WARNING ATI users: Catalyst 3.8 can kill your Card / Monitor

[Evnas]

One other thing I noticed before my card died that I forgot to mention. Most of the time my monitor wouldn't shut itself down after the time I specified in power profile. I had it set to shutdown after 15min

Ive had that happen with a number of video cards, and i dont think it would be a cause of your failing video card. With my old GF2 MX400 in, my monitor doesnt shutdown at all unless i physically turn it off, lol

 

[nealric]

Interestingly enough, I rolled back to the 3.7s and started experienceing problems A lot of games crashed and I got artifacts.(my card is brand new so it went straight to the 3.8s)

Im now on the leaked omegas with no problems (and oced)- and my card is not that hot (at least using the touch test since I dont have a probe on my card)

 

[Albuquerque]

To be very honest, I'm having a hard time believing that a $300 monitor would burn out because of high refresh rate... This specifically goes to Mr. Overkill.

Someone already posted this, but it seems that nobody read it: nearly all monitors built after about 1998-1999 have some form of logic control built into them that verifies the VGA signal is of an acceptable nature. I've got a seven year old (1996 Mfg date) Hitachi 21" monitor that has this technology built into it.

If you select a resolution, a vertical sync rate or even a horizontal sync rate that is outside of the monitor's electrical specification, it blacks out the screen and says "Invalid Freq.". On my newer NEC 21" monitor, it says "Signal Error". And on the Viewsonic P810 21" monitors at my office, it says "Scan Freq too high" or something to that effect.

Even my el-cheapo Daewoo-remarked Seanix 15" monitor has a little feature like that. It gives you something like "Freq out of range".

So to tell me that you bought a $300 monitor that blew up because of a refresh rate issue, you must mean that you bought a 15" monitor from about six years ago from Packard Bell before they went belly-up. Because any current or even recent $300 monitor will have the auto-detection logic in it just for these such occasions.

I have a very hard time believing any modern monitor would die from even a forced refresh rate. The only thing I could possibly see is somehow sending a refresh rate of like 1000hz at like 5v signal strength which would burn out the analog inputs. I don't think any of these video cards has the capacity to send a 5v signal, and certainly not at a refresh rate faster than the RAMDAC onboard.

 

[MassiveOverkill]

It doesn't have to be too high of a refresh rate that kills your monitor. It could be an offset refresh rate that does it as well. So if a normal refresh rate was say 70, 72, or 85 Mhz and the new drivers forced a 70.57, 72.57, or 85.57 refresh rate, that could kill your monitor.

SilverSinkSam, alot of peeps have their emails and other info in their bio's. That can be matched up quite easily.

I'm all for collecting PERTINENT facts. You have yet to show that overclocking affects refresh rates. If you can show overclocking affects refresh rates then it would be pertinent information, but it's not. You might as well ask people whether they're gay/straight, if they're a virgin, male/female, and how many pets they own. Where's that 'beating a dead horse' smiley when you need it.

Albuquerque, my monitor was built in mid 2000. If you think I'm lying about the type of monitor I have and the fact that it actually went out............... I see SilverSinkSam lives in my neck of the woods. If he lives on the East Coast like I do, he's more than welcome to come check out my new 50 pound paperweight to satisfy yours and DragonPrince's skepticisms.

I don't know what's worse, losing my gaming monitor, or being accused of purposely trying to create a scam. This is the second time I've been accused of it.........the first time is when I first joined this board. I was more than happy to furbish all my personal information to the mods that time to prove my innocence.

I'm willing to call all you BSers out that don't know what you're talking about. Put your money where your mouth is. I'm currently using a MagView 17" monitor (also built in 2000) that I swapped out after my Sun monitor died. I was experienceing the same symptoms with it until I went back to 3.7 Cats. Luckily I switched in time as it's still alive. I'm willing to have someone in my area act as a witness (forum member obviously, one that both sides can trust) and I'll reinstall the 3.8 cats and then they can observe this monitor eventually die (I'd give it 3 weeks at most). The monitor can stay at this person's house for the entire test. If it does die, then all you naysayers get to foot the bill for this monitor and the Sun I just lost. You can prepay SilverSinkSam and if my monitor doesn't die within that time, he'll send you back your money, if it does, then he forwards the money to me. If it doesn't die, I'll gladly come here and post and apologize to everyone for being paranoid.

So there you have it, put up or shut up.

 

[NCtramp]

Artifacting is usually the first sign of an overclock gone bad...you OC'd ?

NO O/C EVER!


Read my posting again !!

 

[Michfan]

This problem is very well documented over at the Rage3d forums with the same flaming going on in some aspects. I find it odd that after a driver release that all of a sudden people are having monitors dying on them. I too have noticed weird monitor flickering since installing the 3.8's. Games that run at the same resoultiona and refresh rate as my desktop used to never have the monitor reset itself but now it does. Some have conjectured that the secondary adapter is being forced in some modes as opposed to the primary adapter that should be used causing the monitors to display out of whack refresh rates. Catalyst Maker stated that they have found no reason to believe that the 3.8's cause more heat than the 3.7's do in their own testing. Over at the driverheaven.net forums theres a test setup one guy used using a thermal probe and the different drivers and found out that them temperature increases were mainly from the leaked Dell and Asus drivers. The Omega 3.8.5's are based on the Dell leaked drivers. In his testing the 3.8's actually ran cooler than the 3.7's. I have no doubt something very weird is going on but no one can really pinpoint it at this time. I find it TOO coincidental that after a driver release that so many people's monitors or cards have died.

 

[kct2]

You have yet to show that overclocking affects refresh rates.

You have not shown that running the card out of spec does NOT affect refresh rates. You can't put the burden of proof on others, you are the one making the claim that it absolutely DOESN'T, other people are just saying it COULD have in impact, and the only way to determine that impact is to collect ALL the information, seemingly important or not.

There should be a survey - but is should collect all information, including:
windows version
windows service packs installed
directx version
PREVIOUS directx version (whether you went from 8.1 to 9.0, or 8.1 to 9.0b, etc)
current video driver version (omega or ATI)
previous video driver version
other installed components and their driver version
basic system config (motherboard, RAM, etc)
installed tweaking/overclocking/general system utilities/software
and card overclock/mod status.

I'm sure there are things I missed...

The key is to find something common to all people who had lost monitors, if in the end the only commonality is using the 3.8 driver, then that is telling, but if there is another common factor, then that has to be investigated.

Also, you had know there was a risk when you overclocked your card, or put a barton heatsink on it. As soon as you did those things you had to know you were operating outside ATIs specifications and, therefore, warranty protection. You have to know that ATI cannot test every card in every out of spec condition, so they don't know if running a card in that state will cause a problem or not and neither do you.

 

[Oklahoma Wolf]

nearly all monitors built after about 1998-1999 have some form of logic control built into them that verifies the VGA signal is of an acceptable nature. I've got a seven year old (1996 Mfg date) Hitachi 21" monitor that has this technology built into it.

I have a very hard time believing any modern monitor would die from even a forced refresh rate. The only thing I could possibly see is somehow sending a refresh rate of like 1000hz at like 5v signal strength which would burn out the analog inputs. I don't think any of these video cards has the capacity to send a 5v signal, and certainly not at a refresh rate faster than the RAMDAC onboard.

I'm an experienced electronics tech... actually this type of protection has been built into most monitors ever since the early multisync days in the mid 90's. However, sometimes these circuits (because of poor or simplistic design) can't act fast enough to protect against several non-supported rapid refresh/res changes, which is what I think we are seeing here (it's how fast the display cycles between resolutions and refresh rates, not necessarily the res/refresh itself). Very expensive monitors can and do fail this way - I've seen it happen. Fixing them ain't no party either... I usually turn down these jobs unless it's my monitor

 

[MassiveOverkill]

You have not shown that running the card out of spec does NOT affect refresh rates.

It doesn't. I don't have to prove it. If there was a danger of overclocking affecting your refresh rates, video card manufacturers would publish warnings about it. You would see documentation about it in any of the hundreds and hundreds of forums. Like I said, find me ONE, just ONE credible documented instance that overclocking affects refresh rates. It doesn't. Like I said before, it's about as relevant as the owner of the card being straight/gay, or male/female.

As for my video card, I could care less about the warranty on it........I'm not talking about my video card, I'm talking about my monitor. My monitor costs twice as much as my video card. The problem has already been made clear about the refresh rate. More than one person has stated that the do no overclocking or modifications to their card whatsoever, so right there is your proof that overclocking isn't what's causing it.

 

[MassiveOverkill]

I'm an experienced electronics tech... actually this type of protection has been built into most monitors ever since the early multisync days in the mid 90's. However, sometimes these circuits (because of poor or simplistic design) can't act fast enough to protect against several non-supported rapid refresh/res changes, which is what I think we are seeing here (it's how fast the display cycles between resolutions and refresh rates, not necessarily the res/refresh itself). Very expensive monitors can and do fail this way - I've seen it happen. Fixing them ain't no party either... I usually turn down these jobs unless it's my monitor

Thank you!!!

 

[kct2]

I'm not talking about my video card, I'm talking about my monitor.

And what is your monitor connected to?

What are you blaming the dead monitor on?

I'd say the answer to both is...VIDEO CARD

So you are really talking about the video card.

More than one person has stated that the do no overclocking or modifications to their card whatsoever, so right there is your proof that overclocking isn't what's causing it.

You can't deny that there are many people using the 3.8 driver without any problem at all. So if your one or two instances of someone using an unmodded card and a damaged monitor proves overclocking/modding is not the cause, then the vast majority of people using the driver with no incident obviously proves that it is not the drivers fault...right? same logic

If you are so sure that overclocking has no relation to this problem, then the instances of monitor failure will be distributed between overclocked and non-overclocked cards, so it will be easily ruled out as the cause. Why not just collect the info?

Perhaps it is the utility you used to accomplish that overclock. Maybe it's an old version of some utility that didn't get fully uninstalled...

I know why you want to supress this information, you expect some restitution from ATI. Even though you already said you expect no warranty from ATI. If ATI is not responsible for the card, they also aren't responsible for any damage it may have caused, even if it was entirely their fault. The warranty agreement is a legal contract, one that you chose to remove yourself from, therefore ATI is not bound by that in any way. This is not true for the people who have not modded their cards and still are under ATI's warranty coverage (of course that's only if they purchased a BBA retail card.) If you bought a retail card from a board partner (i.e. sapphire) then you need to take your claim up with them. If you have an OEM card you purchased it without a warranty - part of the reason for the lower price - and have no one to blame but yourself.

Very expensive monitors can and do fail this way - I've seen it happen.

Oh - wait - so this is something that happens, perhaps regularly? So it has happened before the 3.8 driver, and it will continue to happen after it.

Why don't you call your monitor manufacturer and complain to them for a poorly implemented protection circuit?

 

[Albuquerque]

So what was this whole "put up or shut up" thing about? I'm running Cat 3.8's, I'm running far-overclocked speeds. I'm running a Pro bios. I'm running a 21" NEC monitor.

My monitor works. My monitor doesn't tick in and out of resolutions while gaming. My machine has been using 3.8's since they came out. I game on it every day. I use refresh rates in excess of 100hz.

Where exactly is MY problem?

 

[harryinny3]

WELL, For the image quality gained and the better FPS i have gained with the 3.8 drivers, And only 7 degreesF increase in temp. I will stay with the 3.8


Harry

 

[micamica1217]

kct2,

just a thought.....

if you OC your cpu (FSB), your warranty on the vid card may be null and void as well.

mica

 

[kct2]

Well, if it overclocks the AGP bus then yes, that is probably true. I'm not sure what your point is.

 

[MassiveOverkill]

kct2, you're obviously employed by or affiliated with ATI, because your diligent defense of them goes above and beyond normal ATI fan loyalty. I refuse to talk to YOU any further on the issue. Tell your employers to ready yourselves for a class action lawsuit.

Albuquerque. I'm really glad that you're not having problems with your monitor. You're saying there isn't a problem with the new drivers. I'm saying "Yes there is, and I'm willing to prove it" by taking a perfectly good working monitor and subjecting it to cat 3.8 drivers on an UNMODDED, UNOVERCLOCKED STOCK card, and watching it die from the incompatible refresh rate problem.

 

[MassiveOverkill]

I just called ATI at the tune of $1.25/hour and guess what, Chris Murdock, service rep states that "They haven't heard any problems relating to this issue". He further stated that ATI would not compensate anyone for monitor damage.

I also asked Mr. Murdock if he's familiar with website and corresponding forums on Rage3D, which he had no problem identifying. He still had no clue about the problem I was talking about. I had to ask two or three times to be referred to a supervisor. He referred me to Steve Botelho, who is his supervisor, but he's conveniently unavailable (this is at 2:00 PM EST, so lunch should be over). For all I know these names could be made up.

I'm not going to argue anymore with those of you who have blinders on or who are being fed by ATI. SilverSinkSam, I'm not going to argue over creating a thread where we can gather info about peeps who's monitors died, I'll find another venue. I mean no disrespect to you in particular.

 

[Oklahoma Wolf]

Oh - wait - so this is something that happens, perhaps regularly? So it has happened before the 3.8 driver, and it will continue to happen after it.

Why don't you call your monitor manufacturer and complain to them for a poorly implemented protection circuit?

I should clarify... when I do see this happen, it's usually from other causes; mainly time causing various components to go out of spec. I've never seen an instance until today where a monitor has died from drivers forcing several out of range settings within a matter of seconds, as it sounds like is happening here.

The simple fact is, monitors are designed in anticipation of having these settings changed often. The case that some hardware driver might try to do this outside sync range several times in a few seconds wasn't always planned for. It's not poorly implemented, just unanticipated.

 

[DaBigJ]

I just called ATI at the tune of $1.25/hour and guess what, Chris Murdock, service rep states that "They haven't heard any problems relating to this issue". He further stated that ATI would not compensate anyone for monitor damage.

I also asked Mr. Murdock if he's familiar with website and corresponding forums on Rage3D, which he had no problem identifying. He still had no clue about the problem I was talking about. I had to ask two or three times to be referred to a supervisor. He referred me to Steve Botelho, who is his supervisor, but he's conveniently unavailable (this is at 2:00 PM EST, so lunch should be over). For all I know these names could be made up.

I'm not going to argue anymore with those of you who have blinders on or who are being fed by ATI. SilverSinkSam, I'm not going to argue over creating a thread where we can gather info about peeps who's monitors died, I'll find another venue. I mean no disrespect to you in particular.

ATI can hardly admit even to knowledge of this problem, because it suggests that ATI's drivers are responsible for killing many expensive video cards AND expensive monitors. If they even so much as confirm this problem, they would be admitting responsibility for people's equipment.

 

[HeXenViRii]

This just plain makes me mad , err , add my toasted 9500->9700 to the list