Reballing discussion and tips

[Dannyx]

Re: Reballing discussion and tips

I was actually thinking something like a sandwich maker would be more akin to a bottom plate heater. Like I said, I had a proper heater plate at my old shop which the boys took away when we closed down, but I could probably get my hands on it again if I asked nicely I've got one of those K-probe thermometers you mention and a hot air gun from my soldering station, so I was thinking of combining all these things to make my own crude machine....maybe with an Arduino or a micro of some type

[momaka]

Re: Reballing discussion and tips

So you're saying sticky tape that molds tightly around everything is not recommended and I should just use foil which "sits" on top of the board, covering the components ?

Not even that. Don't cover anything. Again, only exception is electrolytic caps very near the chip. And even those, you don't cover - you just put up a sheet of foil loosely around their area (but do NOT cover the caps with the foil).

TBH, I also had the same feeling that COVERING stuff does more bad than good, but I dared not contradict the popular opinion AND my much older and supposedly more skilled coworker and god of all laptops, G

Indeed it does more bad than good.
Take a moment to think about thermal stress in a PCB: if you heat one are and not another, the area being heated will expand and the area not being heated will not expand as much and rely on getting heated from the heated area. This will create uneven thermal stress and could warp your board. You are much better off heating the whole board (with the area in which you will be doing BGA rework in the center of the heat, of course.) That being said, balancing top heat and bottom heat will also make your board warp a lot less. Proper support of the board is another must. On that note, do NOT use those steel or aluminum "jigs" with mounts on them. Instead, let the board sit "loose" on top of some rails that support it. This will allow the board to expand and contract with the heating and not warp because of being held with the jig.

That's something that'd be an interesting read indeed and would also serve as backup to any arguments that may arise if the aforementioned contradictions over reballing ever occur between me and G Sooner or later I'll have to take over this joint and do them myself...

Well, I actually found one item that may be potentially useful to you. It's a suggested reflow profile from a Kingbright AA3528VRVFS purple SMD LED (the one I used in my AN5272-based headphone amp.)



Of course, for different SMD chips of different sizes, you might have to vary those temperatures a bit.

But again, a first step would be to get one of those type-K thermometers like TM-902C if you don't have a reflow machine and want to practice on your own with a regular hot air wand (and perhaps something for bottom heat... like a toaster? )

Speaking of which...

Uniqu3 was using a freaking toaster!

Exactly! And he did quite a number of successful reflows too. So really it's just a matter of practice and knowing what you are doing.

[Dannyx]

Re: Reballing discussion and tips

I just put the foil to act as a "separation wall" between them and the chip, but nothing more (do NOT cover the caps in aluminum foil - that will only make them cook faster!)

So you're saying sticky tape that molds tightly around everything is not recommended and I should just use foil which "sits" on top of the board, covering the components ? TBH, I also had the same feeling that COVERING stuff does more bad than good, but I dared not contradict the popular opinion AND my much older and supposedly more skilled coworker and god of all laptops, G

I found a few awhile back and thought I had saved them... but I guess I didn't.

That's something that'd be an interesting read indeed and would also serve as backup to any arguments that may arise if the aforementioned contradictions over reballing ever occur between me and G Sooner or later I'll have to take over this joint and do them myself...

[momaka]

Re: Reballing discussion and tips

I also noticed it's standard practice to cover the area around the chip to be removed with some sort of foil. This guy uses aluminium tape but I've seen others doing it with foil. I suppose it's to protect the components around from the heat of the IR lamp or to keep them from moving. I always wondered what'd happen if you tried it without the foil entirely...

I actually do NOT recommend that method at all - not unless you want to warp your board even more.
I don't know who came up with the idea, but it was clearly someone that doesn't understand heat flow... much less BGA rework. It's almost like those video card oven bake reflows - they don't make much sense, but everyone that hasn't a clue how to fix anything does it.

The only time I ever use foil around components is if there are electrolytic capacitors very close to the chip - in which case, I just put the foil to act as a "separation wall" between them and the chip, but nothing more (do NOT cover the caps in aluminum foil - that will only make them cook faster!)

That being said, don't trust everything you see on YouTube, even if the person seems to show a "successfull" reflow. Instead, try to find some papers / documentation / technical guides on BGA rework that are written by actual companies and/or labs. These will often contain detailed temperature curves and techniques on what to do and what not to do. I found a few awhile back and thought I had saved them... but I guess I didn't.

[Dannyx]

Re: Reballing discussion and tips

That's the sort of stuff I like to hear

You dount mention the sizze and pitch of the grid array you're thinking of dealing with. Magnification is a boon regardless.

I don't know what that is. I mean I know what pitch is - it's the space between the balls - but what I don't know is what it is in practice...I dunno...what most GPUs use I suppose. Does it vary greatly or is it a more or less common one that you're mostly likely to use ? I know there are different sized balls, but I've only seen this guy use 1-2 sizes in general.

We'll refer to our laptop guy here as "G" from now on BTW for brevity's sake

We don't have a microscope around here sadly, so the best we have are those large clamp-on desk magnifiers with LED rings under the lens. G's seems to magnify just slightly more than mine, since mine was obviously ordered at a later date and I asked for one exactly like I had at my former shop. The ones these guys use have all had some issues with the ball joint at the head and they can no longer stand up straight - they're all droopy like and old dong so I avoided that model entirely. Anyway, waffle waffle...not the most ideal tool for working underneath, because you have to get it pretty close to the table to get it in focus and by that point it's a tad too low to be able to stick the soldering iron or meter probes in there, at least not easily...

Use the right stencil, make sure it's clean. Use quality solder paste and flux from someone like Amtech it will change the game. Lower air pressure helps the keep the balls from coagulating.

G's got a bunch of different stencils he keeps together in a box and selects the one he needs for a particular chip. I'm not sure HOW he determines at a glance which stencil fits what chip - it's probably easy. I haven't seen him lining them up onto the chip though. I've seen some sort of vice which clamps onto the chip and lines up the stencil for you, but either he doesn't use one or I missed that part, so that's another hole I need to fill in.
G uses Kingbo for flux and brushes it on, both on the chip itself to get the new balls to stick and on the board afterwards when putting it back. I haven't seen him applying it when REMOVING the chip though and I've heard that's indeed not mandatory.

Have patience and understand that there's gonna be failures to learn from.

I'm a patient man indeed and like tinkering, so I believe I'm fit for the job Mistakes and failures are always an option, too bad you can't afford too many of them

If you lift pads when you're removing the chip you're not doomed per say but you're basically doomed.

Do these tend to come off that easily ? It's funny that it takes me quite a lot of heat, pushing and prodding to get SMD stuff like capacitors and diodes off boards and the pads don't even budge so I can go hard (ok, not really), but these pads seem fragile...
Let's not forget the device is probably dead to begin with if I'm reballing it in the first place, so even if I do rip something it shouldn't be an issue - we just call it irreparable...

Lay down a puddle of flux place the chip and start with your air hovering up off until the flux begins to agitate/bubble/smoke then move in close on it, obviously don't bump it. Once the balls melts it will snap it self into position by itself.

We're talking about using a hot air station to solder the chip back ? G's BGA machine uses IR. He lays it onto the board and aligns it under the magnifier, presumably to get it as close as possible to the original position. I assume he uses the markings on the board which most boards seem to have to get it in place as closely as possible. He then turns on the lamp (and maybe the bottom plate as well) and lets it "bake" for a couple of minutes - not sure about the exact numbers here. I can see a graph on the display similar to the load graphs in "task manager" indicating the temps. The machine's got two K probes he sticks onto the board and close to the chip and I think it stops automatically. He most likely created some presets to help him eliminate any small human glitches. He seems to stop it when the flux is starting to smoke. I'm kinda expecting to hear a microwave bell here *ding* "enjoy your meal"
Do you need to press down on the chip or wiggle it in this phase ? I haven't seen him do it and I'm very skeptical about those balls adhering to the board just under the weigh of the chip itself.

[QUOTE=2andrewd;892324]
I'm assuming you've used hot air and have some experience using it to desolder/solder smd components. If not, start there./QUOTE]
Yes, I have worked with large and small SMDs in the past 4 years, which is not a lot but enough to get the basics down. Obviously the gear you're using also makes a huge difference and this is something that's been disappointing for some time now. My former shop couldn't afford to get the latest, meanest and state of the art tools, but what they had did the job and paid for themselves nonetheless, so naturally when I moved over to this new place which is immeasurably larger I was expecting they would have a whole different league of tools and equipment, but no....if anything it's actually worse as far as I'm concerned, so I'll sadly not get the chance to play around with something high end to get an idea of how much of a difference it makes, at least not for years to come now...
This is the machine we've got....I don't speak russian but it allowed me to have a look at the different screens and buttons on the thing.
Cheers and thanks

[2andrewd]

Re: Reballing discussion and tips

You dount mention the sizze and pitch of the grid array you're thinking of dealing with. Magnification is a boon regardless.

Use the right stencil, make sure it's clean. Use quality solder paste and flux from someone like Amtech it will change the game. Lower air pressure helps the keep the balls from coagulating. Have patience and understand that there's gonna be failures to learn from.

If you lift pads when you're removing the chip you're not doomed per say but you're basically doomed.

Putting it back together i.e soldering a BGA isn't that difficult compared to things like running wires underneath them (so don't lift any pads) IMO. Lay down a puddle of flux place the chip and start with your air hovering up off until the flux begins to agitate/bubble/smoke then move in close on it, obviously don't bump it. Once the balls melts it will snap it self into position by itself. It's not easy but it is possible if you're willing to put in the time.

I'm assuming you've used hot air and have some experience using it to desolder/solder smd components. If not, start there.

[Dannyx]

Re: Reballing discussion and tips

I also noticed it's standard practice to cover the area around the chip to be removed with some sort of foil. This guy uses aluminium tape but I've seen others doing it with foil. I suppose it's to protect the components around from the heat of the IR lamp or to keep them from moving. I always wondered what'd happen if you tried it without the foil entirely...

[dj_ricoh]

Re: Reballing discussion and tips

Uniqu3 was using a freaking toaster!
Bga work it's all about bottom heater then you help with top head just a little to reach from 150c to 220c.
Hot air station it's still ok.
I took the thermostat from an ir6000 cuz I don't know how to setup the machine and use a hot air station instead of the top heater that came with the machine lulz

[Dannyx]

Re: Reballing discussion and tips

Well it will be a long while until I'll get a chance to mess around with the BGA machine, unless maybe we get a second one AND I get proper permission of course, so I might as well try to build one. I almost started this project a while back, back when I was at my old shop where we had a heater plate normally intended for heating touchscreens for removal but which we rarely, if ever, used, so I was thinking of combining that with the hot air handpiece from our soldering station to create a crude and ghetto BGA machine so we could try our hand at it

[momaka]

Re: Reballing discussion and tips

I think the best way to practice would be on some old, yet functional boards that you're not too scared of breaking. That way you'll at least know if you did it right and the thing still powers on, as opposed to doing it on dead boards which didn't turn on to begin with. With no option to visually inspect the joints, this is the only feasible way that I see.

Well, yes and no.

For starters, you'll want to learn how to successfully remove chips and clean pads without ripping out traces (i.e. getting the temperature right and also learning which iron will work with your braid for pad cleaning.) So for that, you don't actually need a working board at all, because you will know if you damage/rip any traces.

But once you got that technique down, then it will be good idea to try your skills on an older working board to see if you can pull it off (i.e. remove a BGA chip, then put it back on with leaded solder and test the board.)

Only problem is that chip technology may be slightly different now than it was 10+ years ago when a "practice" board like this is likely to date from.

Not really.
BGA technology hasn't changed much in the last 10 years. If you go back 20 years, sure - you'll likely find that the BGA balls are much larger and more spaced apart, in addition to some boards using leaded solder that is easy to work with. But 10 years ago (~2009 as of writing this), that was well into the lead-free era. In fact, the lead-free BGA crap dates back as far back as 2005-2006 on some boards... so technically even some of those will work for practice (think: late Pentium 4 / early Core 2 Duo era boards).

[Dannyx]

Re: Reballing discussion and tips

so for example, I may give you some specific advice that you may find doesn't work well with you.

Not having done this before (only witnessed it), I welcome every bit of advice and tips others may have and out of all of them I'd put together and develop my own method once I actually DO get into it. It's one of those things that can't be universally agreed upon, except for maybe some generics which always stay the same.

Practice, practice, practice. The more times you do it, the better you will get - both in terms of your own skills and in terms of adjusting temperatures/profiles, and whatnot.

I think the best way to practice would be on some old, yet functional boards that you're not too scared of breaking. That way you'll at least know if you did it right and the thing still powers on, as opposed to doing it on dead boards which didn't turn on to begin with. With no option to visually inspect the joints, this is the only feasible way that I see. Only problem is that chip technology may be slightly different now than it was 10+ years ago when a "practice" board like this is likely to date from.

Consider placing your "machine" (whatever that setup may be ) in an area that you can ventilate easily

LOL, yeah, that's a no-brainer. Someone should tell that to this chap who does it in a small stuffy room with no air vents, all the windows closed, smoke everywhere...WTH, how can you live like that ?! Glad I don't work in the same building as him

[momaka]

Re: Reballing discussion and tips

No problems.
I think the reason it didn't get much attention is because reflowing/reballing is not such a "big thing" anymore, mainly due to sourcing of new chips. Also, it's a bit like soldering - everyone eventually develops their own style and gets used to their own set of tools... so for example, I may give you some specific advice that you may find doesn't work well with you.

All in all, the few (generic) tips I have are:
- Don't be afraid to let the board take its time to heat up. Too little heat, and like I said, you may damage stuff (when doing a chip pull) or not fix the issue (when doing a reflow.)
- Wear safety glasses when overlooking a hot board. Although I've never had small caps blow up and shoot out, the board is a big hot surface, so I just like to play it safe.
- Practice, practice, practice. The more times you do it, the better you will get - both in terms of your own skills and in terms of adjusting temperatures/profiles, and whatnot.
- Consider placing your "machine" (whatever that setup may be ) in an area that you can ventilate easily. It's not necessary to add flux when you remove stuff (and most of the time, also when you reflow chips). But it is necessary for putting on new chips. And this will give off a fair amount of smoke and/or fumes. Obviously it's not a good idea to breathe this stuff in all day, every day. So air the place out or have some kind of a duct to suck out the fumes.

[Dannyx]

Re: Reballing discussion and tips

Ok thanks for the reply...I was afraid this thread was completely dead before it even got 1 answer Generic and unspecific topics tend to do this, I agree.

[diif]

Re: Reballing discussion and tips

I agree with momaka, there is nothing hard about placing the chip back.
Once I'd got my profile set up (big thanks to Th3_uN1Qu3). It's a fairly straight forward job.

[momaka]

Re: Reballing discussion and tips

You should dig up some of Th3_uN1Qu3's old threads. Some of them have pretty interesting and useful info regarding reflowing/reballing (especially the one showing his home-brew rework "machine". )

Believe it or not, the hardest part of a reball IS removing the chip. Too little heat, and you'll rip traces. Too much heat, and you might popcorn the PCB or damage the chip (which I suppose may not be an issue if the old one is dead anyways.)

Putting on the new chip IS the easy part - just line it up with a naked eye and that's it. As long as the chip is reasonably close to where it should be... once those solder balls start to melt, they will "pull" the chip to where it should be (unless you really have it way off.) And if you are using leaded solder on the new chip (why wouldn't anyone), you won't have to heat the chip/board as much as before, so less chance of damaging something.

[Dannyx]

Good day folks. Here's a huge topic which I haven't touched upon until now, mainly because I had no way of doing it "properly" but thought I'd start a discussion now. We didn't have a reballing machine at my former shop, but we do here. It's only a "baby" one (a T890 -feel free to look up some pics) compared to other much bigger and industrial stuff that's out there, but I became more and more interested in it lately after watching one of my (much older) colleagues do it. It's a skill that seems to have a lot of potential in the field nowadays, so I'll eventually have to do it myself one way or another - either as a trainee of this guy if he agrees to it, which I doubt (who's stupid enough to let others steal their gimmick ?!) or on my own as a learning experience on some broken boards or something like that if I ever manage to reach that level....I'm getting chucked to the sideline here, as all these old bastards keep everything to themselves - no fair, but that's how it's always going to be in large companies ! That's what I hate about this joint...but I digress >_>

Not actually having done a reball operation myself hands-on before, the way I grasped some of its basics was the same way as the rest of the people and that is from the internet. One thing that's puzzled me though and seems to me as the trickiest part and is not quite clearly illustrated is how you align the chip back onto the board to solder it. I know about the stencils and the balls and the flux and stuff, but what about putting it back ? Breaking something is always easier than putting it back together, so I could probably get the chip off the board no problem, even as a beginner, but putting it back on those tiny balls.......yikes :|

I never got to see this chap putting the chip back - I always had to leave for various reasons just before he got to the good part. So..."how do they do it ?" The thing that's funny is that his methods and tools are far from professional or laboratory-grade - more like "DIY gadgets": he uses stuff like wooden skewers to hold or move the hot chips around, a piece of friggin folded paper to scoop up and pour balls over the chips in a metal bowl you usually mix salad in...you get the idea So if he can get away with THAT, I'd surely be able to pick it up as well Any good resources on reballing ? What are some of the dos and don'ts ? What tools and materials really make the difference here ? Etc. Cheers guys.