Ryzen 1500x AB350m odd crashes
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=5653
Printed Date: 24 Dec 2024 at 1:48am
Software Version: Web Wiz Forums 12.04 - http://www.webwizforums.com
Topic: Ryzen 1500x AB350m odd crashes
Posted By: Robmeng
Subject: Ryzen 1500x AB350m odd crashes
Date Posted: 21 Jul 2017 at 3:17am
|
" rel="nofollow - Hi guys New build for my partner, ok if i link to tomshardware? i have a thread there but it's not received many replies at all and i'm at a loss as to what i should do next http://www.tomshardware.co.uk/answers/id-3471189/build-freezing-games-occasionally-windows-gpu.html there are some screeshots linked (imgur) All the steps i've taken are documented, I would appreciate any help what so ever at this stage. |
Replies:
Posted By: MisterJ
Date Posted: 21 Jul 2017 at 3:29am
|
Robmeng, I suggest you post on the MSI forum. 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: Robmeng
Date Posted: 21 Jul 2017 at 3:39am
| The only MSI product involved is the GPU which is tested and working fine on another system. My mobo is Asrock. |
Posted By: MisterJ
Date Posted: 21 Jul 2017 at 4:15am
|
" rel="nofollow - Sorry, Robmeng, my bad! I suggest you post all your system specifications here as in my signature. I cannot find your memory in the Qualified list but I see no 8G sticks. So, I assume, you are running dual channel. What slots are you using? Are you running at 2400 MHz or OCing? Do you ever get a BSOD? If so, please compress the Minidump folder in the Windows folder and upload somewhere we can access it. I assume the Hynix SL308 is an SSD. Have you loaded the latest drivers? Is it installed in the M2_1 socket? I will look some more on the Tom's site. Explicitly what BIOS are you running (3.00?). 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: Robmeng
Date Posted: 21 Jul 2017 at 4:48am
|
" rel="nofollow - Asrock AB350m pro4 latest bios MSI GTX 970 gaming (factory OC) Ryzen 1500x (stock) 2x4GB Corsair Vengeance LPX DDR4 2400 CMK8GX4M2A2400C14 2x 1TB WD green drives (storage) and a 120GB Hynix SL308 canvass (OS drive) (the two games mentioned below are installed to this drive also) PSU 450W Seasonic G Series 80 plus gold Hybrid modular Windows 10 x64 Updated Main issue is when playing games (Team Fortress 2/ L4D2 mostly) the game will lock up, mainly when loading into a server, or just after loading into the server. But not always, she can play for hours sometimes without issue. We get a variety of errors in event viewer, the latest being - The description for Event ID 13 from source nvlddmkm cannot be found. Either the component that raises this event is not installed on your local computer or the installation is corrupted. You can install or repair the component on the local computer. If the event originated on another computer, the display information had to be saved with the event. The following information was included with the event: \Device\Video3 Graphics Exception: ESR 0x404490=0x80000001 the message resource is present but the message is not found in the string/message table Now as for what i've tried: Fresh clean install of windows. Reinstall GPU drivers using display driver uninstaller beforehand. move ram from A2 B2 to A1 B1. I switched my Evga GTX 980 SC 2.0 into her machine and the freezes persist, using the GTX 970 in my machine i have no problems, so i can rule out the card itself being defective. It's passed through Furmark 1080p test ok multiple times. I've ran memtest for 3 passes with no errors. Used both the XMP profile and mobo default ram speeds. I'm using AMD's balanced power plan but with the pci-e setting off. I've tried debug mode in Nvidia control panel and prefer maximum performance. Temps are fine according to all monitoring software. I was thinking the 450W PSU could be an issue but i believe it is a good PSU and the components shouldn't draw near the full capacity? |
Posted By: Robmeng
Date Posted: 21 Jul 2017 at 4:49am
|
Stress tested PSU in occt for 15 minutes. Done the gpu and CPU tests too. Crashes are not 100% repeatable sometimes she can play for hours other times it will crash again and again. Fresh windows 10 install from usb. Issues since the start. Mobo updated, gpu drivers cleanly installed multiple times, different gpu used. Mainly manifests in Team Fortress 2 (game freezes completely but can task manager to windows easily) event viewer will sometimes log a gpu driver fail sometimes it won't. Occasionlly when viewing videos on youtube the screen will flash green for a split second. At this point i'm thinking power supply - CPU - MOBO or RAM. Hopefully it's a software issue though. Attempted to play TF2 about an hour ago, i loaded in the XMP ram profile in UEFI beforehand, game played for a while then BSOD (first time it's happened on TF2) Currently Reign of Kings without issue (has been running fine for days now tbh) Some screens from HWINFO while playing RoK (can play for hours on this game) http://imgur.com/a/v5xBb" rel="nofollow - http://imgur.com/a/v5xBb and bios voltages http://imgur.com/a/cLkq7" rel="nofollow - http://imgur.com/a/cLkq7 There has been a variety of windows error logs and sometimes none at all. Currently running another series of memtest so i will upload the minidmp asap.
|
Posted By: MisterJ
Date Posted: 21 Jul 2017 at 5:15am
|
" rel="nofollow - Thanks, Robmeng. I do not see anything strange except the VRM temperature but I have seen others complaining about very high temps. Is it really hot to the touch - careful, almost boiling water! My VRM is running 42 C right now. Windows is bad about logging lots of errors that don't seem to mean much. Your BIOS is 3.00? Some memory help in AGESA 1.0.0.6a (3.00 BIOS). How many sticks of memory are you running? Will the games run with half the memory? I am thinking of testing sticks and slots. Use F12 in the BIOS to create a .bmp screenshot file on a FAT formatted USB stick. Are all failures in games? If you let the system sit a the Welcome screen all night, is OK the next day? My system hung on this if memory was in A1 or A2 - RMAed the MB. 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: Robmeng
Date Posted: 21 Jul 2017 at 5:49am
|
" rel="nofollow - Hi I've updated to the very latest 3.00, have tried A1 B1 and A2 B2 Ram is 2x4GB yes https://www.scan.co.uk/products/8gb-(2x4gb)-corsair-ddr4-vengeance-lpx-black-pc4-19200-(2400)-non-ecc-unbuffered-cas-14-16-16-31-xmp Currently running memtest - http://imgur.com/a/tQsAO I've noted that the timings are different from the rated timings of the ram sticks and by extension the XMP profile. i assume XMP profiles are bugged as of now? However CPU-Z gives me this http://imgur.com/a/KZPl5 Benchmark in memtest http://imgur.com/a/fNTAm Memtest result (should i run over night?) http://imgur.com/a/FAoXZ the VRM feels cool to the touch, temps seem to read above 90 all the time. Just to confirm i have installed the cooler the correct way around? http://imgur.com/a/lPw1m I have uploaded the only DMP file in the minidmp folder, dated the 17th (this was a critical CPU error i believe) i'm not sure off its relation to our issues. https://www.dropbox.com/s/rk5u40e1zmbeer8/071717-5750-01.dmp?dl=0 I will try running with a single stick. thanks. EDIT: https://www.dropbox.com/sh/vbepei36w13s8dh/AADlvQ2QaTYXAiT-wl0Ay0Tsa?dl=0 Bios screens (hopefully dropbox will work correctly)
|
Posted By: MisterJ
Date Posted: 21 Jul 2017 at 9:03am
|
Robmeng, the bug check code is 0x00124 which MS says is almost always HW. Perhaps the power supply needs consideration. Is it possible to borrow another one-maybe a little more power? Perhaps the seller will RMA yours. Here is a https://answers.microsoft.com/en-us/windows/wiki/windows_10-update/bugcheck-code-124/98c998d2-447a-40ce-ae1f-8211e355f14d" rel="nofollow - MS link on the error. Please DL the free version of http://www.softnology.biz/" rel="nofollow - Thaiphoon to display the SPD and XMP data on your memory - takes some time. As far as I know the XMP data is OK. I suspect that the Advanced Boot Training (trains memory) is changing memory timings. You can try disabling it. Hard to say on your boxed cooler. My R7 did not have one and I used a water cooler. I think if it fits and your CPU temps are OK, you're good. Looks like maybe more room with AMD logo at the top (90 degrees clockwise). I would not fool with it unless there is a problem. 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: 21 Jul 2017 at 10:45am
|
Given your HWiNFO screenshots, the VRM temperature at 91C is not good, although I don't know how long HWiNFO was running at the time of the screenshot. I see the CPU_OPT and Chassis3 fans are at 0 RPM at times, I would suggest not configuring them to turn off at all, although I have no idea what there location is. Your CPU temperatures look fine, although I see you have no CPU power saving optons enabled, so you are constantly using a fair amount of power for the CPU. Your 450W PSU does not seem to be have any major drops on its 12V and other rails, but again I don't know how long HWiNFO was running before the screenshot. Your PSU's power may be at the edge of just sufficient, meaning little to no headroom. That is mainly for the video card. I see one of the Performance Limit Reasons for the GPU switched to Yes, but I cannot tell which parameter that is, the title for that line is cut off. That means whatever that parameter is, it went beyond a standard limit of some kind. ------------- http://valid.x86.fr/48rujh" rel="nofollow"> 
|
Posted By: Robmeng
Date Posted: 21 Jul 2017 at 8:03pm
|
" rel="nofollow - Hi, thanks for your reply HWiNFO was running for around an hour whilst playing Reign of Kings (no crashes) I believe the VRM temp to be false as i touched said area on the mobo and felt little heat. Was also getting the same reading on startup when the area was cool. I am using ryzens balanced power plan however i disabled cool and quiet in the bios for testing with the game freezes. I can't rule out the PSU, however it's handle multiple stress tests just fine, the error code seen in the above posts was a one time event seemingly unrelated to the game issues. Where do i stand i.e. returning PSU and paying the difference for the next size up? I'll look into the GPU performance, however i am sat at my own PC right now (intel build) and i have - performance limit utilization - set yes and my own system is perfectly stable. The card i have also tested in my PC so i can rule out the GPU at least. EDIT: AIDA64 trial. http://imgur.com/a/mL6Dd Lowest 12v reading was around 12.1v I have one intake fan, on exhaust fan, the CPU cooler fan and the GPU fans, all run so i assume that is a false reading?
|
Posted By: Robmeng
Date Posted: 21 Jul 2017 at 8:05pm
|
" rel="nofollow - Hi MisterJ I cannot for sure rule out PSU, however that error was a one time event that occured outside of gaming, while not ideal it could have just been a hiccup? i will look into exchanging the PSU for the next one up. I will look into the Thaiphoon program, thank you. EDIT: Updated dropbox with thaiphoon results https://www.dropbox.com/sh/vbepei36w13s8dh/AADlvQ2QaTYXAiT-wl0Ay0Tsa?dl=0 Running Heaven now, will bench and post result after at least one hour
|
Posted By: MisterJ
Date Posted: 22 Jul 2017 at 4:30am
|
" rel="nofollow - Robmeng, it is hard to say if you can ignore the BSOD. Something bad happened and others have not reported a 0x124 and I have seen none. Usually one leads to another, so please send any more dumps you get. Your memory is Hynix (mine too). I have had no problems but Ryzen seems to like Samsung memory best (especially B die). I would like to hear results of testing with one stick. Please try A1, then A2, etc. When all four slots are tested then try the other stick. I suggest you use the game that has been most likely to yield a hang. Hopefully you will see a pattern such as one stick failing or one slot failing. In my case both A1 and A2 failed requiring an RMA of the board. How did the temperatures look in AIDA64? I am not familiar with Heaven - use OCCT, Prime95 and W10 memory test (mdsched.exe). 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: Robmeng
Date Posted: 22 Jul 2017 at 9:11pm
|
" rel="nofollow - Hi there, I'm near positive it's a faulty ram module. Three hours worth of memtest found nothing, neither did windows memory test. However i tested for two hours - i would use one stick - launch game, play for a while - exit game -repeat. I would then shutdown the pc - restart - repeat previous steps. I used only slot A2. I did this for both sticks, one froze 9/10 times, the other didn't freeze once. My partner has played TF2 for two hours+ this morning without issue. I assume the game uses memory in such way as to expose the fault? Time to RMA? Rob
|
Posted By: Xaltar
Date Posted: 22 Jul 2017 at 9:35pm
|
That is some intensive troubleshooting Robmeng, great job finding the guilty culprit. I would return that RAM yes. ------------- 
|
Posted By: Robmeng
Date Posted: 22 Jul 2017 at 9:47pm
| It was a difficult one, the issue only manifested in certain situations! |
Posted By: MisterJ
Date Posted: 22 Jul 2017 at 11:13pm
|
I agree with Xaltar, Robmeng, great sleuthing! You will need to ask your vendor for a RMA and they will give you the Authorization number and instructions. You will need to return both sticks. You can ask for an advanced RMA where you give them a credit card number and they send the new memory. You then return the old memory and they charge your card if you don't. Some vendors do this and some don't. Good work. 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 |