Might have fried my 6800 GT

[kryptobs2000]

I tried the vmem mod, soldered one wire to it, looked like it was good to me, didn't put the other wire, or the pot on just the single wire. Nothing else was connected, the resistor that it was soldered to did not looked burned or anything. So I put it back in my comp, to go check that it was the right resistor before I connected the pot and did something stupid. And at first it wouldn't boot up, made some error beeps. So I just yank the wire off. Still the error beeps. So I take my recently soldered caps off, even tho they were connected perfectly fine, I had the same problem before when the caps were connected yet one of the leads had been broken, so I thought maybe it was a similar issue.

So... took those off, comp. booted up, no error beeps. But now I got blocks all over my screen . I check the vddr voltage and I get no reading. Think this might be fixable?

[kryptobs2000]

Got a pic of it

[stealthbomber]

Huh?

You are suppose to solder the wire to the pot and then ground wire, you may have shorted the board if the soldered wire made contact with the cards PCB, or anything conductive like the metal case...

[kryptobs2000]

no, wire goes to one side of the resistor, then to the pot, then the other wire (which I hadn't soldered on yet) goes to the other side of the resistor onto the pot. The wire just stuck up in the air, it didn't touch anything and I rapped the end in tape, so I know that end did not touch anything. it might have been touching the pcb, but the pcb is not conductive, why would this have caused a problem?

here's the mod: http://www.vr-zone.com.sg/?i=1622&s=1

edit: The only thing that is conductive that is in the area, is that point to the upper left of the resistor, which I am 99% sure it was not touching, I looked at it very carefully and it didn't look like it was even close to it, or if I maybe burnt a hole in the pcb and it was connected to something, but the pcb looks pretty much untouched. Also, could be possible that I just connected both sides of the resistor, but I don't think I did that either. Does this look repairable if any of those 3 things happened or am I 6800 GT-less?

[OnLine]

What does that resistor on the board read now, perhaps the resistor got burned thanks to the soldering and perhaps it can be replaced ? It should read 43.2 kohm [ for most GT's ].

Edit: LOL , I mean it should read ~43 Kohm, my multimeter is not the absolute truth.

[kryptobs2000]

oh, meant to say that earlier to see what it's supposed to read, mine (sadly) does read 43k not gaining much confidence lol. I think most likely, it probably was touching that metal circle (w/e it is, I'm not sure) to the upper left. So I'm guessing this would cause a short, any way to remedy this? And what exactly is a short? I just know like... if you send a million volts to something thats meant to take 5, it will get overpowered and die. But what physically happens? It gets burned or the circuit breaks from the load etc. If thats the case, I'd think it's be impossible to fix. If just some mosfet fried or something tho, I could fix that.

[matt9669]

Generally the PCB is connected to ground at numerous points, it's possible that if you exposed the copper layer underneath and something touched it that it caused a short. "Blocks on your screen" sounds like corrupt RAM, i.e. your RAM chips aren't reading back the values written to them correctly, or aren't writing the correct bits etc.


A "short circuit" just means you short-cutted the intended circuit path, and if that shortcut has less resistance than the original path (generally a sc is to ground, meaning really low resistance) then you get a current surge. If the card still boots into Windows I'd highly doubt you actually "fried" anything, you may have just made the RAM's operation unstable i.e. some transistors are no longer switching according to spec.

But, I'm no computer engineer . . . hey wait a minute . . .

[reject]

have u tried rolling blutack over the area to see if theres any solder sparks on the card?

[kryptobs2000]

Thanks Matt for the great explanation. Right now at the voltage checkpoint for the memory reads 0, so maybe I broke a connection somewhere and the ram is not getting any voltage, or either the ram was fried with too much, or something in between and now gets 0. I do have solder on the smd caps you can see below, where I connected 2 caps. But I should beable to get that off.

have u tried rolling blutack over the area to see if theres any solder sparks on the card?

Not sure what blutack is, or what 'solder sparks' are either, could you please explain?

edit: thanks for the help guys

[kryptobs2000]

Sorry, got off topic lol, umm... here's what these smd caps look like. They don't look as bad as the picture makes it seem, cause it obviously is alot closer up than you can see with the human eye, but still, think I can make it look not so bad? Sorry for it being a little blurry



edit: added a better pic, it looks unblurry and all and here's my nf7-s vdimm just cause I'm happy I got it done so easy this time

[enzoR]

what value does the resistor read? i think your card is perfectly fine. take off that resistor and solder a 50k VR in its place. adjust the VR so that it reads 43Kohm that the guy above said it should be on 6800gt cards.

[kryptobs2000]

It does read 43k, and at the vddr check point, I get a 0.00 reading where as it used to be 2.05

[OnLine]

That spot the metal circle at the top left, which You accidently touched with the soldering iron, is most likely the same spot as the top of the small resistor.

Which means, You most likely haven't done a short [ in this area ].

I measured the resistance between those 2 points on my board and it say 0 ohm.

[kryptobs2000]

yeah, mine reads 0ohm, if that means anything. So what should I be trying to fix it. That small circle is the only thing I could have possibly touched and I am pretty sure I dind't even touch it, it is just the only thing it might have touched, so if that would cause it to short, then thats what I did. Otherwise I have no idea. Anything else that could have happened to make this not work? Maybe the resistor isn't in contact, but I really doubt that, it's pretty solid and on there. I don't really have any ideas.

[OnLine]

yeah, mine reads 0ohm, if that means anything. So what should I be trying to fix it. That small circle is the only thing I could have possibly touched and I am pretty sure I dind't even touch it, it is just the only thing it might have touched, so if that would cause it to short, then thats what I did. Otherwise I have no idea. Anything else that could have happened to make this not work? Maybe the resistor isn't in contact, but I really doubt that, it's pretty solid and on there. I don't really have any ideas.

Hm, not an easy case but if your card became dead after the soldering and You're 100% sure You didn't short anything else I would try to remove that resistor and solder on a 100K pot set to 43 Kohm.

Just to be 100% sure 43Kohm is in contact with the board.

[matt9669]

I'm not sure exactly why you're reading 0V for the RAM - that's confusing, as the board wouldn't be running at all were the RAM really recieving 0V . . .

As for the SMD caps, yeah there are some solder blobs there, but nothing you can't fix - just use your preferred desoldering method (one ghetto way to do this is get a can of air and blow the excess solder off once it gets hot).

have u tried rolling blutack over the area to see if theres any solder sparks on the card?

Blutack is a sticky material some peeps use to mount parts to boards (as opposed to say, hot glue) - I've never used the stuff myself. By "solder sparks" I believe reject means stray solder blobs that may have flown onto the board in the course of soldering. Generally solder "pops" like that don't stick (i.e. melt), but they might have found a home somewhere causing a short or the like

[kryptobs2000]

here's the vddr check point



as I said, I get a 0.00 reading. I'll check one more time in a last attempt.

And I got some desoldering braid somewhere around here, I could use that to get rid of the crap, and I'll sand it down... got some fine grit sandpaper around here too.... can't find anything though

[i found nemo]

"c550" on your mobo looks pretty screw'd up....

[matt9669]

Oh I trust you're reading the voltage correctly, wasn't implying you were an idiot - I just can't figure out how the circuit functions like that