System crashing on Fatal1ty AB350 Gaming-ITX/ac
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=7912
Printed Date: 06 May 2024 at 2:04pm
Software Version: Web Wiz Forums 12.04 - http://www.webwizforums.com
Topic: System crashing on Fatal1ty AB350 Gaming-ITX/ac
Posted By: Sakphul
Subject: System crashing on Fatal1ty AB350 Gaming-ITX/ac
Date Posted: 04 Mar 2018 at 4:19am
|
" rel="nofollow - Hello everyone, I have a very strange error and I believe that it is caused by some wrong bios setting. My System: Processor: AMD Ryzen 7 1700 Motherboard Model: ASRock AB350 Gaming-ITX/ac RAM:2x G.Skill DIMM 8GB DDR4-2400 (F4-2400C15-8GNT) Video Card: ASUS ROG STRIX RX 470 4G GAMING OC PSU: Corsair SF450 CPU Cooler: Noctua NH-L9x65 SE-AM4 Casing: Silverstone Sugo SG05 SSD: SanDisk Extreme Pro 480GB (SDSSDXPS480G) Operating System: Microsoft Windows 10 Professional (x64) Build 16299.248 (1709/RS3) When I try to play a game the system crashes after a couple of seconds. With crashing I mean a hard crash like unplugging power. I have to restart the system on my own (It does not restart on its own). If I try to only put load on the gpu via furmark it is working (>1h). Also loading the cpu only (via Prime95) works like a charm (> 6h). If I do both the system crashes (sometimes it lasts for about 5 minutes). What I tried: Reduce GPU frequency to about 1000MHz By reducing the frequency for the gpu the system is able to run for a couple of hours, but stil randomly crashes. Exchange PSU I tried different PSU's (other 450W models and also a 650W model) but all show the same behaviour. Test RAM with memtes Memtest86 was running for about 6h on all 16 cores and shows no errors (it reached to 8 1/2 passes). Reduce RAM speed to 2133MHz System still crashes. Exchange Motherboard I thought that maybe the motherboard is broken. So I did a RMA and got a new one. System is still crashing with the new board. Disable AMD turbocore System is still crashing. Undervolt/underclock CPU GPU on default. Systems tability gets a little bit better. But still crashes after a couple of minutes Reassemble system I reassembled the system after exchanging the motherboard and also checked the wiring. Nothing wrong as far as I can see. Try clean Windows/Ubuntu I tried on a freshly installed Windows and Ubuntu. System crashes on both OS's after a couple of seconds. I have verified the gpu is working in my old system. It is working like a charm. So I now think that maybe there is some error in the bios setting. I tried disabling Over Temperature Protection but this did not work. I also set all coolers to maximum speed (maybe some temperature spike) but the system still crashes. I now have no idea what could be wrong. My simple guess is that Asrock B350 ITX cannot handle a Ryzen 1700 and RX470. Therefore it's broken beyond repair... Or does anyone has a idea what could be wrong? |
Replies:
Posted By: superneo
Date Posted: 04 Mar 2018 at 4:40am
| " rel="nofollow - are you using onboard wlan? |
Posted By: Sakphul
Date Posted: 04 Mar 2018 at 4:55am
|
Well, yes and no. I am connected via wired ethernet. But I havent disabled wlan inn bios.Also did not connect antennas. |
Posted By: JohnM
Date Posted: 04 Mar 2018 at 7:50am
|
" rel="nofollow - Which BIOS version? Let me see if I understand your problem. You can stress either the CPU or the GPU and it's fine but if you stress both simultaneously you get a crash "like unplugging the power". Also, where you say "System still crashes" that means it crashes if you stress both CPU and GPU but is fine otherwise. Is that correct? It sounds like a power problem. How are you providing the CPU power - 4 or 8 pin connector? (I think I can guess as I have the same PSU but I'd like you to confirm it.) When it shuts down, do the fans stop and lights go out? Have you tried removing the motherboard from the case and powering it up sitting on the cardboard box?
------------- ASRock Fatal1ty AB350 ITX P4.90, AMD Ryzen 5 2400G, 2x8GB Corsair CMK16GX4M2A2666C16, 250GB Samsung 960EVO, 500GB Samsung 850EVO, 4TB WD Blue, Windows 10 Pro 64, Corsair SF450, Cooler Master Elite 110 |
Posted By: nanohead
Date Posted: 04 Mar 2018 at 8:20am
|
" rel="nofollow - Agree with JohnM , Sounds like a power problem to me. Almost like when you try to combine the current draw of both the CPU and the GPU it causes the power supply to go into safety mode. Your power supply is a bit on the small side for all those components, but it shouldn't just cut out, unless its a defective unit. Another dumb question. Are you powering the motherboard with an 8 pin or 4 pin connector from the power supply. If its the 8 pin, are you using the correct 8 pin? the one for the motherboard (2x4 pin) instead of the PCIe 8 pin (6+2 pin). Are you only using 4pins? Do you have the correct PCIe connector in the video card? If you've switched everything as you said in your original post, then its either miswiring of the mobo with the PS, a bad PS, or possibly a bad GPU. It could also be a bad PCIe connector too I suppose |
Posted By: MikeBE
Date Posted: 04 Mar 2018 at 11:56am
|
I agree. The power draw of the RX 4XX cards is supposedly quite high, and I would use a power supply of at least 750 watts. ------------- -Mike Elek http://photographytoday.net |
Posted By: Sakphul
Date Posted: 05 Mar 2018 at 1:55am
|
Bios Version is 3.10 "System still crashes" means that it is crashing while loading CPU and GPU both at the same time. Only loading CPU or GPU alone is fine. I have connected the CPU Power via the 8-Pin connector. When it shuts down all lights go out and fans are not spinning anymore. System has completely shut down. I can try running it outside of the case. But I fear that it wont make much of a difference. Nevertheless I will give it a try. |
Posted By: Sakphul
Date Posted: 05 Mar 2018 at 2:01am
" rel="nofollow -
Since I tried multiple different power supplies (other 450W brands and a 650W) I do not think that its related to bad Powersupply. My Powersupply can run the GPU and a A10-5700 (has almost the same Power draw as the Ryzen 1700) just fine (was my old build). Since I switched the Mainboard it also cannot be a faulty PCIe connector. Only if all Asrock B350 have faulty PCIe connectors. |
nanohead wrote:Posted By: Sakphul
Date Posted: 05 Mar 2018 at 4:37am
|
Just tried running the Mainboard outside of the case on a cardboard. Only CPU, RAM, GPU and SSD connected. Nothing else. It still crashes during benchmark. |
Posted By: JohnM
Date Posted: 05 Mar 2018 at 10:46am
" rel="nofollow -
That suggests to me that it's the PSU shutting down rather than a failure of the VRMs to provide sufficient power to the CPU. Can you measure the power drawn from the supply socket with something like a Kill A Watt meter? If not I'd try a higher powered PSU or a lower powered GPU. ------------- ASRock Fatal1ty AB350 ITX P4.90, AMD Ryzen 5 2400G, 2x8GB Corsair CMK16GX4M2A2666C16, 250GB Samsung 960EVO, 500GB Samsung 850EVO, 4TB WD Blue, Windows 10 Pro 64, Corsair SF450, Cooler Master Elite 110 |
Posted By: Sakphul
Date Posted: 06 Mar 2018 at 4:34am
|
Posted By: JohnM
Date Posted: 06 Mar 2018 at 6:22am
|
" rel="nofollow - I'm out of ideas, then. Sorry. RMA? ------------- ASRock Fatal1ty AB350 ITX P4.90, AMD Ryzen 5 2400G, 2x8GB Corsair CMK16GX4M2A2666C16, 250GB Samsung 960EVO, 500GB Samsung 850EVO, 4TB WD Blue, Windows 10 Pro 64, Corsair SF450, Cooler Master Elite 110 |
Posted By: Sakphul
Date Posted: 06 Mar 2018 at 6:40am
" rel="nofollow -
|
Posted By: JohnM
Date Posted: 06 Mar 2018 at 8:55am
|
In case you hadn't realised, I'm just a customer, like you and everyone else here. I have no affiliation with ASRock, AMD or anyone else. I just use their products and, in general, I'm very happy with them. ------------- ASRock Fatal1ty AB350 ITX P4.90, AMD Ryzen 5 2400G, 2x8GB Corsair CMK16GX4M2A2666C16, 250GB Samsung 960EVO, 500GB Samsung 850EVO, 4TB WD Blue, Windows 10 Pro 64, Corsair SF450, Cooler Master Elite 110 |
Posted By: Sakphul
Date Posted: 07 Mar 2018 at 4:17am
|
Posted By: JohnM
Date Posted: 07 Mar 2018 at 5:11am
That's ok. Advice is offered in good faith in the hope that it will help. Sometimes it does. Sometimes it doesn't. But it's always free. You just need to try to put your frustration to one side because I really don't care whose products you end up with, though I hope you find a combination that works for you. There's a new non-beta BIOS (P4.50) out today. It might help, though it probably won't. ------------- ASRock Fatal1ty AB350 ITX P4.90, AMD Ryzen 5 2400G, 2x8GB Corsair CMK16GX4M2A2666C16, 250GB Samsung 960EVO, 500GB Samsung 850EVO, 4TB WD Blue, Windows 10 Pro 64, Corsair SF450, Cooler Master Elite 110 |
Posted By: fauxpaus
Date Posted: 26 Apr 2018 at 3:01am
|
" rel="nofollow - Hi all. I'm having a similar problem to OP, other than that when I stress test with CPU and GPU simultaneously, I do NOT have a hard shut down. I have hard shut downs only while on WoW. At random times. The long and torrid story is at the following link: https://us.battle.net/forums/en/wow/topic/20762267338?page=1
I've tried everything, replaced all components down to the motherboard. There is a VERY extensive list of things I've tried on the link. I'm a desperate girl now and about to lose my mind. I'm curious if the OP ever resolved the issue? If anyone has any suggestions for me...please...let me know. |
Posted By: Xaltar
Date Posted: 26 Apr 2018 at 3:24am
|
I don't see any mention of your BIOS version, I may have missed it with that DXDiag copy paste in there. I also don't see a power supply make and model listed on your thread, that would be helpful. Some power supplies, even ones rated as 500w+ either lie (state maximum power instead of nominal power) or are what is know as dual rail PSUs (have their power rating split in 2 making them in essence half the rating if only one portion is connected). Generally, a hard power off is indicative of a power issue. Either the PSU is inadequate or there is a physical short somewhere. It is quite common for there to be an extra mounting standoff under the board that should have been removed and is shorting for example. Unfortunately issues like this can be very tricky to pin down. There are so many possible causes that it makes it tough to work out. We are more than happy to help you work through all the steps to getting things working correctly though  First thing to check is that there are no shorts, no screws or misplaced standoffs under the motherboard. It is an annoying job taking things apart to look but if it solves your issue it will be worth it. If it does not, at least we can rule it out and move on. If you can also give us your BIOS version and PSU make and model, that would be great too. ------------- 
|
Posted By: fauxpaus
Date Posted: 26 Apr 2018 at 3:46am
Apologies. That's a super long post. The info is all in there but buried under an avalanche. Regarding the physical short: I've now taken the entire rig apart and put it back together multiple times (twice yesterday). There are definitely no screws rattling around. Currently the only things under the board are the 4 standoffs on the corners. Regarding my PSU: My PSU originally was an older 750W that I thought was failing, swapped to a 550W that wasn't rated (before I knew better), then most recently I just installed within the last 2 weeks a Corsair CX750 cp-9020015-na (https://www.amazon.com/Corsair-Bronze-Certified-Non-Modular-CP-9020015-NA/dp/B008RJZQSW) which is a dedicated single +12V rail. There has been no change in the frequency of my shutdowns between any of those. I'm on BIOS v 4.6 (posted 4/24/18). |
Posted By: Xaltar
Date Posted: 26 Apr 2018 at 4:50am
|
" rel="nofollow - Thanks for the info, that PSU should be fine :) Are you using the wifi on your board to connect to the internet/network? You can try downgrading your BIOS to version 3.40, that was the last BIOS that was specifically tuned for Ryzen 1xxx CPUs. The newer ones are to support Ryzen 2xxx CPUs and seem to cause issues for some people using Ryzen 1xxx CPUs. It appears to mostly effect boards with wireless integrated so you definitely fit the bill. I am sure there will be a new BIOS update soon that will be better tuned for both generations but for now, stay away from 4.xx. Make sure you download the instant flash version from your product page, extract it onto a FAT32 formatted flash drive then use the instant flash option in your UEFI. This is the safest flashing method. BIOS 3.40 is the one you want. Let us know how that goes for you. -------------
|
Posted By: fauxpaus
Date Posted: 26 Apr 2018 at 5:00am
Thanks. I'll give that a try... I'm plugged in via ethernet and have all the wireless components disabled, actually. Don't know if that makes a difference.
|
Posted By: Xaltar
Date Posted: 26 Apr 2018 at 6:08am
|
It doesn't, it has to do with the wireless adapter itself, enabled or disabled it still seemed to cause issues with Raven Ridge CPUs (R3 2200g and R5 2400g), the fix then caused issues with older CPUs..... This seems to be a fairly common issue with wifi enabled boards. Good luck with the back flash. I just got done doing my dailies on WoW myself, I know where you are coming from, having your PC let both you and your guildies down sucks  -------------
|
Posted By: cristy6100
Date Posted: 26 Apr 2018 at 6:38am
|
Well here's the thing, today I tested my 2400G with X370 ITX/ac (rest of the config in my sig) with a very old, like over 14 years Pentium 4 power supply, a chinese no name 350W PSU. It will not crash even with the iGPU and CPU stressed with IBT and for the GPU furmark, HWInfo says about 130-140W CPU+SoC power draw, the CPU has only 4pin CPU power, I then switched on the Wi-Fi (it seems many people have problem with Wi-Fi again no problems for the last 2 hours or so, but Wi-FI was working for me from the get go. I went from UEFI 2.10 to 3.xx then 4.xx when I bought the 2400G to replace the A10 9700 I had previously, Wi-Fi worked from the moment I bought the board to this day with no problems or lock-ups. But the throttling is there when both CPU and GPU stressed but no crashes. (these will be fixed with new AGESA code) If you already RMA'ed the board and got a new one, then the motherboard is not the problem, I ruled out the PSU for you with my testing, the CPU cannot be defective (chances are very slim). So what remains is the Dedicated GPU/RAM and the SSD and hard drives you are using. May I suggest unplugging all secondary drives from the system and leaving just one (if possible use a clean hard drive for this testing and not an SSD), make Windows 10 x64 USB with Windows Media Creation Tool from Microsoft, be sure after install to patch and update Windows, its critical you update Windows with all the available updates, Raven Ridge/Pinnacle Ridge need the latest Windows updates to function correctly. When Windows April Update will be released it will include all updates for Raven and Pinnacle Ridge but until then make sure to update. Also note AM4 does not support any OS older than Windows 8, and when using Raven Ridge/Pinnacle Ridge only Windows 10 x64 is officially supported, if you use Linux/Windows 8 or 8.1 you might have problems, also note that Windows 7 is not supported at all. If you still have freezes or hard-locks, please remove video card and leave only the iGPU and test again and report back here. If after doing all I said above you still have lock-ups/freezes the problem is your memory. My Corsair LPX works without any addition config, it boots straight away @ its default 2400 and I can overclock it up to 3333MHz. With the same memory my Bristol Ridge A10 would not boot past 2133MHz, but Ryzen seems to like Hynix memory Cheers
------------- Main Rig: AMD Ryzen 2400G | ASRock Fatality X370 ITX/ac UEFI 4.90 | Corsair LPX DDR4 2400@3200 1.35V | Corsair RM650i PSU | CM ML120L AiO Cooler |
Posted By: fauxpaus
Date Posted: 26 Apr 2018 at 7:07am
" rel="nofollow -
Hi, I'm actually a bit confused about your reply. I built the entire thing from scratch in Dec. Since then I have replaced - power supply 2x (Jan and 2 weeks ago) - CPU (last week) - GPU (2 weeks ago) - motherboard (yesterday) - HD: The story is as follows - The original was not an SSD - just a standard- and until I got the SSD I had no secondary drives. So, the original had Win 10 installed from the beginning. My out of the box original install was Win 10, kept updated, etc. Over the course of the past months I performed multiple wipes of the HD, following the directions as you stated (I did make a win 10 x 64 usb via windows media creation tool from microsoft, installed all patches and updates available). There is 0 chance that Win 10 was not updated/is not updated, and I'm not using any of the other versions of Windows stated in your post. It was hypothesized that, although unlikely, based on ONLY crashing while playing WoW and not any other games/stress testing/minding my own business, the problem could be "the hard drive", and some magical read/write issue, or a faulty section of the HD. Based on this, I was eventually convinced to get an SSD (1 week ago). I performed an install of Win 10 on that SSD via the same method (Win 10 x 64 USB with Windows Media Creation Tool, patched and updated). At that time, I did NOT have a secondary drive plugged in at all. I had the same hard shut downs anyway. I actually just plugged in the old HD yesterday. Regarding the RAM: I've tested the ram pretty extensively, with no problems. I've run the super long version of the windows memory diagnostic (the multi hour option). I've tried using each stick independently, both in each slot for a total of 4 combinations. You said GPU/RAM/SSD/HD, but I believe that I've ruled those out given the steps I've taken?
|
Posted By: Xaltar
Date Posted: 26 Apr 2018 at 3:05pm
|
@Cristy: She is using an R5 1600 not a Raven Ridge CPU so there is no iGPU to test with ;) @Fauxpaus: Cristy was under the impression that you were using an R5 2400g/R3 2200g, probably because I mentioned them. The BIOS backflash to 3.40 is where I think you will find success. Numerous people have reported issues with 4.xx when using Ryzen first gen CPUs, interestingly, much like yourself, only with particular games. Battlefield 1 is a common one. The one thing I noticed was that they all appear to happen during online play. I have been digging into your issue -------------
|
Posted By: cristy6100
Date Posted: 26 Apr 2018 at 3:28pm
Oh, yes I was under the impression you were using a Ryzen 2200G, and I read the three pages :), somewhere along the way I got confused. Well with a Ryzen 1600 you should have no issues whatsover. Well then we can rule out the MB/PSU/HDD-SSD/CPU you already said you checked for shorts, so the only thing remaining is the RAM, to you have the posibility to test with a Corsair LPX kit preferably under 2666MHz?; or any DDR4 kit under 2666MHz (high speed memory kits come with different dies than low speed ones) ------------- Main Rig: AMD Ryzen 2400G | ASRock Fatality X370 ITX/ac UEFI 4.90 | Corsair LPX DDR4 2400@3200 1.35V | Corsair RM650i PSU | CM ML120L AiO Cooler |
Posted By: stree
Date Posted: 26 Apr 2018 at 6:23pm
|
Been following this, sadly with nothing to suggest. But I did notice this in a recent post from Fauxpas, which at the least needs clarification because it is a bit ambiguous.not intentionally I am sure! The bit that caught my eye was " So, the original had Win 10 installed from the beginning. My out of the box original install was Win 10, kept updated, etc. " Now this could refer to a recent 1709 ISO from MS, in which case false alarm, but it could refer to an older version such as a boxed CD version 1703 but with all updates to present. If the latter then I suggest that therein lay at least some of the problems. The only other thing of course is the usual , make sure it has the latest chipset and radeon drivers, installed in the right order. One more item: I have found that the various powerplans can have a effect, I now choose performance and it seems to cure many annoying issues. Then again, I am on a different build, so may not apply universally. ------------- ASRock X370-ITX BIOS 4.50 R5 2600 Cryorig C7 EVGA GTX 950 75w 2x8GB Ballistix Sport LT 2933 960Evo M.2 256GB, Firecuda 1TB Win 10 Pro 64 1803 G-Unique Archdaemon 300 Watt Lian-li Q21B |
Posted By: fauxpaus
Date Posted: 27 Apr 2018 at 5:01am
Sorry for any confusion. I will clarify (hopefully). On my prior PC (super old) I had Windows 10. I made a Win 10 USB install via Win Media Creation Tool. Cut to scene two, early Dec after Black Friday sales, all new parts arrive for all new PC! Included in this pile of parts is a 100% blank, new, 1TB WD HD. Upon putting all my new components together for my all new PC, I proceeded to install windows on that shiny, new, completely blank HD. I used the USB, then did all updates etc. That HD was subsequently wiped multiple times, and Windows 10 was again reinstalled via USB (from a new USB actually, but still using same procedure w/Windows Media Creation Tool) and updated appropriately. Shut downs continue. Then came the SSD - new from the store - plugged in by itself, Win 10 installed from yet ANOTHER new USB (using the same Windows Media Creation Tool). I have never had a boxed CD version of Windows. Regarding the power: When I first put together everything on the new computer, I was using the default power plan. Upon my first few shut downs, I changed to the "performance" power plan. After more shut downs, I downloaded the AMD chipset drivers for the "AMD balanced power plan", which is what I have it set to currently. There have been no changes in the frequency of my shut downs with any of those power plans. |
Posted By: cristy6100
Date Posted: 27 Apr 2018 at 4:41pm
|
" rel="nofollow - Did I mention that unstable AC current to the house can cause shutdowns and hard-locks? The PSU will reset itself sometimes when the PC is on if to low/high voltage is coming on the AC line and that will cause a hard lock of the PC, other times it will just shutdown without no warning. Can you maybe test the voltage of your AC line? Have there been problems with AC delivery in our area? AC is usually delivered in US at constant 50Hz (or very close to 50Hz) in Europe 50Hz and 60Hz but not all PSU are rated to go at 50Hz some are just 60Hz, and plugging them in a 50Hz outlet will cause them to heat up much more, fail or behave oddly. Again this is just an ideea but I have seen things like this happen, I have seen my AC go to sub 40Hz an cause my PSU to heat like sh*t and behave very odd. I still think its your RAM memory but its worth checking if you have stable power in your house Also DO NOT tamper/check AC lines by yourself, I can tell you by experience that you wont like a "sting" from 230V or even 110V for that matter, if you have a friend who is electrician you can request his help, for messuring the AC frequency you will need a advanced multimeter or an osciloscope anyway. First check with other memory modules, if nothing else works, move your PC in a grounded outlet in another part of the house. Also I can suggest removing the LAN cable when doing this testing. Lan cables when connected to faulty/ground looped equipment can cause the PC to lock also because of the motherboard short/ESD protection kicking in. Cheers
------------- Main Rig: AMD Ryzen 2400G | ASRock Fatality X370 ITX/ac UEFI 4.90 | Corsair LPX DDR4 2400@3200 1.35V | Corsair RM650i PSU | CM ML120L AiO Cooler |
Posted By: fauxpaus
Date Posted: 28 Apr 2018 at 5:57am
So - I don't think there is an unstable current to the house. I can play other games and do other things without any issue (and other parties in my household use computers with fairly high power usage and have run them from the outlets I've used without any problems). In the last hour I've been logged into other games and played them successfully. Right now I have another game on my second monitor. Then I log into WoW and hard shut down, instantly. I can't believe that my AC current is suddenly failing only during the few seconds that I decide to log into WoW. Right now I can't log in AT ALL, I'm not even getting 30 minutes of play time before a crash. Since Tuesday, I log in and shut down in <10 seconds. This morning, I got new ram that was on the QVL for the motherboard - put it in - shut down in <10 seconds. After that I took everything out of the case and put it into another case. I used different cables. I logged in and shut down in < 10 seconds. |
Posted By: cyphacipher
Date Posted: 20 Jun 2019 at 7:15am
| Sorry to necro this, but has this ever been solved? I"m on my 2nd RMA (3rd board) for this exact same issue and noone knows what causes it. |
Posted By: fauxpaus
Date Posted: 20 Jun 2019 at 8:21am
| FWIW I never resolved the problem...I went through 2 boards and 2 chips and every iteration of new parts...eventually I ended up selling the system and getting an intel chip. |