll see if I can. Unfortunately, the reason for purchasing the Strimmer (yes yes, I did get one with LEDs, I admit) was because I can't fit the 24-pin connector through the grommet because of the motherboard size, so it can't reach. I'll have to see if there's some other way to connect it. I may try with the Strimmer first just because it's easier to test that way, and then look for a way to remove the Strimmer if the problem is still there.

 

Supposing I do bypass all non-essential connectors and the problem still exists, would that indicate the PSU, or could power to some other component like the GPU be causing it?

[RollinLower]

1 minute ago, Brandon Shinobu said:

I'll see if I can. Unfortunately, the reason for purchasing the Strimmer (yes yes, I did get one with LEDs, I admit) was because I can't fit the 24-pin connector through the grommet because of the motherboard size, so it can't reach. I'll have to see if there's some other way to connect it. I may try with the Strimmer first just because it's easier to test that way, and then look for a way to remove the Strimmer if the problem is still there.

 

Supposing I do bypass all non-essential connectors and the problem still exists, would that indicate the PSU, or could power to some other component like the GPU be causing it?

i suspect the PSU. if power only drops to your GPU, you would get a display driver crash/freeze, but i don't think the system would shut down completely like this.

[TheStyne]

I'm not an expert, but in my experience when you have a faulty GPU, CPU, RAM, etc then the problem presents itself through a blue screen or freeze, as windows tries to pick up the pieces of the missing hardware. In this case, since the whole computer shuts off I really suspect a power delivery problem, either in the 24 pin or the entire power supply. I suppose it could also be some kind of problem with the wiring in your home, so you could also try another outlet and make sure it isn't connected to a power strip or extension cord. I doubt it, though, considering it happens randomly as opposed to when under load.

[Brandon Shinobu]

5 minutes ago, RollinLower said:

i suspect the PSU. if power only drops to your GPU, you would get a display driver crash/freeze, but i don't think the system would shut down completely like this.

 

2 minutes ago, TheStyne said:

I'm not an expert, but in my experience when you have a faulty GPU, CPU, RAM, etc then the problem presents itself through a blue screen or freeze, as windows tries to pick up the pieces of the missing hardware. In this case, since the whole computer shuts off I really suspect a power delivery problem, either in the 24 pin or the entire power supply. I suppose it could also be some kind of problem with the wiring in your home, so you could also try another outlet and make sure it isn't connected to a power strip or extension cord. I doubt it, though, considering it happens randomly as opposed to when under load.

Sounds good. Well, sounds like a plan, heh. I'll see what I can find out through this. I do have my computer hooked up to a UPS ( CP1500AVRLCD ), and I can't help but note the UPS is for 1500w, and the PSU is 1600w, but I'd think if that were a problem it would have happened before now.

 

Granted... before now I didn't have a power-hungry GPU constantly hooked up. The lights do flicker occasionally in my apartment, but I don't know what causes that. It's been very hot here lately, so I figure it's fluctuations from people using it up.

[RollinLower]

1 minute ago, Brandon Shinobu said:

The lights do flicker occasionally in my apartment, but I don't know what causes that. It's been very hot here lately, so I figure it's fluctuations from people using it up.

well, if anything your UPS should help to even that out.  and your rig should not be drawing 1600 watts in this config, or you must be pushing some bonkers overclocks lol.

superflower is a legit brand though, as you mentioned, they are the OEM for some high end brands. 

[Brandon Shinobu]

5 minutes ago, RollinLower said:

well, if anything your UPS should help to even that out.  and your rig should not be drawing 1600 watts in this config, or you must be pushing some bonkers overclocks lol.

superflower is a legit brand though, as you mentioned, they are the OEM for some high end brands. 

No OC'ing going on at all, aside from the built-in OC'ing that the AMD stuff does on its own, Precision Overdrive Boost, or something? The GPU is a 3090 that... yeah, didn't come cheaply. It was only after I had hooked that back up this problem seemed to start. I built a system that I hoped would last me another 6+ years, heh.

[Guest]

1 hour ago, Brandon Shinobu said:

 

• Lian-Li 24-pin power Strimer cable extension (power struggled to reach without an extension)
• CoolerMaster 90-degree 24-pin power adapter with built-in capacitor (bought to reduce strain on the 24-pin mobo connector

So have you tried it WITHOUT all of this extension?  This is the first place I would start.