Asus strix b450 itx with 3700x random shutdowns.

B

bufodr_T

Supreme [H]ardness
Joined
Nov 5, 2005
Messages
6,157
Bios version 2501

Systems works fine. Then all of the sudden shuts down. The rgb lights on the back of the motherboard are lit but system doesnt turn on til i unplug the power cord, hold down the power button, plug the power cord back and then it boots and runs like normal. Then it randomly does it again.
Once while i was using premiere and handbrake at the same time the other while i was playing gta v

Bought the mobo off another member here and i dont think its the problem Cpu is new. Ram is good as never had a problem in my old system with it.
Specs
Asus b450i itx mobo
ryzen 3700x with stock wraith max cooler
evga 2080ti black
16gb corsair vengeance ddr4 3000 cl15
1tb wd m.2 with windows 10 pro
2tb seagate firecude
InWin A1 plus with stock gold rated 650 watt power supply.

I tend to believe the power supply is a dud. I ran the 2080ti with an 8700k on a sf600 with no issues.
What do you guys think? I have ordered a new sf600 to replace the stock one to try and will be here sunday.
I havent done a clean install of windows yet although i have installed all the drivers for the am4 motherboard.
Would the motherboard be overvolting and cause this?
 
Just noticed something else.
The Shadow of the Tomb Raider plays fine.
As soon as i change the setting to HBAO it instantly shuts off and i again have to unplug the power cable and plug it back in to reboot.
 
Makes me wonder if it is where i havent done a clean install of windows. Maybe is should try that first.
 
Can you find any threads on Asus CS forum or anywhere else with the same cpu/mobo citing similar issues?
Guess after the new psu you’ll know.
 
All symptoms point to a bad psu, but it shutting down the instant i click HBAO in the setting menu gets me. Not once, but EVERY time.
 
Have you tried updating chipset drivers to the one dated July 30?

It fixed some nvidia related crashes and while it seems like a long shot, its probably worth a try before swapping out more hardware.
 
just reinstalled windows and still does it. When running cpu intensive tasks alone it performs awesome with no shutdown. Only when it is gpu intensive and added pwer it seems to shutdwn. I have a cosair power supply arriving tomorrow.
 
bufodr_T said:
All symptoms point to a bad psu, but it shutting down the instant i click HBAO in the setting menu gets me. Not once, but EVERY time.
Click to expand...

Probably draws slightly more power from the PSU causing the shutdown. I would think the Corsair PSU should be somewhat better.
 
Definitely the psu. Took a new 550 watt antec gold psu, connected it to the gpu, and used a paper clip to power it on. Been going under full load for about 3 hours and no shutdowns.
 
Makes perfect sense....Soon as the gpu started drawing higher power it turns off. Any power supply can go bad....but ill be honest Inwin wouldn't be my first choice for a setup of mine:) You can put it back together now
 
Keep in mind that it might not be a bad PSU, but rather the PSU is designed with a limit of OCP (overcurrent protection) that is below what the computer is asking for.
In other words, a brand new PSU, that is not faulty, may exhibit the same symptoms: Total shutdown under certain loads.

These new systems with more and more cores have very high instantaneous current needs - you've got 8 cores / 16 threads and thousands of GPU "cores" all possibly ramping up within milliseconds of each other (such as clicking on "enable HBAO" in a game).
This puts an incredible strain on the PSU and might trigger the OCP which results in full and instantaneous shutdown.

If the PSU itself is not faulty, then the only solution is to buy a beefier PSU that has higher OCP in its design or find a similar wattage PSU that has a higher OCP limit. It really has nothing to do with the PSU wattage rating, though in general a higher wattage PSU will have a higher OCP limit simply due to it being higher wattage.
For a real world example, the computer listed in my signature below doesn't "need" a 1000W PSU because under all loads it uses less than 750W average at the wall according to a Kill-a-Watt meter. But, the system would suddenly power off under certain loads/games (reproducible every time) when using the 850W version of the same brand/model of PSU, and I tried two new PSUs before giving up. After switching to the 1000W model, the shutdowns stopped. Perhaps I could have bought another PSU brand in the 850W range and fixed the shutdown problem as well, but without knowing what the OCP limits are, it would be a crapshoot. I wanted to continue to use the Prime Titanium series so I went up to the next wattage rating instead.

Yes, it sucks because you end up spending more money than you should need to spend and you certainly do not need the higher wattage. But it's the OCP, which is a hidden thing no PSU manufacturer tells you any details about that gets you. Not to mention that GPUs/CPUs/MB manufacturers don't tell you what kinds of current draw spikes they can generate so you are left in the dark, literally.
This kind of thing is only going to get worse as more and more cores are added to CPUs and GPUs. Manufacturers will need to address it in time. I don't know if it is a lack of enough capacitors in both the PSU and MB/video card or what, but these current spike shutdowns are getting more common these days.
 
Ok fellas. Did it again just now when going to play some gta v. Now i know its not the power supply.THis time it wouldnt post or power on for at least 30 minutes. Then it finally posted an ran. Im kind of at a loss of ideas for now. That leads me to think its a bad mobo or cpu. Opinions? I can buy an x570 itx board if i need to, i just dont want to keep buying shit lol
 
question are you running the 2080ti off 2 8 pins on the same cable or two different cables? still sounds like OCP is being triggered. one thing you could try is underclock the gpu to say 1350mhz see if the system crashes again. at those clocks is should not trip ocp. only other thing i'd say is check your cpu voltage and see if it's too high, possible the vrm's are tripping some sort of protection.
 
Last edited:
When i tested it a few minutes ago outside the case i used a 550 watt antec gold power supply. I just tried it with the sf600 and the only thing that lights up is the rgb and standby led. I plug the antec 550 watt psu i have and it boots fine. Whatever it is has apparently fried the SF600. I ran an overclocked 8700k at 4.7 and the same 2080ti with no issues. Good thing i ordered the psu from amazon.
 
The 2080ti is an evga black edition at stock clock. The cpu is a stock clocked 3700k. No way this system is using enough juice to fry a sf600.
 
I don't know what it is...old unit looked at internally...like chip OCP/UVP/OVP...some of that bullshit went bad for no reason...start and just quit...id at least look at the rails though.
 
I just upgraded to the same mobo and a 3600. They did just release a new mobo firmware a few days ago I believe that I upgraded to.
 
I updated to the newest one last week and it appears to have fixed the issue. AFter that GTA V wouldnt run. I had to remove MSI afterburner and install the beta so that i could oc my 2070s and play GTA V.
 
bufodr_T said:
I updated to the newest one last week and it appears to have fixed the issue. AFter that GTA V wouldnt run. I had to remove MSI afterburner and install the beta so that i could oc my 2070s and play GTA V.
Click to expand...
awesome! Yea, I might be switching back to a nvidia card from my 580 as warcraft has minor freezing issues upon start up.
 
One thing about power supplies you need to know is it single 12v rail amps and how much does your 2080Ti need under full 3D load and overclocked ..
 
Looks like they just released a new BIOS firmware over a week ago. It says it improves system stability among other things. At least ASUS has been rolling out the firmware updates quickly for this model.
 
You must log in or register to reply here.
Back
Top