This wasn't supposed to work...

[Th3_uN1Qu3]

Replaced a nvidia 8400 in a Packard Bell laptop i just got. Altho the chip moved no problem when i poked it, it took away three pads on the edges when i lifted it... All 3 with visible traces.

My first time lifting pads in quite a while. Normally, the board would have been scrap, but having no other choice, i decided to try and fix them. Notice the really tiny wire in the middle of each of the lifted pads.

Was quite the adventure to get wires soldered to the traces AND then tin them without them getting unsoldered again. But the end result? Soldered the replacement chip on and everything works fine.

[3d0]

Re: This wasn't supposed to work...

I don't see any lifted pads :P
Well done!

[dellxps15]

Re: This wasn't supposed to work...

u did really a great work!!!! what king of wire did u use for repair ? 0.1mm or normal awg30 ?

[Th3_uN1Qu3]

Re: This wasn't supposed to work...

I didn't have wire thin enough, so i cut a piece of fresh desoldering wick and pulled the wire from that. Took me about 1 hour to solder 3 tiny pieces of wire... But in the end, it was worth it.

And just in case you didn't believe me, here it is all back together. I like that, just like its 17" brother, this one has a little subwoofer too so it sounds nice. Actually what i did is i traded its big brother in working condition for this one with bad video, and some cash.

Now to take more pics and put it up for sale, as i don't have much use for it. The reason is right there at its left and it's called Lenovo R61i. That one deserves a topic of its own, sometime later.

[Agent24]

Re: This wasn't supposed to work...

Nice solder job. I can just imagine how much fun that was...

[trebo]

Re: This wasn't supposed to work...

Replaced a nvidia 8400 in a Packard Bell laptop i just got. Altho the chip moved no problem when i poked it, it took away three pads on the edges when i lifted it... All 3 with visible traces.

My first time lifting pads in quite a while. Normally, the board would have been scrap, but having no other choice, i decided to try and fix them. Notice the really tiny wire in the middle of each of the lifted pads.

Was quite the adventure to get wires soldered to the traces AND then tin them without them getting unsoldered again. But the end result? Soldered the replacement chip on and everything works fine.

great job, perseverance at its best, enjoyable ending

[Th3_uN1Qu3]

Re: This wasn't supposed to work...

Nice solder job. I can just imagine how much fun that was...

"Fun" is an understatement. When i saw what happened i was like #@!&, but then i thought i've got nothing to lose anyway, so i put a new tip on my iron and went ahead with it. And... it worked.

[dellxps15]

Re: This wasn't supposed to work...

wich bga station do you have? i got one and after one reball, vga give same error

[sorinac]

Re: This wasn't supposed to work...

if vga give same error after reball the it's most possible faulty

[Th3_uN1Qu3]

Re: This wasn't supposed to work...

wich bga station do you have? i got one and after one reball, vga give same error

Ghetto'd it together myself. Hot air station + stand + 41x41mm BGA nozzle + preheater = BGA rework tool. Wouldn't attempt removing big chips like those on desktop cards, but for laptops it works fine. I only use leaded solder too, i reball even chips i buy new. Lower soldering temperatures make it easier to work with higher success rates. Broke only one board in the past year.

I don't do too many reballs so investing in a proper BGA station would be pointless. I do 4-5 a month at most.

Agree with sorinac here: chip's faulty if it does the same after reballing. Nvidia chips dated 2006-2008 are to replace on sight with new one, for them a reball doesn't last more than a reflow, because the problem is with the solder bumps inside the chip, not the balls under the BGA.

[orientalsniper]

Re: This wasn't supposed to work...

What do you use for stenciling?

[Th3_uN1Qu3]

Re: This wasn't supposed to work...

I use direct heat stencils.

[orientalsniper]

Re: This wasn't supposed to work...

Sorry for hijacking, but got any link for the product you are using?

[Th3_uN1Qu3]

Re: This wasn't supposed to work...

What do you mean? Can't figure out which product you are asking about.

[orientalsniper]

Re: This wasn't supposed to work...

The direct heat stencil tool you are using.

[Th3_uN1Qu3]

Re: This wasn't supposed to work...

As in what do i use to reflow the balls? Same hot air station. The key to successful reballing using a hot air wand is to use as low a temperature as possible. For this i take a piece of leaded solder and increase the temperature until it liquefies with the wand at 5mm from the solder, and use that temperature, no higher. That way i'm certain i won't overheat the chip. Don't worry if it takes a couple minutes for all the balls to stick, it's better to be a little patient.

When there's a few stubborn ones, i put some flux over them and go with the wand one more time in that area, and they're all good.

For removing the stencil, i let it cool down completely then set the temp to 250C and the air on max, and blow on the stencil for 10-20 seconds. This activates the flux without melting the solder, so the stencil comes off easily. Some people prefer removing it while it's still warm, but if you do it too soon you damage the balls and have to do it all over again. It's safer this way.

[orientalsniper]

Re: This wasn't supposed to work...

I actually meant which stencil tool are you using, can you tell me your model or the page where you bought it from? Do you vouch for it? This is an example: http://www.zeppelincomputers.com/dat...stencils_4.jpg

[Th3_uN1Qu3]

Re: This wasn't supposed to work...

I actually use the same crap you posted in the picture. Cheap and nasty but it does the job. Getting the balls reflowed on the sides where the holders are can take a while tho.

Some people prefer taping the stencil to the chip with aluminum tape and reball it without using a holder/jig, you may look into that.

[xboxhaxorz]

Re: This wasn't supposed to work...

Replaced a nvidia 8400 in a Packard Bell laptop i just got. Altho the chip moved no problem when i poked it, it took away three pads on the edges when i lifted it... All 3 with visible traces.

My first time lifting pads in quite a while. Normally, the board would have been scrap, but having no other choice, i decided to try and fix them. Notice the really tiny wire in the middle of each of the lifted pads.

Was quite the adventure to get wires soldered to the traces AND then tin them without them getting unsoldered again. But the end result? Soldered the replacement chip on and everything works fine.

My vision isnt the greatest but there doesnt appear to be any lifted pads obviously because you repaired it duh

Trace repair is TIME CONSUMING, as you now realize and using the desoldering braid is a great way to repair them

When doing trace repair it is always best to use lead free solder for the repair so that when you use lead solder balls you dont take it high enough to melt the trace repair solder

The other option in trace repair is NOT repairing the trace lol and simply solder the trace to the solder ball on the chip which i tend to avoid and perform actual trace repair

So if you can do trace repair, which you have that would put you on a expert soldering level

[Th3_uN1Qu3]

Re: This wasn't supposed to work...

When doing trace repair it is always best to use lead free solder for the repair so that when you use lead solder balls you dont take it high enough to melt the trace repair solder

Now that's a good idea... Thanks. I used 60/40 solder for the repair as that's what i use on everything (well, the balls are 63/37 but apart from that...). Worked fine but that's to note down for the future.