I said i'll never try reballing...

[Th3_uN1Qu3]

Re: I said i'll never try reballing...

Thanks. It works quite well actually, with one little gripe: Only the largest suction cup works properly.

After my first FAIL (the DV9000, because i didn't have the proper nozzle for the hot air station, and i couldn't heat the whole chip uniformly), i can now proudly announce my first VICTORY!

HP DV5, ATi HD3450 (216-0707011) video chip. Board had been reballed before, but likely overheated or something, as the corners of the GPU were raised. It fired up from time to time with corrupted gfx, most of the time it just shut down by itself after 1 second.

I replaced the chip with another one, that i bought pre-balled from ebay. In fact, i got lucky, as the seller didn't mention that, so i also got stencils and solder balls, but that's what i got. Replacement chip seated itself perfectly, and IT'S ALIVE!!! I was kinda worried about the NB, as it too has raised corners, but it appears to work. We'll see once i fully assemble it and run some stress tests.

[xboxhaxorz]

Re: I said i'll never try reballing...

oh cool, enjoy
yes i figured that it should be sold alone too

[Th3_uN1Qu3]

Re: I said i'll never try reballing...

tweezers wont work, as you will probably lift pads

That's exactly why i got it. Btw, i got it off you, as it was the only one i found on ebay that didn't come bundled with all sorts of other (more expensive) crap i don't need right now, like a lot of stencils or a can of flux.

[xboxhaxorz]

Re: I said i'll never try reballing...

tweezers wont work, as you will probably lift pads
the vacuum tool provides usually just enough umph to lift

[Th3_uN1Qu3]

Re: I said i'll never try reballing...

The vacuum pen-like tool wasn't that expensive... and it can lift surprisingly large/heavy stuff for its size so i likely will be using it for more than BGAs. Definitely gonna come in handy.

[Agent24]

Re: I said i'll never try reballing...

EEVblog 437 (at 8:50 mark) shows Dave using some blue tack on the end of a screwdriver. The actual lifting at 11:54 mark.

Heh, I saw that too.

I just use tweezers. If it's a large part, grab it by a corner\edge. So far haven't needed anything else.

[retiredcaps]

Re: I said i'll never try reballing...

The lift tool is due to arrive tomorrow,

EEVblog 437 (at 8:50 mark) shows Dave using some blue tack on the end of a screwdriver. The actual lifting at 11:54 mark.

[xboxhaxorz]

Re: I said i'll never try reballing...

lol its called vacuum suction pen

[Th3_uN1Qu3]

Re: I said i'll never try reballing...

I'm back, i've been quiet as it's not the nVidia G86-730-A2 chip which arrived, but an earlier order. A BGA holder for reballing. The lift tool is due to arrive tomorrow, so i'll have something to play with until the chips arrive, i can lift the bad ones and clean the boards.

[xboxhaxorz]

Re: I said i'll never try reballing...

Thanks guys!
About desoldering wick... I use (I guess) it's the goot wick - I know it is bad... I'll try today to do a laptop motherboard, and I'll get to all of You soon...

Yesturday I found a website which is selling this:
http://uk.farnell.com/itw-chemtronic...-bga/dp/892087
Is it good for cleaning?

Thank You again!

Goot is horrible wick
Chemtronics
MG
Those work well

But i have personally not found a wick that works as well as this stuff
http://www.youtube.com/watch?v=3fceVikYw1o

Note: most of our videos show us using products we sell, so you may consider this advertising but the videos also show our techniques so while i do not do any direct advertising on forums our videos show products in use

[ravenns]

Re: I said i'll never try reballing...

Cool guys!

I use 2 probes one on the left and right side of the chip (if You are looking from above at the chip), sometimes up and down sides of the chip (if You are looking from above at the chip).

Like this ( * - marks the spot where the probes are):

-------*
///////////////////--------////////////////////
//..................//--------//...................//
//..................//--------//...................//
//......CHIP......//-OR---*//......CHIP.......// *
//..................//-------//...................//
//..................//-------//...................//
///////////////////-------///////////////////
-------*

My profile:
0. Kingbo around the chip
1. Preheat until 150-160/170*C
2. Hot air - to 260*C until reading on the thermometer is at 190*C
3. Hot air - to 300 - 330*C max then poking the chip untill ready
4. Lifting
5. Cooling the board on it's own until 150*C then turn on the fan
6. Cleaning the chip with ZD-99 knob set at 1-2 o'clock
7. Also, cleaning the pads on board with the same settings of ZD-99
8. Put solder balls on, and then hot air set to 260*C about 5min until all the balls are melted
9. Next, Kingbo flux both board and chip
10. Aligning the chip?!? ****this is tricky for me if anyone have any suggestion....****
11. Bottom heater until 150-160*C
12. Hot air set to 230-240*C until chip "sits" about 60sec more after chip sits
13. Cooling on it's own until 150*C, then fan

This is all, if anyone have anything to add, or correct me, please do so...

Thanks in advance!

[Agent24]

Re: I said i'll never try reballing...

Without something to spread out the airflow in the nozzle you will have big problems - by the time you melt the balls on the outer edges of the chip the center will be overheating. You could test with a probe in the center and one by the edge. Or use an IR thermometer...

I don't think tape will help. Isn't that the tape used to protect surrounding components from heat? It's just going to block heat from reaching the IC properly, I suspect.

I saw a good thread on another forum about this same issue. I'll try to find it again, it was quite interesting.

[Th3_uN1Qu3]

Re: I said i'll never try reballing...

I use 2 probes on opposite sides of the chip. I've never used probes ON chip.

Are both of your probes on the top? If so, try moving one to the bottom of the board, and do your best to keep both temperatures equal.

Also, I used to cover chip with capton tape, but all it did is that the SMT components were lifted during heating... (I'm talking about SMT ON the chip)

There's no reason to do that.

I will try to use aluminum foil tape to cover whole chip in order not to cause crack of the chip (mirrored square). Sometimes I get little balls of solder over chip, which I guess is from second leyer of the chip (beneath the chip itself). I'm talking about solder balls which are inside of the chip. This happen rarly, and it happened in my first attempts with high temperatures...

It means you got the chip too hot, or you heated it up too fast.

Now, I'm affraid that using of hot air station will cause breakage of the chip? I use ZD-939 hot air station with BGA nozzle which doesn't have any net for proper disipation of hot air, but it has (I guess) zones - center zone which is the hottest, and the other zone away from the center which is less hot... And now I'm a little unsecure about this center zone will overheat the center of the chip. I hope You understand what I'm trying to say...
Does anyone use hot air station for reballing?

You can try with a temperature probe and see if you got any hotspots. All i can say is that i'm using exactly the same station and so far so good. I've done many reflows in the past couple years and had only one chip die, and that's the nVidia 8400 in the DV9000 i bought a few weeks ago, which i posted about in this thread.

I've used exactly the same technique and temperatures with a couple other chips since. I've done a nForce NB in an Acer 5520G and it came back to life. Also an Intel NB/IGP in a Fujitsu E8410, which behaves exactly the same after reflowing (doesn't work unless i press slightly on the NB, in which case it works just fine), but the Fujitsu had obvious water damage, some of it right on the NB heatsink, so the NB is most likely corroded and requires replacement. I've ordered it already.

So with my DV9000, i think it was just a dud GPU. And i think that's what's in the mail today... I'll go pick it up and come back with news.

Do You think it's a good idea to cover the whole chip with aluminum foil tape?

No. As a matter of fact i hardly use it at all, only when there's connectors close to the chip i'm reflowing, so i don't melt them.

[ravenns]

Re: I said i'll never try reballing...

I use 2 probes on opposite sides of the chip. I've never used probes ON chip.

Also, I used to cover chip with capton tape, but all it did is that the SMT components were lifted during heating... (I'm talking about SMT ON the chip)

I will try to use aluminum foil tape to cover whole chip in order not to cause crack of the chip (mirrored square). Sometimes I get little balls of solder over chip, which I guess is from second leyer of the chip (beneath the chip itself). I'm talking about solder balls which are inside of the chip. This happen rarly, and it happened in my first attempts with high temperatures...

Now, I'm affraid that using of hot air station will cause breakage of the chip? I use ZD-939 hot air station with BGA nozzle which doesn't have any net for proper disipation of hot air, but it has (I guess) zones - center zone which is the hottest, and the other zone away from the center which is less hot... And now I'm a little unsecure about this center zone will overheat the center of the chip. I hope You understand what I'm trying to say...
Does anyone use hot air station for reballing?

Do You think it's a good idea to cover the whole chip with aluminum foil tape?

[Th3_uN1Qu3]

Re: I said i'll never try reballing...

Thanks guys!
Very helpful!
Regarding my hommade machine... I use 2 temperature probes, and get even heat during preheat. When I apply above heating with my hot air station with a nozzle on it, I get slightly different temperatures on those two probes (~5*C). A few days ago I thought I should buy 2 more instruments, so I can cover all four spots around the chip.

I don't think you need 2 more probes. I use just 2 myself, one on the bottom under the core, and one on the top, on the side of the chip. I've lifted a bunch of chips cleanly with this arrangement. What will help tho, is getting one of those thin thermocouples that go slightly under the chip. Omega thermocouples i think they're called. I have another system in my mind, but i'll keep my mouth shut until i've tried it myself.

A few months ago I bought a Kingbo flux, and couldn't beleive how easy it is to work with it. Before Kingbo, I used Chinese Amtec (fake), and all it did is damage...

I have that, and it seems to work alright for me. You just have to remember to clean it afterwards if you used too much, otherwise it can mess up high-speed signals.

Tryed liquid flux from Check republic or Poland (don't know exactly) it did some good lifting, but Kingbo is THE KING!

It's from Poland i think, i have some of that too. I tend to use it for other things tho, like SMD rework. It's got too much isopropyl alcohol in it, and bubbles a lot. The advantage is that it's (supposedly) no-clean. The disadvantage is that it leaves this sticky residue all over and you have to clean it anyway.

[Agent24]

Re: I said i'll never try reballing...

Yesturday I found a website which is selling this:
http://uk.farnell.com/itw-chemtronic...-bga/dp/892087
Is it good for cleaning?

I've not used it but I expect it should be, Farnell are a good company. And it says it's for BGA, too.

[ravenns]

Re: I said i'll never try reballing...

Thanks guys!
Very helpful!
Regarding my hommade machine... I use 2 temperature probes, and get even heat during preheat. When I apply above heating with my hot air station with a nozzle on it, I get slightly different temperatures on those two probes (~5*C). A few days ago I thought I should buy 2 more instruments, so I can cover all four spots around the chip.
A few months ago I bought a Kingbo flux, and couldn't beleive how easy it is to work with it. Before Kingbo, I used Chinese Amtec (fake), and all it did is damage... Tryed liquid flux from Check republic or Poland (don't know exactly) it did some good lifting, but Kingbo is THE KING!

About desoldering wick... I use (I guess) it's the goot wick - I know it is bad... I'll try today to do a laptop motherboard, and I'll get to all of You soon...

Yesturday I found a website which is selling this:
http://uk.farnell.com/itw-chemtronic...-bga/dp/892087
Is it good for cleaning?

Thank You again!

[xboxhaxorz]

Re: I said i'll never try reballing...

I had a few HP Pavilions (GPU) and playstations come in with the same stupid ball grid array issues, but I'm reluctant to get involved in the process... I mean if they come from the factory and this happens to them after a couple of years what chance do you have with trying to work something at home...

It would be impossible to give any sort of guarantee when, in my opinion, the devices have a congenital defect that you cannot rectify... well, aside from trying to crap in extra heatsinks or fans to stop the chip from detaching itself..

Having said that, I did a reflow of a PS3 for a friend with a 15 euro heatgun, using Gilksy's tutorials on youtube.... And I was more surprised than anybody else when it worked..

While there are no guarantees there is increased longevity, the main problem is the solder combined with poor cooling

HPs are cheap thus they have the most issues
Panasonics are hella expensive and rarely have issues plus all chips are bga style since they dont want a chip falling out of a socket while its falling from an airplane yet they suffer very few bga issues

with a PROPER rework process and changing the solder from lead free to lead you can in fact have a long lasting machine

combine that with copper shims and mx4 compound you can get at least a few yrs

keep in mind that intel has announced not only gpu but cpu will now be bga
also the ps3 rsx is actually a pbga meaning it has bga on top of bga, the cell chip is still plain bga

[xboxhaxorz]

Re: I said i'll never try reballing...

The chip must be heated 4 times: to remove it, to clean it, to reball it, and to reflow it. Whether you are heating too much or for too long at each step is a different question.


If you are lifting pads, it may be that you did not heat enough when you removed the chip and it loosened them enough that when you started wicking you finished them off. It could also be that you overheated them instead.

With a DIY system this is quite possible, if it is not calibrated properly, or you are getting uneven heating.

Make sure you your temperatures are accurate and uniform as possible and you are following the proper reflow profile.


When cleaning the pads, you shouldn't need to "scrub" the pads with the wick. Just lay it down, apply heat, then remove. Move to the next area and repeat. Once you get the hang of it you might be able to start moving it around, but I think it's risky.

Overheating and physical stress are the enemy of all pads and traces! Make sure you have the area clean (lots of paste flux), a nice big clean wick, and a good iron.

I don't think you really need the solder bulb at all.

correct info ^^

you can help yourself by cleaning the board right after its removed or simply leave it preheated and start cleaning

this is useful if your iron is not suitable, but you should cover the area around the site since the components will be liquid

[xboxhaxorz]

Re: I said i'll never try reballing...

I am a newbee reballer too, I need an advice about cleaning pads on the board, and about aligning the chip in place after reballing. I get all of the 2 steps right, but in the end I get non functional Mobo or graphic card.
Firstly I built my own reballing machine which is consisted of hot air station above, and a grill below. I managed to do the clean lift, but I don't know if the chip is "healty" after all of those heating.
I usualy do like this:
1. Preheating below to a 150*C, then I apply above heat to 230*C I pick up the chip (this is the first heating of the chip)
2. Cleaning the chip (the second heating of the chip)
3. Heating the chip to melt the balls (third heating of the chip)
4. Aligning the chip with the markings on the board and heating (fourth heating of the chip)
So is it possible that in any of those heating steps the chip is burned?

Also, my problem is when cleaning the pads on the board (I use solder bulb and gently melting residues of solder with that bulb), and afterward use copper wick, and no matter how gently I go with the wick, I always rip off some of the pads on the board, but on the chip I manage to do a nice and clean and shiny pads.
I've seen many videos on YT about cleaning the pads on board. Some use cotton swabs while still hot, some use a piece of silicon to scrub the exces solder.
Can I use just the bulb method instead, and leave the wick?
Also, when using the wick I scrub the insulating warnish between two pads...
Also about aligning the chip?
Any idea?

Lifting pads happens to everybody, HOWEVER it rarely happens to us now and that comes down to a few things

operator error
soldering iron has low wattage
not using the best wick
not using enough flux

cleaning a pcb is alot harder then a chip due to the pcb being more fragile

if you do LIFT pads and they have no traces its worth a shot to put the chip back on as they may be ground pads

the video at the top is the best way to clean sites
http://******/15XorQA

you have to be gentle, if you are using pressure to clean then you need a better iron or turn up the heat higher

most components do have a limit of heat cycles and while it may still work the life is definitely reduced

we do the same amount of heat cycles as you and there is really no way around it plus we rarely have dead chips

when cleaning the chip the heat is usually quick, and when melting the balls you dont go that hot since your using lead balls so its not as bad, finally when placing the chip you would not need to go as hot again due to the lead