Powercolor Radeon 4670 VPU Recovery Errors

Hawky

Distinguished
Apr 4, 2009
26
0
18,530
I had gone over to a friend's house to uninstall his old Geforce 8200 GT and install the new Radeon 4670, trying as best as possible obviously not to touch the 4670's connectors I had touched it by the edges and tried to not apply any pressure to it as it was installed. Screw in and all, turned the PC and it worked perfectly.

The computers specs are as follows:
Operating System: Windows XP Professional (5.1, Build 2600) Service Pack 3
System Motherboard: G31M-ES2L
BIOS: Award Modular BIOS v6.00PG
Processor: Pentium® Dual-Core CPU E5200 @ 2.50GHz (2 Cores)
Memory: Kingston DDR2 2046MB RAM
PSU: 400w (card usually only takes 300w)

The powercolor URL to the GPU.
http://www.powercolor.com/eng/products_features.asp?ProductID=4560

The VPU errors occured while in game only. We had tried using the updated drivers from the website itself, powercolor, however after an hour or so the VPU error had kicked in. Uninstalled them and used the disc's drivers, games then went for a good 5ish hours at least before the VPU error kicked in.

We also don't have any heat issues, GPU-Z came up with 60 degrees when in game and besides that, the case is sometimes wide open (no side panel) to allow more airflow as needed.

The most updated driver ver from powercolor is:
Driver + Ver: HD4650/ HD4670 Series PCI-E & Catalyst 9.11 + Version 8.67.
Date: 26/11/2009

Now, I myself am also using a Radeon 4670 only I am using ASUS.
I have never had a single issue and my driver ver+date (dxdiag) is:
Driver Version: 7.15.0010.0176
Date: 2/4/2009

Lucky enough for me, ASUS provides the newest + older drivers unlike Powercolor, so I am unable to find any older Powercolor drivers. We also tried the ATI Radeon drivers but had installation errors, we tried again and this time currently the setup has not generated any errors.

The ATI drivers worked, but then produced another VPU error an hour later.

Anyone got any ideas at all? This is plain frustrating.
 

Hawky

Distinguished
Apr 4, 2009
26
0
18,530
I believe the issue is a PSU failure. I'll run a Memtest just incase, though, I believe that's fine.

He's using a 400w "PowerTek PTI-400AR-T". Now, I haven't been able to find the exact model online but I did find one 770w version on newegg and that power supply got reviews like...

Newegg

Cons: Inserted a EVGA 8800 Ultra 768MB today. After 5 minutes of playing Crysis Warhead on GAMER settings, it smelled like burnt rubber. I sniffed around, smell was coming from a PSU. Turned it off immediately

Cons: The first one blew up and killed my $150 mobo.

Cons: Not worth $1...don’t risk buying this power supply. It has too many defects. Integrated transformers can not withstand to much head...50C? Too much heat? Yes, way too much heat for this power supply...So don’t take a change of burning your motherboard or anything else for that mater.

Cons: After about 4 months it blew up and killed my board.

Any other ideas? Anyone? =/
 

Hawky

Distinguished
Apr 4, 2009
26
0
18,530
Now more issues are occuring, after 2days with dealing with VPU errors - now when my friend plays at max resolution, things look a bit pixellated and graphical quality has lowered. When playing at a diff, lower, resolution - things look decent again.

Is this graphical issue with the monitor, or another effect of the PSU? I'm starting to think I bought a failed card.

Any other ideas?
 

Hawky

Distinguished
Apr 4, 2009
26
0
18,530



In summary, I have tried the following solutions:
*Tried 4 graphics drivers for it and didn't help.
>Powercolor latest, powercolor disc, ATI latest, ASUS latest.
*Tried turning off the windows paging file and didn't help.
*Made sure the graphics card wasn't overheating, nothing here.
>GPU-Z came up with 60 degrees when in game and besides that, the case's side panel is sometimes removed.
*Tried registry & file sweeping to get rid of ALL graphics driver remains ever used on the system, didn't help.
*Tried reseating the graphics card on the PC's motherboard, didn't help.
*Tried to test system memory via memtest86.com and had no errors.
*Tried disabling VPU but that only caused blue screens of death.
*Tried various graphics driver+game settings and nothing.

Nothing worked, but that's a good enough place for you to start lol.

And I think it's either:
A) His power supply
B) A faulty card

What's your power supply?
 

Hawky

Distinguished
Apr 4, 2009
26
0
18,530


The HKC isn't a brand I would trust, especially when it comes as a default PSU for a case. Those things are usually cheap, crappy and generic.

Avoid the generic plague, they have bad cheap hardware, and sometimes incorrect voltage/wattage readings on the side of the PSU.

Generic ones tend to literally blow up in smoke after a few days to a few months, if your lucky to keep it going that long, and can take your motherboard/graphics card with it etc.

If you could, also try this registry fix, it disables ATI driver logs from logging which can collide in games.

Windows Registry Editor Version 5.00

[HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Services\Atierecord]
"eRecordEnable"=dword:00000000
"eRecordEnablePopups"=dword:00000001

Copy+Paste that quoted text into notepad, save to your desktop as a .reg file, double click it and merge.

Currently I'm trying this fix, see what happens, it's been confirmed it has helped a few on the AMD forums.

Edit. Didn't help =/
 

Hawky

Distinguished
Apr 4, 2009
26
0
18,530
Orly? I'm using an ASUS Radeon 4670, no issues at all, and the 4670 was deemed "The mainstream market" card at late 08, early 09.

I'd still perhaps look into your PSU, maybe see if you can flash your Graphic card's BIOS with another manufacturers, make sure all your other drivers are up to date, and your chipset+BIOS drivers work well for the card.

May also have to check your memory sticks, as they may not be giving the correct voltages/mhz, so you'll have to check what the right volatages/speeds for both the memory and motherboard, then either underclock/overclock to get 'em correct.
 

Distorted_13

Distinguished
Dec 18, 2009
12
0
18,510
I have a overclocked cpu (2.8 Ghz => 3.6 Ghz... to high? =D) when i did that my ram automatically became overclocked too. i tried to get the standard speed and it still crashed. Now i'll try putting my cpu to normal speed and see if that works.
 

Distorted_13

Distinguished
Dec 18, 2009
12
0
18,510
Now i feel stupid cuz i didnt think of that before... my psu cant handle a cpu overclocked that much with that graphics card. but performance with overclocked cpu + graphicscard downclocked is still the best so ill keep playing this way.
 

Hawky

Distinguished
Apr 4, 2009
26
0
18,530



I'd also look into heat.
 

Distorted_13

Distinguished
Dec 18, 2009
12
0
18,510
When playing games like crysis the heat of my graphics cardcan get to 80°C, but it doesn't cause any problems... yet. I'll take care of that later. And my cpu is actually very stable like this, it doesnt go higher then 40°C when playing games.