Seeking advice - Crash issues
Printed From: ASRock.com
Category: Technical Support
Forum Name: AMD Motherboards
Forum Description: Question about ASRock AMD motherboards
URL: https://forum.asrock.com/forum_posts.asp?TID=5574
Printed Date: 22 Dec 2024 at 2:01pm Software Version: Web Wiz Forums 12.04 - http://www.webwizforums.com
Topic: Seeking advice - Crash issues
Posted By: nocturnalsandwich
Subject: Seeking advice - Crash issues
Date Posted: 14 Jul 2017 at 12:21am
I am going to link my Reddit post because I already went into great detail about the things I have done to troubleshoot this.
https://www.reddit.com/r/buildapc/comments/6mwbhc/new_build_keeps_crashing_first_time_amd_build_for/" rel="nofollow - https://www.reddit.com/r/buildapc/comments/6mwbhc/new_build_keeps_crashing_first_time_amd_build_for/
My PC Part Picker list is right here.
https://pcpartpicker.com/user/NocturnalSandwich/saved/spzJxr" rel="nofollow - https://pcpartpicker.com/user/NocturnalSandwich/saved/spzJxr Basically the PC crashes when on Windows desktop or while playing games. I've ran through every device and used Windows to update the drivers. I downloaded the "all in one" driver but am afraid that might roll back some of the updates Windows did. I updated to the new AMD chipset, updated to the newest Bios, have tried the RAM in every slot individually and together, reset Windows to a clean install, booted with just keyboard/mouse plugged in, unplugged and remounted all parts and wires, and have replicated the error in all those scenarios as well as with the old GPU (GTX 650Ti). I ran sfc /scannow which found corrupt files and can't repair them. I ran /RestoreHealth and it needs me to manually download the files, but I am not sure which ones are corrupted. When I use the update tool it updates everything but the newest Adobe Flash for Windows 10 (I am running 64 bit Pro) which runs into an error when updating. I've run MemTest86 and it found zero errors through one pass before it told me to close the program. One thing I am responsible for up; I bought RAM not listed on the QVL list for the AB350M Pro4, however I have run it up from 2133mHz on every increment. It runs stable up to 2400 @ 1.2v DRAM with Ryzen Master setting the SoC voltage to 1.15, and I was also able to run stable all the way up to 3200mHz at either 1.35v or 1.4v (voltages I found here on a forum post about RAM issues that were stated as running stable). The crash is persistent through all of those which leads me to believe it is not a RAM issue. I am truly at a loss, any guidance would be appreciated. I spoke to the techs on the phone and they insist it's an issue with the RAM due to me running it too high and tinkering, but the computer ran stable and I was gaming/streaming to Twitch for 3 hour sessions on 3 separate days before this started happening. At one point I got 5 mobo speak beeps but remounting the CPU cleared this error (and I checked for damaged pins, there are none).
|
Replies:
Posted By: wardog
Date Posted: 14 Jul 2017 at 1:19am
" rel="nofollow - I'll take my best stab at helping.
But first. So I don't have to go referencing your prior doings on anothe forum, can I ask the we start fresh with a machine that is reset to UEFI Defaults?
I have my way of looking at and doing things so having UEFI Defaults to beging keeps me and maybe even you from being tripped up in/by this or that "random' change I myself may not be aware of yo having made.
That work for you? I'll go mano a mano against that computer of yours.
If so, write down your current OC settings. Should we fail. Which we wont!
|
Posted By: nocturnalsandwich
Date Posted: 14 Jul 2017 at 1:27am
" rel="nofollow - I reset to factory settings for Windows, and for the ASRock motherboard. I am currently on the 2.5 Bios patch, would you like me to roll back? I am not loading my XMP profile, simply leaving it on auto, and I am doing no overclocking the RAM came at 3200mHz and it is downclocked by the motherboard to 2133mHz and the Ryzen is just running at the standard factory tuning.
I am not home right now, working for ~3 more hours. I will do my best to answer but I won't be able to tinker the hardware and settings further than what I explained above until a bit later. Thanks so much for replying.
|
Posted By: nocturnalsandwich
Date Posted: 14 Jul 2017 at 1:29am
" rel="nofollow - My prior troubleshooting description copied from my other post:
Describe your problem. List any error messages and symptoms. Be descriptive. The PC boots and runs stable on Windows for basic web browsing and daily use. Whenever I boot a game (tested PUBG, CSGO, OW) I am able to play for as long as I want, but the PC crashes when I close the game app to return to the desktop. After this happens the PC even crashes sometimes sitting on Windows just using Discord and Chrome. No specific error message but the order of events for the crash are as follows; First few times it crashed the monitors shut off, the USB devices with LED lights shut off (but the laser on the mouse stayed on?), and the fans/CPU keep spinning. The GPU fans haven't turned on even when it is working so I don't know if the GPU is still running when this crash happens. NOTE: this happened with a 1050Ti and the 650Ti I slotted in while waiting for the new GPU. The next few times it has happened after troubleshooting and tinkering the screen simply freezes, I can still engage and disengage CAPS and NUM lock lights, and as before the fans and CPU are till running. Again, this happens when the game closes, and after rebooting from that it will sometimes happen on the desktop.
List anything you've done in attempt to diagnose or fix the problem. Initially I assume the Ryzen 1600 was too much for the old and overused 650Ti and according to GPU-Z I was running the GPU at nearly 70C so logically I just assumed GPU overheating due to being so old. When the problem persisted with the 1050Ti slotted in I got to digging for answers on Google. My CPU is running at a cool 35-45C idle and 50-60C while in games. The new GPU runs so cool the fans don't even need to turn on. My RAM is DDR4 Corsair Vengeance LPX 3200mHz which is a clock speed ASRock advertises is compatible with their mobo. I didn't realize there was a list of specific RAM models among specific manufacturers that are confirmed to work, and the RAM I bought happens to not be on there. However, I've spoken with ASRock over the phone and discussed the RAM support issue. He said 3200mHz will put it in a boot loop, which it did at 1.2DRAM volts in bios settings. I've since walked it up at 1.2v, 1.35v, and 1.4v and I get the computer to boot and run Windows stable at anything from 2133mHz to 3200mHz when set to 1.35v or higher. However the crash occurs no matter what when I close a game, and will compound into crashes just sitting on Windows randomly if I reboot the PC too quickly after crashing. I've now walked it back down to 1.2v @ 2400mHz since that seems to be the manufacturer and user recommended settings for the time being. I've got my RAM slotted to A2 and B2, recommended by users and ASRock tech support, and basically the guy on the phone told me at this point they wouldn't be able to troubleshoot it any more and they offered to replace the board under warranty. I don't want to do that, so I did more digging and found a user who turned his SoC voltage up to 1.15 in order to get his 3200mHz RAM running stable at ~3000mHz. Ryzen Master already has my SoC clocked to that by default so I am not sure it's the RAM at all at this point.
|
Posted By: wardog
Date Posted: 14 Jul 2017 at 1:51am
Let's begin with a through listing of the components you're using, including the PSU's Make Model and age.
Not that this applies to you but it seemingly never fails. A User doesn't list their PSU and after 4-10 pages of diagnosing it turns out the PSU is 12yrs old and loads more heavily on the +3.3v and +5.0v rails. Well duh
|
Posted By: nocturnalsandwich
Date Posted: 14 Jul 2017 at 2:03am
For detailed model numbers reference this link, it's just a direct listing of my parts.
https://pcpartpicker.com/user/NocturnalSandwich/saved/spzJxr" rel="nofollow - https://pcpartpicker.com/user/NocturnalSandwich/saved/spzJxr
I'll list everything I have.
CPU: Ryzen 5 1600 3.2gHz 6 Cores
GPU: EVGA GTX 1050Ti FTW ACX3.0 Cooling (SSC) 4GB GDDR5
MOBO: ASRock AB350m Pro4
RAM: Corsair Vengeance LPX 3200mHz 2 x 4GB
SSD: Western Digital Blue Series M.2 2280 plugged into the 6gb/s SATA interface slot
PSU: EVGA 80+ Bronze 450w https://pcpartpicker.com/product/MfvZxr/evga-450w-80-bronze-certified-atx-power-supply-100-b1-0450-k1
EDIT: it's all new out of the box.
Anything else?
|
Posted By: nocturnalsandwich
Date Posted: 14 Jul 2017 at 2:55am
Just found the PSU first available date on Amazong as Date First Available | May 4, 2016 |
---|
|
Posted By: MisterJ
Date Posted: 14 Jul 2017 at 3:33am
nocturnalsandwich, I agree with wardog - reset to UEFI defaults (F9 in BIOS). I am concerned about the XMP 3200 MHz DRAM clock. Defaults will allow it to go back to SPD settings, which I suspect are 2133 MHz. Once the crash problem is gone, then wardog and others will help with the memory speed. I would like to hear more about your crashes. Please look in the Windows folder for a Minidump folder. If there is one, please compress it and upload where we can all have a look. Enjoy, John.
------------- Fat1 X399 Pro Gaming, TR 1950X, RAID0 3xSamsung SSD 960 EVO, G.SKILL FlareX F4-3200C14Q-32GFX, Win 10 x64 Pro, Enermx Platimax 850, Enermx Liqtech TR4 CPU Cooler, Radeon RX580, BIOS 2.00, 2xHDDs WD
|
Posted By: nocturnalsandwich
Date Posted: 14 Jul 2017 at 3:51am
I don't want you all to misunderstand me; changing my RAM clocking was simply a way to try to find out if the RAM was the issue. I've already reset to factory settings and the RAM is running at 2133mHz and the crash is still happening. The crash originally happened on factory settings and I only ever changed the RAM settings to troubleshoot the issue.
As I stated above, I am not using the XMP profile, and even when the XMP profile was loaded the system defaulted it to 2133mHz. This isn't an issue with getting the RAM where I want it or overclocking, I am fine with it running stable and just working. I was able to get the RAM booting stable up to it's factory 3200mHz at every increment along the way but the crash still happens. So I've ruled out the RAM being an issue.
I have no issues posting, booting, installing Windows, running Windows. The crash happens when I begin to actually use the computer and game/multitask normal desktop apps. I was just trying to list things I did to check this since my RAM doesn't appear on the QVL list that was a priority to troubleshoot.
|
Posted By: nocturnalsandwich
Date Posted: 14 Jul 2017 at 6:25am
Screen shot of my crash logs.
http://imgur.com/a/pnso3" rel="nofollow - http://imgur.com/a/pnso3
|
Posted By: wardog
Date Posted: 14 Jul 2017 at 7:46am
nocturnalsandwich wrote:
Screen shot of my crash logs.
http://imgur.com/a/pnso3" rel="nofollow - http://imgur.com/a/pnso3 |
Oh hell!
Before proceeding ANY further, get yourself a 550-650w psu. Name branded like Seasonic, Corsair, EVGA etc etc.
|
Posted By: MisterJ
Date Posted: 14 Jul 2017 at 7:57am
nocturnalsandwich, thanks but that is not really useful. I want to see the Minidump files if there are any. Please look in your Windows folder for a minidump folder, compress and upload. Thanks and enjoy, John.
------------- Fat1 X399 Pro Gaming, TR 1950X, RAID0 3xSamsung SSD 960 EVO, G.SKILL FlareX F4-3200C14Q-32GFX, Win 10 x64 Pro, Enermx Platimax 850, Enermx Liqtech TR4 CPU Cooler, Radeon RX580, BIOS 2.00, 2xHDDs WD
|
Posted By: parsec
Date Posted: 14 Jul 2017 at 9:34am
nocturnalsandwich, have you installed the video driver yourself, or was it done via Windows update?
Even if you installed the video driver yourself, Windows 10 may have updated it since then. The only problems with Win 10 I've had is when I allowed it (before I could stop it) to update the video driver.
This is EVGA's driver download page, just select the appropriate card (10 series), the OS version, and uncheck Show Latest Driver to get a full list of driver versions. You can find your driver version in Programs and Features:
https://www.evga.com/support/download/" rel="nofollow - https://www.evga.com/support/download/
Given what you wrote in your first post, that you have Windows install all the drivers, would I trust Windows to install the appropriate drivers? No, that is the last thing I would want to happen. You are reluctant to and didn't run the AMD All in One driver installation package? Running that should have been the FIRST thing you did after installing Windows 10. Later in your post you said you "... updated to the new AMD chipset", does that mean chipset drivers? Was that via Device Manager, as you described? Any time I install Windows, I make sure the network interface is disconnected/disabled, to prevent Windows update from installing anything before I install the correct drivers.
You said in your first post that sfc /scannow returned with non-repairable errors, and all I can see is you reset Windows to default settings. Do you still have sfc returning non-repairable errors after the Windows reset? If so, you have a corrupted Windows installation, which is another variable in the problem. The main problem is the crash when a game is closed, or apparently just web browsing after a crash somehow (restarted the PC?) We still have not seen the reason for the crash, as stated above, which may only contain information on the driver involved during the crash, which is a good clue. That is essential information.
Your video card does not seem to need any extra power cable from the PSU connected to it, is that correct? The card only uses power from the PCIe slot?
------------- http://valid.x86.fr/48rujh" rel="nofollow">
|
Posted By: Jm62383
Date Posted: 14 Jul 2017 at 9:54am
" rel="nofollow - What setting is your power management on? In Nvidia control panel set it to "prefer maximum performance" and see what happens.
|
Posted By: wardog
Date Posted: 14 Jul 2017 at 9:55am
" rel="nofollow - MisterJ, I'm considering this is power related.
His sfc /scannow was unable to fix some corruption. NTFS has been around for years. For all intents it is bulletproof now. And rather adept at correcting itself if not too far over the hill, which again, doesn't happen often enough to not be discrediting this as power related.
IMO. You please stay on the track you're on. My guess? It isn't writing a MiniDump. He didn't mention BSODs, the lack of which are considered hard events severe enough to cause it to instead immediately shut down and or reboot
IMO, since we are troubleshooting over the internet no less, his PSU is where I myself would begin now knowing his 1050ti GPU involved. At best, he's on the very edge with it. At worst, it'll be the reason behind what he described.
|
Posted By: nocturnalsandwich
Date Posted: 15 Jul 2017 at 12:21am
" rel="nofollow - First off, thanks for all the replies. I was not paying too much attention last night and just going at it again.
I'll try to address you all one at a time.
Wardog: The PSU was brought up by ASRock support when I was speaking to them last night. We were on the phone quite a solid while and he went over all the steps he thought to rule out the motherboard. I explained the failed scannow results and speculated on a corrupt installer myself. He offered to replace the board via customer service as the best options to rule it out. In his best judgement he didn't want to say it was RAM or a PSU until we ruled out a faulty board. However, I am at the ready to buy a new PSU or RAM if the new board encounters the same crash. I've tested every wattage estimator on the web with my components and added in other devices accounting for USB devices etc and no single estimator thinks it would ever need more than 290watts on the high end. I am reluctant to think the 450w can't push that, but it is definitely possible and something I've considered. I have used both Windows "High Performance" power plan, the new AMD Chipset Power Plan for "Ryzen Balanced", and set my games to "Max Power" or however it is phrased in NVIDIA Controller. I just don't know how to check the hard numbers coming out of the PSU.
MisterJ: My apologies, an IT career friend told me he dealt with similar crashes in his professional environment and they were due to bad motherboards. He suggested those logs be a place I look to get hints for what type of hardware error could be crashing Windows OR if the error was actually due to bad software/installation. I'll look for what your asking tonight.
parsec: Thanks for chiming in. When I first wrote the post you're referencing I hadn't done a clean install of Windows yet. At that point my worry in installing the AMD all in one driver from the ASRock support page for my motherboard was reverting drivers Windows had already found updates for and causing potentially more issues. I overrode this by just clean installing Windows. I ran the all in one tool and then even updated beyond that with the new AMD chipset driver w/ power plan. ( http://support.amd.com/en-us/download/chipset?os=Windows+10+-+64 ) I also made sure Windows was updated as currently as possible before trying to have it install newer drivers AFTER I ran the all in one and Chipset installers. SFC /scannow still does return corrupted files. I downloaded the Windows automated repair tool that runs scan now and patches by itself but it stalls (doesnt freeze just stops making progress) at 33% under "Checking Windows Store" and I wasnt able to use the DISM /RestoreHealth tool to auto fetch the files. To explain my crash as clearly as I can, it is a "black screen" crash that either consists of the image freezing on both monitors followed by the signal being lost accompanied with the USB devices losing power to LEDs and functions; or a pure screen freeze (that locks until I hard reboot) with the USB devices still receiving power and CAPS/NUM working as intended. During both of these scenarios the CPU and chasis fans all still spin and the computer seems to be running which is what made me think it may be a GPU issue. I'm going to go ahead and follow your link to the EVGA page, I had only used the drivers from the NVIDIA site prior. Will update with fixes, I don't want to assume and be arrogant enough to rule out I've exhausted all driver possibilities but beyond the detail of my initial post I've used Windows and manufacturer websites save EVGA (little blank moment on my part) to double check and download drivers for everything down to my camera, monitors, mouse, keyboard, headset, speakers, etc. My GPU is the EVGA FTW edition of the 1050Ti 4GB card, it is the only card that requires a 6 pin connector in the 1050 line. It runs at 125w MAXIMUM and the Ryzen runs at 65w MAXIMUM and those are the two biggest power draws in the entire PC. I am going to include some images below that may be relevant.
Jm62383: I've adjusted the power settings in Windows to all the highest settings, as well as in the NVIDIA app.
Possibly relevant to everyone: http://imgur.com/a/d1E36
|
Posted By: MisterJ
Date Posted: 15 Jul 2017 at 1:43am
" rel="nofollow - nocturnalsandwich, I will look at any minidumps you have and see if I can make some sense of them. Are the irreparable disk errors on your system disk? If so, I would recommend that you install a fresh copy of W10 after a full format (not Quick). If the format fails, them you may well need a new drive, or at least install W10 on another drive (after full format). Good luck and enjoy, John.
------------- Fat1 X399 Pro Gaming, TR 1950X, RAID0 3xSamsung SSD 960 EVO, G.SKILL FlareX F4-3200C14Q-32GFX, Win 10 x64 Pro, Enermx Platimax 850, Enermx Liqtech TR4 CPU Cooler, Radeon RX580, BIOS 2.00, 2xHDDs WD
|
Posted By: nocturnalsandwich
Date Posted: 15 Jul 2017 at 1:55am
There are no disk errors at all, it passes all scans and checks. I fully reformatted the drive on the install I mentioned above which took place Wednesday. I was able to secure another Windows install USB from a coworker for the same version and bit rate, just in case my install USB contains corrupted files.
Before proceeding with any reordering of parts I'm going to exhaust all the software troubleshooting you guys give me including installing the EVGA drivers and using this new confirmed clean install of Windows that functions with out of the box settings for my coworker.
|
Posted By: MisterJ
Date Posted: 15 Jul 2017 at 2:17am
Great, thanks nocturnalsandwich! I assume by "bit rate" you mean architecture (32 or 64 bit). You should be able to run a scan of your install USB and identify any errors (probably the install would fail if there were any). I don't think you need to do that right now. I have lost track of your setup. Can you outline what drives you have and which one is the system disk, please. Perhaps after your SW pursuit, you could get down to a minimal configuration for testing - using one monitor for example, UEFI defaults, one memory stick, etc. Enjoy, John.
------------- Fat1 X399 Pro Gaming, TR 1950X, RAID0 3xSamsung SSD 960 EVO, G.SKILL FlareX F4-3200C14Q-32GFX, Win 10 x64 Pro, Enermx Platimax 850, Enermx Liqtech TR4 CPU Cooler, Radeon RX580, BIOS 2.00, 2xHDDs WD
|
Posted By: nocturnalsandwich
Date Posted: 15 Jul 2017 at 2:52am
Yes, architecture haha.
Exact setup link is here: https://pcpartpicker.com/user/NocturnalSandwich/saved/spzJxr" rel="nofollow - https://pcpartpicker.com/user/NocturnalSandwich/saved/spzJxr
The links from this page include model numbers and all the like. There is one drive, SSD with Windows on it. Image of the drive and all my components are in this Imgur album ( http://imgur.com/a/d1E36" rel="nofollow - http://imgur.com/a/d1E36 )
As far as minimal configuration goes; I've testing with one monitor this entire time using HDMI, I'e replicated this crash with both RAM sticks individually AND in each slot individually, AND have done that at every mHz and safe voltage in an effort to rule out RAM compatibility issues. I've also replicated the error with a GTX 650Ti AND the new 1050Ti.
|
Posted By: MisterJ
Date Posted: 15 Jul 2017 at 4:45am
nocturnalsandwich, OK, looks like you have done a great job of debugging! Once you settle the video driver question, sounds like it is MB RMA time. Thanks and enjoy, John.
------------- Fat1 X399 Pro Gaming, TR 1950X, RAID0 3xSamsung SSD 960 EVO, G.SKILL FlareX F4-3200C14Q-32GFX, Win 10 x64 Pro, Enermx Platimax 850, Enermx Liqtech TR4 CPU Cooler, Radeon RX580, BIOS 2.00, 2xHDDs WD
|
|