Hyper Transport Sync Flood Error, possibly caused by PSU?

WorldSyncError

Commendable
May 31, 2016
11
0
1,510
So I've been getting this error for... I think about 2years now. I just kind of ignored it but recently it's been annoying me a lot more because it will crash my system multiple times every day, especially when I'm running GTA5 or any heavy CPU & GPU intensive games.

quick PC specs:
OS: Windows 7 Home Edition 64bit
motherboard: MS-7713
CPU: AMD Athlon II X4 640 3.00Ghz, 4cores
GPU: MSI R7 260X 2GB
RAM: crucial 2x4GB DDR3 1600MHz
HDD: 1.5TB Seagate ST31500541AS
BIOS: American Megatrends Inc. A7713MLN.106

This is a prebuilt MedionPC with a mix of hardware upgrades I did over 5-6years.

I ran BlueScreenView and it highlighted hal.dll which from what I've researched, indicates that it's a hardware problem.

Filename: hal.dll
Address In Stack: hal.dll+f768
From Address: fffff800`0320a000
To Address: fffff800`03253000
Size: 0x00049000
Time Stamp: 0x4ce7c669
Time String: 20.11.2010 15:00:25

Besides that I also checked the HDD for bad sectors and it came back clean. Reset my BIOS to default. Never OC'd anything.

I actually did quite a bit of research and most of it said to adjust my voltages in BIOS but my motherboards BIOS is extremely limited and doesn't let me adjust much besides device boot priority and some other minor things.

The interesting thing however is that even with my 450w PSU, for some reason I cannot power my system with both, front and back fans. As long as 1 of the fans isn't powered, I can start up my PC no problem. If I try connecting them both, 1 out of 3 fans wont spin; front, back or GPU. I bought this PSU from a local PC hardware store for like $50 some time around...hell I think it might be just 2years ago after the previous one died on me.

Is there anything I can do besides buy a new PSU & a new motherboard? The motherboard may not be at fault but I don't like being so incredibly limited in BIOS. From what I gathered, a 450w PSU should be more than enough for my system. I'm surprised it's causing me problems. Should I buy a more powerful PSU?
 
Solution
generally, for newer machines this error will not be caused by a software driver with the exception of the BIOS settings/overclocking
and any loaded software overclocking drivers (or overclocked GPU).

in your case the CPU was reset and was allowed to restart before the power was stable. on the restart the CPU called a bugcheck.
Error Type : BUS error
BUSLG_GENERIC_ERR_*_TIMEOUT_ERR (Proc 0 Bank 4)
System Uptime: 0 days 0:00:09.268


I can not read the BIOS info so I can not see if the system was overclocked, and it only has a partial list of drivers loaded.

for this problem you should focus on what would likely cause your systems CPU to be reset.
IE it overheated or a GPU used too much power from the pci/e slot.
(I assume you...

WorldSyncError

Commendable
May 31, 2016
11
0
1,510


I did a clean install last year. It didn't affect it at all.
Here's the PSU. Also its 420w, not 450w, but I'm sure it should still be enough.
cSFhv8J.jpg
 
That is a bad power supply to use, replace it. Pretty much anything from the good brands is OK to get, XFX, Seasonic, EVGA, Antec, even some lesser quality units like from Rosewill and the cheaper Corsair stuff will be better.

The power supply you have is not a 420 watt one, it's under 300 if you do the math for the voltage and amps it provides. They have a "600" watt one that is actually about 400 watts. Not a good brand at all.
 

WorldSyncError

Commendable
May 31, 2016
11
0
1,510


Here's the dump file: 053116-44959-01.dmp

I'm probably going to just buy a new PSU anyway and see if it does anything. Going to look up and see what brands are popular and I might as well overkill it a bit anyway since I plan to upgrade in the future too.

Found this PSU: http://www.corsair.com/en/cs-series-modular-cs650m-650-watt-80-plus-gold-certified-psu
Reviews aren't too good, but this is the best affordable PSU I could find outside of US. It's still going to cost roughly 130USD(converted). LCpower is like half price but apparently it's a bad brand here in EU. Kinda aiming towards gold rated and modular so I can do some cable management. My cables are everywhere right now and they're too short. I don't really know what brands are good in EU. I'm thinking of just ignoring the bad reviews and try this one out. It's got 3year warranty from a local website.
 
cpu called the bugcheck 0x124 WHEA_UNCORRECTABLE_ERROR


Error Type : BUS error
Operation : Generic
Severity : Fatal
Error : BUSLG_GENERIC_ERR_*_TIMEOUT_ERR (Proc 0 Bank 4)
System Uptime: 0 days 0:00:08.782

looks like the CPU was reset because of a power problem.
then the power was not stable when the CPU was restarted so the CPU got errors on its internal memory bus, the memory controller detected the error and the CPU shut the system down with a bugcheck.

Generally, this would be caused by a GPU pulling too much power from the motherboard slot, or a power supply that does not provide proper power to the GPU. You want to remove any overclocking in BIOS and in software, I would reset the BIOS to defaults or update the BIOS.
I would also test the power supply and check its ratings

The bugcheck is a secondary problem, the Main problem is the power to the CPU went too low and triggered the cpu reset logic. You have to fix this problem and the bugcheck problem will not occur.




 

WorldSyncError

Commendable
May 31, 2016
11
0
1,510


Thanks for the analysis. Resetting BIOS to default made no difference. I'm betting it's the PSU considering I already had problems powering up the whole system before. I'll update when I get my hands on a new one. I'm kind of low on cash right now so I may have to wait a month or so before I can actually replace it. Worst case scenario, I'll have a spare PSU.
 

WorldSyncError

Commendable
May 31, 2016
11
0
1,510
So I bought a new PSU; EVGA Supernova G2 850w. I've been using it for almost 3days and just got another BSOD with the same error. Although the crashes are much less frequent now, they do still occur.

Here's the latest dump file: https://1drv.ms/u/s!AuZV3ExXkryggReYQWv1p8-0WF2M

I'm going to wipe my HDD clean next week and reinstall Windows 7, so I'll see if that does anything but if it doesn't, I really don't know what else I can do besides buy a new motherboard. Admittedly this one I have now (MS 7713) is very old and looks cheaply made. As far as buying a new motherboard goes, I will be waiting for Zen before I do any upgrades to my mobo+CPU.

Edit: Could the front audio have something to do with it? I also bought a new case and I hadn't plugged in any of the front panel cables right away except for the power switch. I plugged in the front audio to the motherboard this morning and ever since, it seems as if it's crashing just as often as it did with the old PSU.
 
That is a good PSU, but you over-did the power by about 300 watts unless you are planning on getting some higher end parts later to use with it.

If you have not done a clean Windows setup, that should be one of the first things to try. Also try running on one RAM stick at a time.
 
generally, for newer machines this error will not be caused by a software driver with the exception of the BIOS settings/overclocking
and any loaded software overclocking drivers (or overclocked GPU).

in your case the CPU was reset and was allowed to restart before the power was stable. on the restart the CPU called a bugcheck.
Error Type : BUS error
BUSLG_GENERIC_ERR_*_TIMEOUT_ERR (Proc 0 Bank 4)
System Uptime: 0 days 0:00:09.268


I can not read the BIOS info so I can not see if the system was overclocked, and it only has a partial list of drivers loaded.

for this problem you should focus on what would likely cause your systems CPU to be reset.
IE it overheated or a GPU used too much power from the pci/e slot.
(I assume you already updated the BIOS)

note: RAM should not cause this error, storage should not cause this error, there are no normal windows files that would cause this error.

most likely cause will be related to the graphics card and its power. Make sure it is not overclocked, overheating and that it is getting good power.
IE the supplemental power is connected with a good connector that is actually providing power and is not run thru a bad connector.
(or a dead lead, which is not likely if you replaced the power supply)

and be sure to update the motherboard bios, the motherboard triggers the logic that resets your CPU when power is not correct (or it thinks the power is not correct based on BIOS settings) guess, you should reset the BIOS to defaults also.

if you have time,
change your memory dump type to kernel :https://support.symantec.com/en_US/article.howto31321.html
google how to force a memory dump with a keyboard, make the registry settings

reboot wait 10 or 15 minutes and force a memory dump, then put the file c:\windows\memory.dmp on a server and I can take a quick look to see if I see any strange settings or errors.







 
Solution

WorldSyncError

Commendable
May 31, 2016
11
0
1,510
The GPU is an MSi R7 260X. Supposedly it uses 150w. 75w from the 6pin power connector and the rest from the PCIe slot.

No overclocking available for this motherboard.

I never updated the BIOS. I can't even get a proper manual for the board, much less a BIOS update. MSi manufactured it. They don't offer support for OEM products. The OEM (Medion) did send me a manual but that's it. The manual lacks some vital info and it doesn't seem like I'm even supposed to have it as it has "confidential" written all over it along with what I assume is commentary from the engineers that were working on it. Even Medion didn't seem to like working on the board based on some of the commentary.




I reset my BIOS last week and formatted my HDD. The BIOS reset seems to have so far solved the issue with the system crashing. The fresh windows install also fixed a driver issue I had with my G502 mouse macros not working.

I noticed my motherboard temps rising up to 65°C on HWmonitor when the PC is under heavy load. I'm not sure how accurate those readings are but 65°C seems quite high to me and they used to be even higher in my old case (68°C). BIOS only displays CPU temps. CPU and GPU temps on HWmonitor seem stable (45°C, 55°C respectively) under load. Cool air blowing out the back of the new case. Cool might be a bit of an overstatement, but it's not warm at all.


I'm going to consider this as solved since the crashes seem to have stopped but the motherboard temps still worry me that the problem will return if they rise any higher. Best I can do for the board temps is buy 2 more front intake fans.