Grey screen after cleaning GTX260M: ESD damage?

Hi,

this is my first post in this forum, so please bear with me if I make any mistakes.

I have the following problem:

When I disassembled my Notebook (Alienware M17x-R1) today for cleaning, I also removed the graphics card (single GTX260M) to dust off the heat sink. After reassembly, my machine would only show a uniform gray screen.

I have now exchanged the card for an identical one and everything is working again.

Is it possible that I destroyed the card with an ESD?
How does the output of an ESD damaged graphics card usually look like?

And-should I try to "bake" the card, like described here: http://forum.notebookreview.com/alienware/385973-how-repair-your-dead-graphics-card-your-alienware.html or would it just be a waste of time?

I have already tried to find general information about the GSOD and its precise causes, but could only find solutions for a fairly recent driver problem on ATI cards and Apple systems.

Thank you!

 

First you should go back in and make sure everything is firmly connected, even things you didn't necessarily touch. Reseat the card again. Do a complete power drain by pulling your CMOS battery, the normal battery, and the A/C adapter, then hold the power button for 30 seconds.

As for the baking thing, it's an older technique, and I'd wait to do that only if necessary.

If nothing seems to work, there are a few on eBay. Buying one from Dell is also an alternative.

 

Please don't toss your video card or any other PCB into the oven, even if it does work. This is stupid and having an Engineer have success and advocate for this method is sad. If you're out of options and you really are considering the baking method, why not just get a hot air gun or find someone that has one and reflow the PCB. But in all honestly, if you're at this stage just get a replacement and not have to deal with whatever else may happen from the oven baking.

Again, DON'T use the oven baked procedure! If you want more info and the proper way of troubleshooting PCBs in detail I'll be more than glad to outline them for you.

 

thousands of people have had success across the entire globe.. it works. Are you aware of the solder problems that cause it in the first place?

did you read the OP? He has a replacement already, he wants to try to get the other one working.

I would try to install it again and double check the LCD cable this time. ESD damage is not repairable via baking, if it doesn't work the second time around PCB repair is probably out of the relm of possibility for you, and I would toss the card, or make myself a keychain.

 

Please, talking about oven baking a PCB gives me a headache already. If you really want to fix it do it right and not a half job such as oven baking the whole PCB. Sure if may have work for some users out there but it is not the proper or the best way to go about doing it.

If you know it is a solder problem and know where it is, then reflow just that part. Why do it to the whole PCB where non solder components are being heated as well. In addition, you don't know if the reflow is going to be ok afterwards. This is why IPC standards are in place. Research it, follow it.

I realize a user doing this is not the same as rules and procedures in a workplace but come one...use some common sense here. Just because a few got theirs to work SOMEHOW, without really knowing how, where or possibly why, you shouldn't just trust it. Hell, might as well toss some other stuff in there to see if they magically pop back to life.

Does anyone here know the details how a PCB is made to even consider doing this oven bake procedure. How many of the successfully users out there know what temperature lead/unlead solder is flowed at, for how long, etc.? How many of them know the tolerance of the PCB in itself, traces, cold solders, etc.? Add to that all the different components on the PCB and whether or not they meet the same requirements (ROHoS, non, BGA, FPGA, etc.).

Sorry, this oven bake procedure is misguided no matter if there are those that got their parts working. It's like testing rockets and hoping a few would work. Know what you're doing, try to do it right and don't be a sheep. There are proper ways today to fix PCB problems.

 

okay, you apparently are not familiar with the subject. reasearch G80 and G92 nvidia core solder problems. There's no reflowing on the PCB board, it's the chip die, and since most people lack a heat gun, the oven suffices.

I don't have to. Unfortunately, we live in a world where not every person with a computer carries a 300W heat gun and knows how to re solder a die the size of a half-dollar. I would advise you educate yourself on a topic before trying to explain it's fallacy.

misguided?

Please.

Again, go goggle the actual problem. I would link you one, but its so common an issue you won't have a problem finding a good source.

 

You don't have to? How arrogant is that. Look, if it's the core that's having solder problems how does one know that baking it in an oven will fix it? Because a few have had it worked? Six Sigma and similar tools were created to avoid ideas and behavior such as this. IPC standards are in place to help, not to obstruct the right way to fix PCBs. The thousands you've mentioned means little when it comes to avoiding such quality defects and inefficiencies.

Can you honestly tell me that by oven baking you can tell that it will in fact come back to life, that the solder has been re-flowed properly and that other components won't be damaged? The fallacy here is I'm trying to advocate the right ways of addressing a defective PCB and have agreed and suggested that the user tried the cables, re-insertion and if it comes to this oven baked idea to avoid it and get a replacement. You may not have a heat gun but when it's at this point the best thing to do is get a replacement/help.

Unless you or any other Engineer can support the oven bake with data I say stick to what is already out there for fixing PCBs.

Honestly, oven baking PCBs. I thought this was done and over with in the 80's, what with all the jokes made from it. In any case, take it as you will, I'm done.

 

you're either not capable or unwilling to familiarize yourself with the specific problem, so I can't see continuing the topic further with you until you do so.

 

Do you seriously think people haven't tried every alternative before the oven-bake? Even the post by the engineer guy stated that it was his FINAL option as his warranty was out and he has tried everything else.

 

I've used a heat gun and tried the oven method for a dead 7950m GTX. The heat gun worked temporarily and then the card died again. I later put it in the oven and let it bake and after that the card never died. Maybe there were other chips on the PCB that had bad solder joints that I wasn't aware of and the oven trick did the job. It could potentially damage other components but you know what? Considering the video card is already a paper weight and if you don't have the proper diagnostic tools to figure out what the problem component is (read: majority of people out there) then the oven trick is the best route to take.

 

I've no idea why.. however for some reason I think he's under the assumption that I'm advocating the OP use the baking method. Considering he missed half the OP the first time this is not surprising.

 

Lozz, what is your problem with me telling people to use the proper ways, tools and such? I don't mean to be a jackass here but seriously, your attitude towards my offering professional advice in addressing a defective PCB is not well taken.

Don't make this out like "fine believe what you want" as though what I have talked about is nonsense. Years of hard work by many in the various professions, including myself, have given us the tools today to properly address defects such as PCBs. To discredit not only me but the tools we have today is disrespectful. If you don't agree with what I have presented please take your time and provide the details as to why you don't agree. I'll listen if there's evidence to show that oven baking PCBs does in fact bring them back to life.

And you're right, what I say isn't going to discourage all from using the method in their own right. That's expected. But at least I have provided some sound advice as to why this shouldn't be done and that there are proper ways to address this.

 

I don't have a problem with PCB diagnosing, what I do have a problem with is trying to say baking is the incorrect approach when the problem has always been the material used on the nvidia die, when the method has consistently worked thousands of times, when you're attempting to suggest fallacy in a method of which you have no idea the original origin of the problem of. That is what I have a problem with. I have not suggested that baking the card at this time will work. In fact I discouraged it because I specifically said it would not work. I'm not even sure why you were attempting to say that I was in the first place.


When these specific cards suddenly die, in almost all cases it's because of the before mentioned defect, there are good reasons Nvidia spent millions on a recall of effected chips. Admittedly, there are exceptions, but there are always with anything like this. It's easy for you to suggest diagnosing, however I would say it's not so easy for the average person wouldn't you agree? Baking works-most of the time, it's just being argumentative to suggest there are better ways, because sure there are, but they're probably not within the reach of most people imo. I would consider myself pretty informed and even I don't think I could attempt a PCB repair myself, especially on something as compact as a MXM3.0b board.

 

Alright, I can see there are some misunderstandings here with the big guns coming on my tail. So let me address this issue properly.

Mardermodus asked if he should try the oven baked method in which TurbodTalon thoughtly provided his suggestion and rightly suggested that he do it only if necessary.

I came aboard and offered my professional advince for him NOT to use the oven baked method if he's gotten to this desperate point; that and there's proper ways of addressing solder problems on a PCB.

You, Lozz, on the other hand came in stating that there are thousands of others that have gotten this to work and that "it works" as a matter of fact. This I don't disagree but merely stated that users who have done this don't actually know or can 100% tell if the problem has been fix. Yes the PCB is back to operation... how, why, how long is all unknowned. That is not how things are done, and not done here with you guys.

You then insisted I lack the knowledge of the why the oven back method was used, of which I must refer back to solder issues. Please don't tell me I haven't read and understood the problem. Why else would anyone else heat the PCB board? Regardless of what the problems are in the thread that I've not read, the underlying problem is the solder. You tend proceed to insult me on what I have commented.

Corthalis/Joker, I'm offering my advice of not using the oven baked method. Why is this such a problem for you guys, especially when I've offerend sound and tested methods? It even amazes me more when I know you guys do all these testing on your laptops using data and yet you get upset at met offereing my advince based on standards. Seriously, if my advice is that bad explain it to me in detail. I have no problem with anyone using the oven baked method, but I'm not going to be a sheep and go along with you guys on it.

 

No. You are wrong. Refer to post 4. I said Baking, as in baking to re-flow solder to the core works. This is a Fact. What I then said, is that ESD damage is not repairable via baking, and I would advise that he did not use the method.

 

I can see you are going to prove your stance no matter what and I will indulge you on it for a while longer.

Regardless of what has been talked about when you are heating something up it is to correct solder and/or expand...in this case, everything that can be. There are no other common sense reasons other than this. Do I really need to read the entire discussion to come to this conclusion, no. Absolutely not. So what is it exactly that you still think I don't comprehend from not having read the anything other then deducing the problem from heating up a PCB? You keep pushing this in my face to make it seem as though I lack the knowledge and understanding of the whole subject.

I don't see anywhere where you've stated you discourage the method. Help me, I'm blind.

Again, regardless of my a-hole advices, TurbodTalons sound advice should be consider.

 

post number 4, can't possibly miss it. well, you can, but it's hard to.

 

You're arguing against the oven baked method and saying it doesn't follow standards which is likely correct. But on the other hand, there's no denying that hundreds of people have done this method be it for a video card or xbox 360 and gotten it to work. It was pointed out to you that having a heat gun on hand or other diagnostic tools is not feasible for a lot of people out there (an oven is in nearly every home). If the card is no longer under warranty and the person is left with no options, it's not a bad way to go. Now I'm only referring to bad solder joints. Other issues like a chip failure from static electricity is an entirely different topic.

 

Lozz, I'm not disagreeing with you. I'm merely stating that it may work and it may not. And even if it did one does not know if it's a good flow. This is why xrays and other tools are used to verify.

Wow, all this because I suggested not to use oven baking.

 

In these kinds of forums, I've whitnessed on more than one occasion where too much information is just as bad as incorrect information. Joker and I had a similar argument not too long ago. If you tell people that baking is a bad idea and is silly, drifters and some of the rest of the forum might take it too seriously as a suggestion, and discount the practice all together. Even though it works most of the time on those specific cards.

 

Sound advice and 100% agree. As to the back and forth bickering going on here... let me resolve this right now.

First, 'baking' - if it works on your first attempt, is going to be a temp solution. Go ask the M1730 crowd - some of them have pulled off 15+ bake jobs on their 8800M GTX SLI. You may ask why 15? Because its not a perm fix.

This happens on many topics...


In any event, this thread is now closed