Sign in with
Sign up | Sign in
Your question
Closed

Monitors losing signal from 5850

Last response: in Graphics & Displays
Share
September 15, 2010 1:39:21 PM

I'm having a problem with my system and my Asus EAH5850 Direct CU.

A few weeks ago my system started to freeze during games and benchmarks. I was playing Fallout3 GOTY, Metro 2033, and running Vantage every now and then. It seemed that randomly during 3D use the system would freeze.

Symptoms:

Both monitors lose signal and go in to power save
keyboard lights stay on but is non responsive
hdd light indicates activity
power button has no affect
all fans spin normally

I'd end up holding the power button down or using the reset button, and Windows would always report "the last shutdown was unexpected" but never have any other info or errors in the logs.

I swapped the card with a Visiontek 3870 OC I had left over from the last GPU setup and I used the system for 2 weeks or so while I had the 5850 sent in for repair and I never had an issue.
I got the card back on Monday and yesterday the issue happened again 3 times, with the last time the system actually shut down softly when I pressed the power button. This time the issue happened when I was working in Windows, either just watching a movie or copying files. I successfully played Dragon Age Origins for an hour or so without issue.

Troubleshooting I've tried:

Uninstalled drivers using driver sweeper in safe mode and re-installing both the Asus and AMD drivers several times before RMA.
reverted back to stock clocks on system and GPU
Uninstalled all other hardware
Uninstalled all "non critical" software

I was thinking it may be a power supply issue, but I haven't found a spare power supply with the necessary connections for the system.

Note: I had a power surge that fried a lot of equipment in my house around the time this all started, including all network equipment on every PC in the house. My mobo has 2 NICS, the connected one at the time was fried. I switched to the other NIC and all is fine without the 5850. I'm connected to a 600 watt UPS.

More about : monitors losing signal 5850

September 15, 2010 10:08:41 PM

Here is the output from the bug check. I have not found any pertinent info about "WHEA_UNCORRECTABLE_ERROR (124)".


*******************************************************************************
* *
* Bugcheck Analysis *
* *
*******************************************************************************

WHEA_UNCORRECTABLE_ERROR (124)
A fatal hardware error has occurred. Parameter 1 identifies the type of error
source that reported the error. Parameter 2 holds the address of the
WHEA_ERROR_RECORD structure that describes the error conditon.
Arguments:
Arg1: 00000000, Machine Check Exception
Arg2: 874db4bc, Address of the WHEA_ERROR_RECORD structure.
Arg3: 00000000, High order 32-bits of the MCi_STATUS value.
Arg4: 00000000, Low order 32-bits of the MCi_STATUS value.

Debugging Details:
------------------


BUGCHECK_STR: 0x124_GenuineIntel

CUSTOMER_CRASH_COUNT: 1

DEFAULT_BUCKET_ID: VISTA_DRIVER_FAULT

PROCESS_NAME: System

CURRENT_IRQL: 0

STACK_TEXT:
8e392cc4 83105fc5 874db4a0 8319ec28 8319ec20 nt!WheapCreateTriageDumpFromPreviousSession+0x32
8e392ce4 83106def 8319ec20 874db4a0 8319ec50 nt!WheapProcessWorkQueueItem+0x56
8e392d00 830a4f3b 8319ec20 00000000 85bb4d48 nt!WheapWorkQueueWorkerRoutine+0x1f
8e392d50 832456d3 00000001 b8cf1ee3 00000000 nt!ExpWorkerThread+0x10d
8e392d90 830f70f9 830a4e2e 00000001 00000000 nt!PspSystemThreadStartup+0x9e
00000000 00000000 00000000 00000000 00000000 nt!KiThreadStartup+0x19


STACK_COMMAND: kb

FOLLOWUP_NAME: MachineOwner

MODULE_NAME: hardware

IMAGE_NAME: hardware

DEBUG_FLR_IMAGE_TIMESTAMP: 0

FAILURE_BUCKET_ID: 0x124_GenuineIntel_PROCESSOR_BUS_PRV

BUCKET_ID: 0x124_GenuineIntel_PROCESSOR_BUS_PRV

Followup: MachineOwner
Score
0
September 21, 2010 12:05:00 AM

Sweet, only talking to myself here...

Anyway, i'm still trying to solve this issue. As I write this I've got the 5850 in my WinXP PC with a P4 D 960 running FurMark. It's running fine after 7 min @84 FPS and 70 degrees C.

The system has a 500 watt Cooler Master power supply, and it seems it's running just fine if Fur Mark doesn't kill it.

I guess the next step is to do a fresh install of Windows 7 on my main PC and put the card back in.

After that the only thing(s) left that would seem to make sense is the power supply.
Score
0
Related resources
September 21, 2010 5:04:47 PM

firebird said:
Sweet, only talking to myself here...

Anyway, i'm still trying to solve this issue. As I write this I've got the 5850 in my WinXP PC with a P4 D 960 running FurMark. It's running fine after 7 min @84 FPS and 70 degrees C.

The system has a 500 watt Cooler Master power supply, and it seems it's running just fine if Fur Mark doesn't kill it.

I guess the next step is to do a fresh install of Windows 7 on my main PC and put the card back in.

After that the only thing(s) left that would seem to make sense is the power supply.



Hi Firebird,
I've been experiencing the exact same issue with my computer. Random loss of video signal to monitor. I've switched out my PSU and the problem persists.. I'm currently in the process of RMA'in my card but from what I'm reading here is it didn't solve the issue for you. Can you let me know your computer specs so we can compare what we have in similar and such

mobo- p6x58d premium
cpu- i7 920
gpu- ati 5970
ram- 6gig redline mushkin
psu- 850 hx corsair
Score
0
September 21, 2010 5:06:13 PM

ahh, my apologies.. i see the specs under your thread.
Score
0
a b C Monitor
September 21, 2010 5:14:59 PM

Try testing your RAM with memtest. If necessary, underclock your CPU or use higher voltages. If the HD 5850 runs fine on another system, then clearly something is wrong with the hardware in the other (unless you're really unlucky with installing drivers). Since the system was able to run a card that takes lesser power, then maybe you have a faulty PSU. Have you tried running the card in the other PCIe slot?
Score
0
September 21, 2010 5:30:55 PM

skolpo said:
Try testing your RAM with memtest. If necessary, underclock your CPU or use higher voltages. If the HD 5850 runs fine on another system, then clearly something is wrong with the hardware in the other (unless you're really unlucky with installing drivers). Since the system was able to run a card that takes lesser power, then maybe you have a faulty PSU. Have you tried running the card in the other PCIe slot?


I have tried several variations of voltage and clock speed on my mem and CPU/chipset with no change.
I have ran memtest over night with no issues.

Last night I successfully ran the 5850 in my P4 machine for an hour of Fur Mark. It kept the CPU usage high, and the GPU pegged and there were no issues.
I re installed Windows 7 with the 3870 in it (prolly not the best idea, should have used the 5850) and when I get home tonight I'll swap in the 5850 and see what happens.
I did not install any video drivers yet, so I'm not anticipating the CCC drivers to cause the issue.

I thought about the power problem, but I don't have another high-end PSU to test with, and I've got my cables ran really nice in the gaming machine. Would make me :(  to have to undo it all to swap in the 500 watt.

@Extole

What brand is your 5970?
Have you been upgrading drivers in Windows and from AMD regularly? Just looking for any similarities between our situations.
Score
0
September 21, 2010 6:02:02 PM

The brand is Diamond. I have tried nearly every driver since 10.4 and till recent 10.9. I ran driver sweeper every time in safe mode etc. I to have ran memtest and all passed. What really gets me about this issue is that i can't repeat the problem on demand.. It is so random. It can happen when I'm doing something simple like browsing the internet or typing up my research paper for 2 hours *face palm*. Then it would run fine for a few days to a week then start up again. I do a lot of gaming also and sometimes able to go on for hours without a hitch. There is nothing in event viewer log that indicates a problem either except for when i do the hard shut downs. I've switched out my PSU 3 times and that didn't resolve the issue. I did have one weird experience once when the monitor lost signal. The video card fan jumped up to 100 percent fan speed which i thought was really unusual. Has that ever happen to you? (i don't believe its an over heating issue either cause ccc and gpuz is telling me the temps are idling at 50c when this occurs)


also, when this issue occurs do you still hear music or your video playing in the back? I was even watching a live stream and heard the guy talking for quite a while after i lost the signal so it doesn't appear my whole system is locking up right away or if at all.
Score
0
September 21, 2010 6:32:00 PM

Well, now our issues are sounding very similar....

I did the same driver rollbacks and updates that you did....
The issue happens while working on the desktop, gaming, watching movies, or just sitting idle...
When it happens I will hear the sound continue for 15 seconds or so
A few times the fan on the GPU has kicked on high when the monitors lose signal and the system locks...
Once I was able to press the power button and I heard the Windows logoff sound and it shut down softly but there was no display...

Not sure if I mentioned, but Asus has auto emails for your card at every status change during the RMA process. My card was in and out in the same day which leads me to believe that "repairing" phase resulted in NPF. I wish they would have indicated if there was a repair done or not.
Score
0
a b C Monitor
September 21, 2010 6:42:30 PM

Send it back in and tell them that they didn't fix anything. Ask for a replacement.
Score
0
September 21, 2010 6:48:35 PM

If I have the same issues again tonight I'll contact Asus again and work on getting another card as an exchange instead of repair.

Makes me wonder if I should have stuck with the good 'ol Sapphire cards or Visiontek. I never had an issue since my 8500 when sticking to those vendors.
Score
0
September 21, 2010 6:56:40 PM

well, I'm going to rma my card in tomorrow. Not sure how long Diamond's turn around time is but as soon as i get it back i'll report back here to let you know if it did the trick or not. Can you let me know if you find any resolution in the meantime as I too want to get this dreaded issue resolved
Score
0
September 21, 2010 7:07:55 PM

Extole said:
well, I'm going to rma my card in tomorrow. Not sure how long Diamond's turn around time is but as soon as i get it back i'll report back here to let you know if it did the trick or not. Can you let me know if you find any resolution in the meantime as I too want to get this dreaded issue resolved


I should know tonight if the re-install fixed anything. If it doesn't I might have the energy to swap in the 500 watt power supply, but I think it's going to struggle with my gaming system. According to my UPS I can pull upwards of 500 watts at times with the 5850 in.
Score
0
September 21, 2010 11:53:20 PM

I just finished running a 30 min. test of Fur Mark without issue.
I will install more stuff and post again after running for a while.
Score
0
September 22, 2010 1:04:31 AM

firebird said:
I just finished running a 30 min. test of Fur Mark without issue.
I will install more stuff and post again after running for a while.



Cool let me know how it goes, cause then maybe it could be something else? like maybe motherboard issue?
Score
0
September 22, 2010 12:55:27 PM

After I ran Fur Mark I re-installed a few apps and tested again....

3D Mark Vantage caused the issue after it's 2nd or 3rd loop.

I had just installed Vantage and the latetest version of PhysX (513) when the issue happened. I un-installed that version of PhysX and installed the version that comes with Metro 2033, but that caused the issue as well. So, I un-installed that version of PhysX and then re-installed Dragon Age and let it install PhysX. My thinking there was that my issues started around the time I installed Metro 2033 and a began experimenting with hybrid PhysX by running an 8800 GTS with my 5850 ( I abandoned it after I started getting the crashing. That was the first time I installed the newer versions of PhysX right from Nvidia's site.)

So this morning and last night I played some Dragon Age and didn't have an issue, but I didn't play long enough to know for sure. I have yet to have a problem under normal desktop use, but I don't have all my apps installed yet and haven't done any of my usual day-to-day stuff.

<sigh> Looks like I won't know for another day or so.
Score
0
September 22, 2010 3:44:05 PM

firebird said:
After I ran Fur Mark I re-installed a few apps and tested again....

3D Mark Vantage caused the issue after it's 2nd or 3rd loop.

I had just installed Vantage and the latetest version of PhysX (513) when the issue happened. I un-installed that version of PhysX and installed the version that comes with Metro 2033, but that caused the issue as well. So, I un-installed that version of PhysX and then re-installed Dragon Age and let it install PhysX. My thinking there was that my issues started around the time I installed Metro 2033 and a began experimenting with hybrid PhysX by running an 8800 GTS with my 5850 ( I abandoned it after I started getting the crashing. That was the first time I installed the newer versions of PhysX right from Nvidia's site.)

So this morning and last night I played some Dragon Age and didn't have an issue, but I didn't play long enough to know for sure. I have yet to have a problem under normal desktop use, but I don't have all my apps installed yet and haven't done any of my usual day-to-day stuff.

<sigh> Looks like I won't know for another day or so.



Hmm i think ill play around with the Physx drivers to then.. I did however do a couple things last night and didn't run into that issue after the changes.. not sure if its fully fixed yet or not cause i only had a day of testing it but here is what i did...

- ( i really hope this wasn't the problem) but my atx 24 pin connector to the mother board is connected to an atx extension cable.. I bypassed that and connected it directly to the mother board and made sure it was secured in the slot.. took a little bit of force to have it click in than i thought it would.

- I also went into my power saving options in the control panel and set everything to full power and made sure that stuff like "turning off hard drive after 20 min" was set to like 500 min or as high as it would go. Also disabled all hibernation settings too.

Again I'm not sure if these changes fixed the issue just yet but ill keep you updated for the next couple of weeks. In the meantime firebird I suggest you double check your atx cable to your mother board and make sure its fully plugged in or bypass the extension if you do have one just as an extra saftey measure. I know it sounds stupid but i read a few threads out there where people were having similar problems as ours and it was due to not plugging the cable in all the way. I'm hoping i just had a bad atx extension *crosses finger*
Score
0
September 22, 2010 4:03:46 PM

I did see that in a post as well, that the ATX extension cable was the culprit. I don't believe mine has an extension, but I'll mess with that and the power settings as well if I have the issues tonight.

Seems strange that Fur Mark doesn't cause the issue but Vantage does. Maybe it's OpenGL or the fact that only 1 core gets stressed with Fur Mark....
Score
0
September 22, 2010 4:33:35 PM

Oh and firebird, when this issue happens to you did you ever see any red led's light up on the card? I've seen it once a long time ago light up but never saw it again. From what i've read is that it could that the graphics card is not getting enough power or faulty card. I will try furmarks tonight and see if i can replicate the problem.
Score
0
September 22, 2010 4:55:52 PM

I have a fan in the side window of the case that blocks most of the view to the inside and I've never noticed the red led. Maybe I'll position the case so I can see the card.
Score
0
September 24, 2010 2:09:07 AM

I checked my ATC power connector and it's not using an adapter and is connected securely.
I changed my power settings to "high performance" and set everything to never shut off.

I had been using the PC for a few hours without issue.

About 1/2 hour into DAO monitors went black and PC froze. Since I've had the new install I'm getting different Windows errors, stating I have a "blue screen" error with the same IDs as before.

<sigh>

I might try putting the SupremeFX soundcard that came with the mobo back in and see if that helps.

I'm wondering if something got fried when I had a power surge that took out a lot of electronics in my house and I'm just having a hard time finding which component....
Score
0
September 24, 2010 9:36:27 PM

So I decided to let my local PC shop have at it and see if they could figure out what was wrong. They called me back after a few hours and said that there were able to get the PC stable through multiple Vantage runs (which wasn't possible before) by lowering the core to 699 MHZ and droping the vcore to .98 V.

I told them they could play with it all weekend as I'll be out of town, but it really seems like the card is bad and Asus didn't fix anything. Unless they call me back with different news, I will be demanding a NEW card.
Score
0
a b C Monitor
September 24, 2010 9:39:16 PM

skolpo said:
Send it back in and tell them that they didn't fix anything. Ask for a replacement.


Could've saved you four days of hassle =P
Score
0
September 28, 2010 6:48:19 PM

Heya Firebirdm
I Just wanted to conclude that i think my issue was dude to that atx cable connecting to the mobo being defective.. I haven't ran into the issue again after replacing that cable. I was wondering how your progress is going so far and if it indeed was the video card for sure?
Score
0
September 28, 2010 6:57:24 PM

I have an RMA # again from Asus, but they told me: 1) we will not give you a new card 2) the last repair was "No Problem Found" and 3) I might be able to contact them while the card is in for repair to request additional testing.

Because that sounds so crappy I'm going to do some power supply testing when I get home tonight. I noticed my 12v was down to 11.5v at idle with the 3870 attached. I'm going to measure it with a voltmeter and keep an eye on it when it's loaded to see if it drops. I would rather have a new P.S. overnighted to me versus waiting on Asus to not do anything again for a week and have to spend more money on shipping.
Score
0
a b C Monitor
September 28, 2010 7:05:57 PM

Hey guys I hope things work out for you but just my thoughts...
firebird, try removing the 8800GTS from the PC. Might be sucking up too much juice for the PSU.

How were you enabling hybrid PhysX? GenL? I've had zero issues with his patch.

At one point I randomly had my monitor go into sleep mode when the login screen would appear. It was fine up to the windows loading splash screen, then go blank as soon as the login screen would appear. Sound still worked through my speakers. I typed in my password and I heard it login, but still no video. If I booted into safe mode I could get in and deleting my drivers would allow me to reboot normally. Then, reinstalling the ATI drivers and rebooting, and again blank screen. I tried a couple different versions of the drivers, always booting into safe, running driver sweeper, rebooting normally and installing and every single time it would blank screen on the next boot after the driver install.

I ended up reinstalling Windows 7 64bit fresh, and haven't had any issues since. Don't know what caused it but a fresh install fixed it.

You might consider partitioning your drive and setting a windows backup on it, then format and reinstall windows on the main drive and see what happens. If it works try restoring the backup and see what happens.
Score
0
September 28, 2010 7:14:09 PM

This might not have been clear before, but I removed the 8800GTS when the lock ups started happening and I've re-installed Windows since but the problem persists when the Asus card is in.

I did have the hybrid setup working, but since the lock ups started at that time I assumed it was something to do with that.
Score
0
a b C Monitor
September 28, 2010 7:16:51 PM

Oh damn so no go in either case... Huh..

You may want to try flashing the 5850 with a different VBIOS and see what happens. I can maybe email you my reference model Sapphire 5850 BIOS.
Score
0
September 28, 2010 7:27:11 PM

I'm starting to lean towards power.

It seems more plausible since I've noticed the low idle current on the 12v rail and the fact that if the voltage and clocks are lowered on the 5850 (according to the PC shop) the system runs fine.

Before I had the 5850 I had a 3870x2 and 3870 in Crossfire over clocked to 850 core with not trouble. I even maxed out my UPS a few times by pulling 600 watts.
With the 5850 I think 350 is the most I've pulled in.
Score
0
September 28, 2010 7:41:06 PM

Hmm... I know this is a little far fetched but have you tried bypassing the UPS to see if that helps at all? Maybe it got damaged in that black out? I'm also thinking this is power related too cause you and I were having identical problems. I know I mentioned this before but i would just go through every cable in your PC to make sure all cables are plugged in tightly so that it clicks in.. The best test tho would be to see if you can find another PSU to test the system on.


-Edit- Did the PC shop try testing the GPU card in a different system and run vantage? Cause if it was stable on a different system that would let us know that the gpu isn't the problem.
Score
0
September 28, 2010 7:50:49 PM

i don't have any other UPS sytems around to test with. Just last night I lost power in the middle of the day with no storms or anything obvious and the UPS took the load fine and kept my machine running. I'm about to make a claim for $1000 worth of damage to various electronics around the house, and I'm not really willing to not use a UPS for anything longer then a test.

When I RMAd the card the first time, took the power supply out and re-installed it, running the cables really nice and neat. I think it took me an hour and a half. I had the issue before that and I have it now (would I really have had cables loose before and after that?). I'm hesitant to dismantle it (looks so nice :)  ).

I will check all connections again and re-seat some when I'm measuring voltage. <sigh>

So much work. Too bad I don't have the $$$ for an entire high-end system of spare parts.....
Score
0
September 28, 2010 7:53:19 PM

They did test it separately, but then had me bring my PC in because they only had a single core P4 system to test with (so sad) and they didn't have any software to stress it with.

More and more seems to point to my system.....
Score
0
September 28, 2010 9:34:24 PM

Actually, scratch the whole ups thought.. I forgot you took it into the shop and was on a different power source and it was still failing.. Oh and I'm not sure if you already answered this but did you ever see any red LED's on the card when this occurred (usually an indicator card is not getting enough power)? I don't know if this is recommended or not and maybe someone on here can shed some light but I'm curious to what would happen if you tried pushing a little more volts into your card. If either it would be more stable, less stable or the same. Reason I bring this up is i think it would help to narrow down it being a power issue or not. Again, this is usually practiced with overclocking the card and of course whenever overclocking there are some risks. If you are interested in doing this, the program i use to send more volts to my card is MSI after burner. Another good test is to see how much voltage your card is getting before doing this and how much voltage you get after you over volt to see if it is correct.

-Edit-
If you do decide doing this definitely read up on it in regards to what the safe voltages are for the card when it comes to over clocking etc.. But, as how everything is going it does seem like everything is pointing to power tho being that the computer shop putting less stress on your system was able to get it more stable sounds like the card isn't getting enough power..
Score
0
October 4, 2010 5:00:18 PM

Went to a little LAN party this past weekend and put the 5850 in another PC. It ran @850-1200 (was using the timings from the 5870 that was in the system). and gamed without issue for several hours.

I have a new Seasonic 650x power supply in my PC and that doesn't help either.

I guess it's the mobo....(runs with a 3870 just fine though)
Score
0
October 11, 2010 3:38:20 PM

Well, I thought I'd update this since I'm not quite ready to close it yet...

I'm getting a new mobo, cpu, and memory this week. Rampage 3, i7950, Scythe MG2, 6GB Mushkin Blackline 1600 and another 5850.

I will post back on the results of the new system.

On a related note, I've also been involved in this post: http://forums.amd.com/game/messageview.cfm?catid=260&th...

Seems like I'm not the only one having this type of issue.
Score
0
October 14, 2010 1:44:35 PM

Well, I finally have a working computer now. I ended up adding my PC to my claim for the lightning/power issue that took out a lot of electronics in the house (and in the area from what the insurance company said).

I replaced my mobo, cpu, memory, and added a 2nd Asus EAH5850 Direct Cu. I went with an Asus Rampage 3 Formula, i7950 with Scythe Mugen R2, 6 GB Mushkin Blackline memory 1600 MHZ.

I finished running all benches last night and using the same HDDs and power supply as on the other board I've been rock stable.

So I'm going to chalk this one up to a bad motherboard (it was very unstable yesterday before I got the new parts....)

I guess a moderator can close this now, and I can't choose a best answer because nobody mentioned a motherboard :(  and I can't select my own answers :) 
Score
0
a b C Monitor
October 14, 2010 1:54:32 PM

This topic has been closed by Mousemonkey
Score
0
!