AMD Athlon II X4 640 possibly bad? But unsure if it is Software issue.

bgunner · May 25, 2014

this is in my second system. specs are as follows:

MB: AsRock A770DE+
Ram: DDR2 2x2GB PNY Optima PC2 6400 and 2X2GB Kingston HyperX PC2 8500 total of 8GB both running at 940 MHz.
CPU: AMD Athlon II 640 OCed to 3.46 with Phenom II 965 stock cooler.
HDD: 1 WD Blue 5400 RPM laptop drive 500GB and 1 seagate 160GB sata 1 drive.
PSU: Antec earth watts AE650W
GPU : XFX ATI Radeon HD 5770 1GB OCed to 930 core and 1348 memory.
DVD: liteon : exact cant remember but can get if necessary
Case: Cooler Master 690 II Advanced with 2 140mm side fans, 2 140mm top fans, 1 140mm front fan, 2 120mm bottom fans and 1 120mm rear fan.
Wi-Fi: Rosewill RNX-N300X Wi-fi card
OS: Win 7 Home Premium 64 Bit

Issue: crashes seemingly randomly. seems to happen most when son is playing video games but not when daughter is on google, PBS kids.org, playing her music.

have stress tested CPU with both Intel Burn Test for 40 runs on maximum stress and Prime 95 for 5 hours and no errors at all.

have tested ram with Memtest86+ for 11 passes and no errors

have tested GPU with OCCT GPU test with error Check enabled, no errors.

have tired each set of ram separately and in different slots and have run memtest86+ on each set of ram with no errors.

PSU was swapped with a Raidmax hybrid 530W with no change <-- spare PSU kept for testing known good.

Have changed motherboards from Asrock A785GM-LE to Asrock A770DE+ and have the same issue. Although this CPU was stable with the following components till I purchased my AMD 965BE CPU: A770DE+, PNY Optima PC2-6400 2x2GB, antec 650W, antec 900 II case, hitachi 500GB HDD, creative SB Xtreme gamer Sound card.

have tried different HDD's with no difference.

Have reinstalled the OS to make sure it is not an OS issue.

at this point the only thing I can point to is a Program conflict or bad CPU.

IF this is a program conflict, what is the best way of diagnosing this? what are the steps to take to find the culprits?

I want to make sure there is no other issue before condemning this CPU. It has only been used for a total of 2 years so I want to be sure it is a hardware issue and not software. All stress tests say the CPU is good so I want to lean towards software but I have reinstalled the OS and have had the same issues.

this system was build out of spare parts from me upgrading my system. This is why the parts are older.

Mouldread · May 25, 2014

Hey,

you can try installing http://www.resplendence.com/whocrashed and see what it tells you. You might be able to narrow it down to at least see if it's a hardware or a software issue. It helped me recently when I started getting BSOD. It turned out it was a GPU driver issue.

ingtar33 · May 25, 2014

the problem is your ram.

AMD cpus in that series are based on the PhII series k10 design. The k10 series were notorious for being unstable with "mix and match" ram. As in you have two different types of ram in your system~ even if they are running at the same speed the k10 is stupid sensitive to that. throw in an overclock as well and you'll get random crashing all day long.

Remove the unmatched pair of ram and you'll see the system stabilize pretty quickly. If that doesnt' fix the problem then i'd start to play with your north-bridge... sounds like it might be under-volted or under clocked... but my first guess is the mix and match ram. A close second will be your northbridge.

bgunner · May 26, 2014

ingtar33 :

you Must have missed this part of the post :

"have tested ram with Memtest86+ for 11 passes and no errors" this equals around 12 hours or so, AND-->

"have tired each set of ram separately and in different slots and have run memtest86+ on each set of ram with no errors." each time tested for 11 passes that equals 12 hours or so per stick/set.

meaning I have tried each set separately. Ran the PNY set only and the Kingston set only. I will try to up the North bridge voltage and see if this helps. Also the Phenom II's and the Athon II's are way less picky than you make them out to be. The Phenom 1's were and so were the Pentium 4's. Please don't think I'm a noob when it comes to hardware compatibility as I have a fare amount of experience with this, hence I knew enough to try each stick and each set of sticks separately.

bgunner · May 26, 2014

Mouldread :

thanks for the suggestion I will try this and see if it will pinpoint any issue and I will let you know with in the next week or two as the system only gets used when my son and daughter are here or I need to print something.

ingtar33 · May 26, 2014

bgunner :

I wasn't questioning if you had tested the ram properly with memtest86. Frankly, i don't expect this type of instability to be present in a memtest run. This isn't about BAD ram, just mix and match ram, which will unstabilize an overclocked k10 (btw: the original phenom, is also a k10 series cpu, and you're right, it was MORE sensitive to this type of thing, but just because it was MORE sensitive, doesn't mean it's not sensitive)

Try with your northbridge first if you don't think the ram is the problem, as i said, these symptoms speak to ram/nb issues, while i think it's slightly more likely to be the ram, that doesn't mean it's almost as likely to be the nb. If playing with the nb voltage doesn't fix it, and playing with the nb speed doesn't fix it, then look at the voltage to the ram. With 4 sticks of ram a k10 typically needed the ram to be volted another +0.1 or +0.2 volts over stock voltage to stabilize. Its just a quirk of the beast really.

bgunner · May 27, 2014

I have also run just the PNY set and crashed, ran just the kingston set and crashed not just ran memtest86+. It is not a ram issue with compatibility. I have already explored this option.

The OC does not include the north bridge since I did not OC it. I left it as close to stock at 2000MHz. I should also mention this issue is there OCed or not.

RobCrezz · May 27, 2014

First step would be to rule out software issue, do a wipe and fresh windows install.

bgunner · May 27, 2014

I have done so. The crash persists. I have tried to have my son change VOIP programs so to rule out those but they the system still crashes. It does not blue screen, the screen will go all wavy and lockup. The GPU is known good so not hardware issue there.

bgunner · May 27, 2014

Update: I have tried to use the WHoCerashed program and it says there are no dump files present. This does not surprise me because of the way it crashes. Thanks for the suggestion though.

Any other suggestions I can try?

have changed all hardware except ram and CPU But have tested ram and tried in all configurations possible.
Have reformated and reinstalled OS. Problem persisted.
Have upped the north bridge voltage from 1.78v to 2.25v But have to wait to see if this fixes it this coming weekend. If it does I will slowly back it down till it crashes and up it slightly again.
Have run intel Burn Test and prime 95 with out errors or crashing
have run OCCT for the GPU test with error checking enabled and past 1 hour without errors and crashing.
Have run Memtest86+ on all sticks individually and together and in sets with out errors. all tests were for 11 passes each.
Have checked the smart function on the HDD's and all is good there.

Now I'm out of ideas... :??:

ingtar33 · May 27, 2014

well, lets see how the upped voltage on the northbridge helps (or not)...

that said, i'll need some more info. what type of crash is it? black screen restart? blue screen of death? black screen shutdown? what types of situations are you in when it crashes?

hopefully the bumped nb voltage will work. but if it doesn't maybe we'll have enough info to take a few shots at this one.

bgunner · May 28, 2014

The PC only seems to crash while my son is play video games. It is not just one game because it happens with all of the ones he plays. LoL, GW 2, Warframe, TF2 all crash the same way and doesn't always happen. Some times he can game all weekend and no crash. When it crashes the screen goes wavy ( usually white and black lines with a few color lines thrown in ) and it will lock up. He has to do a hard shut down in order to get it usable again.

Because of how the screen goes wavy I thought it was a GPU driver at first so I have tried different drivers with no difference.
I also though that maybe the OC was unstable so I removed the OC by resetting the bios and just making sure the CPU voltage was good, Ram voltage and timings were correct and AHCI set. But the crash was still there.
I thought maybe there was a program conflict so I formatted a different HDD and installed the OS to that drive and the crash was still there.

This motherboard is Known good because it is stable with a Phenom II 965 oced to 4 GHz. I just can't seem to figure this one out on my own. I must be missing something, most likely something small in my face kind of thing.

I can not get it to crash by stress testing. I have tried OCCT's power supply test which runs both GPU, ram and CPU stress's but can not get it to crash.

ingtar33 · May 30, 2014

now that i have more info, it sounds like the gpu itself is overheating and crashing... try this.

Pop the side of the case off and stick a room fan in the opening, and see if the problem comes back. if it doesn't then it's "something" (in this case likely the gpu) overheating

bgunner · May 30, 2014

Not an over heating issue: Stress Test temps,
Intel Burn Test max CPU temp, 60°C. Gaming CPU temp 48°C. Idle temps 34°C.
OCCT GPU test with error check enabled max temp 80°C.<- this was a 1.75 hour run, by this time the GPU temp had stabilized for 1 hour or more. Gaming temp 72°C. Idle temps 36°C. -> XFX ATI HD 5770 reference design.
MB max temp 32°C

Case: Cooler Master 690 II Advanced with all optional fans added. 2 top 140mm fans, 1 front 140mm fan, 2 140mm side fans, 2 120mm bottom fans and 1 120mm rear fan.

Son should be here tomorrow to play to see if the NB voltage helped.

EDIT: should mention I have tried a Sapphire 4870 and had the same issue. the 4870 is kept as a spare for testing as known good GPU.

ingtar33 · May 30, 2014

well... the error you're describing is typical to a temp overheat of a non-cpu/mb part... purhaps a ram overheat... though i guess the NB could be overheating, and the motherboard doesn't monitor that temp to protect itself...

It also could be a slight "grounding" error, caused by dirt/dust/debris... and last but not least, it could be a dying hard drive... i had a hard drive go on a clients machine that caused this same type of crash once...

Hate to toss a bunch of ideas out there, but this type of crash is pretty indicative of hardware error not software or driver. driver and ram errors typically lead to blue screens, power errors usually lead to black screens... overheating can cause this if the part overheating isn't protected by the mb/bios, if a part overheats that is actively and accurately monitored typically the pc will just shut down without warning like someone pulled the plug.

I'm not sure it's the voltage to the nb anymore... though test it out... it sounds more like something dying or grounding out, or overheating. My list of "most to least" likely would go something like this...

1) gpu overheating/dying
2) something dirty and causing intermittent power fluctuation/grounding from dirt/debris or even a loose wire
3) something else overheating (something not monitored actively... most likely something on the motherboard, possibly ram, possibly the northbridge)
4) a dying hard drive
5) a dying motherboard
6)
7)
8)
9)
10) a dying psu (i put it way down the list because i've never seen a dying psu cause these symptoms, though i suppose it's possible)

Since you tried another gpu and got the same errors, we can strike the first possibility off the list. the rest we'll just have to test for the best we can. I'd like to eliminate temps completely by popping that side of the case off and sticking a fan in the opening, as it's the easiest (and cheapest) thing to eliminate from the list of suspects (i know your case has great airflow, but i can't tell you how many people have cases with great airflow who still end up with temp issues, so try this and see if it goes away).

You can test the psu by trying another... though having another on hand is the tricky part...

the dirt issue can be solved with some time with qtips and 90% rubbing alcohol... though i suggest you only do this if you are bored and don't mind wasting an afternoon, OR if the inside of the case is dusty...

you can eliminate a loose wire by pinching every wire at it's "end" near whatever plug it's part of and twisting it, if it moves it's likely not connected to the plug properly and might be causing an intermittent short in your system... i had a case fan cause this exact type of freezing issue once... hardest bug in the world to track down, found it by accident, when i was plugging the fan back into the motherboard, felt the wire "move". refrained from plugging that fan in and the problem vanished. Just an example, i mean it could be ANY wire, but it doesn't take much to cause power fluctuation bad enough to cause this type of freeze.

just tossing some ideas out there in hopes we can hunt this down.

bgunner · May 30, 2014

"It also could be a slight "grounding" error, caused by dirt/dust/debris... and last but not least, it could be a dying hard drive... i had a hard drive go on a clients machine that caused this same type of crash once..."

The system is very clean on the inside. Not only have I added a 280 x140mm silverstone magnetic screen to the side fans I make sure They stay cleaned as well. I opened the case last weekend and made my son clean the side, front and bottom filters. I am teaching him how to keep a system up to top performance. This rules out a dust issue.

It is not a HDD as I have changed HDD's in the sysstem.

Ram temps are well with in specs and the heat spreder's on the PNY only reach a temp of 40°C. This is cool considered the voltage is 1.85V. Tested with a thermal couple touching each module. Yes I tested for this. Odd aye?

MB is known as good. can swap the Phenom II 965 in and have no issues. Plus issue was there with the A785GM-LE motherboard.
HDD's have been swapped so this is not the issue.
all dust is clear, power wires plugged in fully
GPU known good, tested and not overheating,

"Since you tried another gpu and got the same errors, we can strike the first possibility off the list. the rest we'll just have to test for the best we can. I'd like to eliminate temps completely by popping that side of the case off and sticking a fan in the opening, as it's the easiest (and cheapest) thing to eliminate from the list of suspects (i know your case has great airflow, but i can't tell you how many people have cases with great airflow who still end up with temp issues, so try this and see if it goes away). "

this would show on temp monitors except the NB temp as this motherboard (A770DE+) does not monitor the temp there. But even when the system is at stock settings the system crashes so over heating the NB, I believe, should not be a concern.

"You can test the psu by trying another... though having another on hand is the tricky part...)"

have already done this as stated in My past posts. I keep a Raidmax 530W hybrid on hand and is known good for testing.http://www.newegg.com/Product/Product.aspx?Item=N82E16817152028&cm_re=raidmax_530w-_-17-152-028-_-Product the only difference is mine is the version 1 without the ability to turn off the the fan LED. (ALL HARDWARE Except CPU and Ram has been changed and issue persists.

"you can eliminate a loose wire by pinching every wire at it's "end" near whatever plug it's part of and twisting it, if it moves it's likely not connected to the plug properly and might be causing an intermittent short in your system... i had a case fan cause this exact type of freezing issue once... hardest bug in the world to track down, found it by accident, when i was plugging the fan back into the motherboard, felt the wire "move". refrained from plugging that fan in and the problem vanished. Just an example, i mean it could be ANY wire, but it doesn't take much to cause power fluctuation bad enough to cause this type of freeze."

I will try this and see if there is any change But Pretty sure the wires are OK, But never hurts to check.

"just tossing some ideas out there in hopes we can hunt this down."

some of your suggestions I have already tried, some are N/A But I will surely try what I have not thought of yet for tests like the wire thing.

Thanks I will get back to you after the weekend as I want to see if the NB voltage helps first then will move on to the wire issue.

bgunner · May 30, 2014

I do have one question though, should I test each wire plugged in or unplugged? The twisting I am referring to here.

ingtar33 · May 31, 2014

well... i'd do it with the system powered off, and unplugged. i'd also unplug the "plugs" from their sockets/pins. all you're doing is seeing if you can find a loose wire that wiggles a bit. even twists freely. That fan wire was actually secured inside the plug. it just was BARELY secured, so if it vibrated a little bit, it sometimes would interupt the power flow. Craziest thing ever really... i mean all it was, was a little play on the wire, but as you know, there shouldn't be ANY play on these wires...

As for other things to try... If the nb voltage doesn't help, i'd simply unplug every device and fan in the system that you don't need for gaming. See if the issue comes back. it's possible it's something on the "outside" of the core of the system... as it looks like you've mostly tackled the "major" stuff. A short in the cd rom drive or maybe a fan with a bad coil... we don't know what it is, but the idea is to strip out as many possibilities as possible. if the problem vanishes then at least you know WHERE to look... heck if the problem remains your list of suspects gets really short...

bgunner · May 31, 2014

Awesome Ideas to try thanks I will get back to you on the results of the tests.
Because it could take a few weeks to test please keep this thread open Moderators. thank you for your patents in this matter.

I play on my system and not on the secondary system. Due to a Physical disability, an inability to sit at that PC, I cant play on my secondary system to test it. This means testing is done when my son, 17yrs, that is here only on the weekends. My daughter that is only 5 at this point, but is here during the week PBSkids.org is her speed for now, can't test the system because it wont crash with her using the system. It will take time to test due to my inability to sit at that PC. And yes I meant sit at that PC, It is in my home but I can nit sit there for long.

ingtar33 · May 31, 2014

no problem. we'll figure it out. this is one of those "ghosts" in the computer that can drive you batty to find. Hopefully we can narrow down where it is.

my biggest fear right now is it's some inherent instability in an AMD driver that happens to affect only the games your son plays. (the chances of that grow exponentially smaller with every game he plays; so i don't even want to think about this as a possibility right now, but it is "something" outside of hardware that could be causing it)

bgunner · Jun 3, 2014

No crash this weekend But because my son had to work both days he didn't get to use it much. I would like to wait one more weekend and have him play more to see if it will crash. It has gone a weekend or two without crashing before so I want to be sure that upping the NB voltage either helped or did not before proceeding.

ingtar33 · Jun 7, 2014

alright... well give us an update on monday, btw: happy father's day!

bgunner · Jun 16, 2014

OK well the system has not crashed since the upping of the north bridge voltage. The one thing I am concerned about is it is not getting the long gaming sessions it had in the past when it was crashing. Basically the same circumstances have not happened to try to recreate the crash situation.

I am now wondering " What is the max safe voltage for the AMD 770 north bridge?". I have done a search on google but only pull up forum threads for the same question. I can't seem to get a data sheet about it or a solid answer from the forum posts. At the moment I have it set for 2.25v But I think this may be high for the 770 chipset but I can't seem to find a solid answer.

ingtar33 · Jun 16, 2014

well... i had an old m4a78t-e motherboard with a phII x4 965be, overclocked out of it's mind. though that board had a AMD 790GX northbridge, it was keeping the cpu/nb voltage at 1.4V all day long for years. no immediate harm. (though eventually one of the ram channels died on the board, which might have been the results of that voltage... that said, it was running at that voltage for over a calendar year

1.8V? which northbridge voltage did you pump up that high? the cpu/nb or the nb? cause i wouldn't run the cpu/nb up that high, wouldn't run the nb voltage that high either... are you talking about the 1.8v northbridge voltage? cause that is what it's supposed to be at.

bgunner · Jun 16, 2014

This is an older DDR2 model so it only has just the ability to just the "North Bridge" voltage. Here is a link to the board for reference : http://www.asrock.com/mb/AMD/A770DE+/

Well when my Phenom II 965 was in it OCed to 4.075 GHz I only had to run the north bridge voltage at 1.25v to be stable. With this Athlon II 640 OCed to 3.45 GHz I had it at 1.1825 or 1.1875 and it was unstable. so i bumped it to 2.25V so see if it would stabilize. Well it seems to have. Not Oced the system is was still unstable at the 1.8+ voltage. So I would like to know what the max safe voltage for this chip set is so I know not to exceed it.

The motherboard try's to set this voltage to 1.1 volts when the defaults are used. So it is not the 1.8 voltage.

EDIT: I found a really small data sheet but does not say safe voltages. http://www.go-gddq.com/down/2011-08/11082014389706.pdf

AMD Athlon II X4 640 possibly bad? But unsure if it is Software issue.

Judicious

Distinguished

Glorious

Judicious

Judicious

Glorious

Judicious

Expert

Judicious

Judicious

Glorious

Judicious

Glorious

Judicious

Glorious

Judicious

Judicious

Glorious

Judicious

Glorious

Judicious

Glorious

Judicious

Glorious

Judicious

Share this page