Personally we find this Ball-Grid Array chip-swap rather horrifying. But if you want to beef up the processor on your 701 Eee PC this is what you’ll need to go through. Not only did [Red Fathom] upgrade to a 1.6 GHz chip, but he managed to get the computer to boot up with the new hardware in place.
BGAs are notoriously hard to solder. This hack pulls it off using just a hot air gun. [Red Fathom] heats the board from the underside until the solder melts and he can pluck off the old chip. He then uses a solder braid and iron to remove extra solder from the footprint. After a little cleanup with a cotton swab and some flux he plops in an Intel Pentium M LV 778. It doesn’t look like he added any solder after the cleaning process. Perhaps he’s relying on the small amount left on the tinned pads of the board?
After the break you can see the soldering process and a video of the new processor booting Xandros.
Soldering video:
Booting Xandros after the swap:
[Thanks Nonya-Biz]
The replacement processor is factory new or bought “Reballed” on eBay etc. the balls on the chip is sufficient to solder on the new chip – especially if the pads has just been soldered before and cleaned, it’s a breeze to remount a new chip.
Reflow and stenciling BGA is easy compared to soldering small-trace stuff or even some analog. There is just no stencils or discussion on it…
I did a OMAP some years ago along with a RF generator..
Action happens at 6:30 and 13:50.
Pretty badass though, it sure does take balls.
^^ I see what you did there. :)
He really should be preheating that board with a proper rig. This time he got lucky, but he will have bad joints that may show up under thermal stress later on.
I was about to say…chips are usually pre-balled (or reballed) so they can be soldered back on.
One of the other concerns when working with reflow is with RoHS (lead-free) solder is cold-solder.
If enough heat isn’t used, there’s a chance that the solder did not melt and crystallize and can lead to the solder cracking over time causing all sorts of strange timing, corruption issues.
In CPUs, it can be anything from memory issues, data corruption to BSOD/KPs.
In GPUs, corruption (texture, etc.) to “flying triangles” (remember the ol’ *vidia issues?)
anyways, awesome vid.
reminds me of that one guy who reflowed and upgraded his ultraportable UMPC.
Never seen someone do BGA like that, heating below the board only, guess he was lucky. If you fix motherboards for a living, you need to be more careful than that.
Anyway, the new processor already had the balls to be soldered, the most difficult part is the reballing itself, which he skipped (no need with a new BGA).
Heating below the BGA is common. Usually the area below the big BGA part is pretty open, full of bypass caps and other small stuff that just sticks due to solder surface tension and tinning adhesion so they don’t doesn’t fall off. Some heat with the board upside down and the BGA falls off by itself because it’s heavier.
As for re-balling, it’s pretty easy if you have a tube of balls and a ball stencil; both of which are not particularly cheap or off the shelf. Flux, then fix the ball stencil, pour the balls, lift the stencil, inspect, then use rework hot air to melt the balls, done.
This is not the real process, but it is a hack. And it must be taken for what it is.
To complain about him is unnecessary.
I have changed perhaps 1000pcs BGA with real equipment and I am impressed that he succeeded.
That is way too delicate for me to even think of,but i wonder if the cpu will have enough cooling or if it will start overheating.
I wonder if the 900A/901 use the same socket. Hmm
Great job on updating the cpu :) but that screen looks nasty :/
also – what’s the name of the background tune? anyone?
http://www.youtube.com/watch?v=i_0G6WPuss4&feature=related
Great skills!
Ballsy, hope it holds up in the long run.
You, Sir, are one crazy mofo!
Loving this hack as this takes some real bravery, altho he seems pretty confident in his methods!
Also loving the jungle vibes in the background, can we get a track ID? :)
@sonofabit too
“Foregone Destruction”, BGM from Unreal Tournament ’99 ;)
Unbelievable. Really, it’s unbelievable. I’m willing to call this as being fake.
I took over 6 minutes for him to pull the CPU off, which does sound rather reasonable. But then it took just under 2 minutes to reflow the new CPU on to it!!! Yeah, sure, the pcb did still have a little bit a residual heat before the reflow, but not much. Solder does not melt so easy when it isn’t attached to the surface that is being heated. In general it will take long to reflow a BGA onto a board than it will to remove a BGA.
The guy barely placed it. OK, maybe he has the hands of Jesus and first try it landed perfectly in place (Murphy’s law tell us that it never happens).
But then the fact that the guy used a Q-tip to apply flux, and thats the worst thing to do because it ALWAYS will leave behind fibers on the tiny barbs on the solder pads. OK, maybe he was using a microfiber Q-tip, but what about the fact that heating a board from only one side causing it to flex and bow, plus not having a proper ramp up and ramp down process? This just does not happen.
And as I already mentioned, it takes longer to transfer heat to solder that is not directly attached to the part you are heating so it does take longer for it to reflow especially when only being heated from the bottom.
I’ve done LOTS of BGA reflows and reballings using a huge range of tools and it does not happen this smoothly going about it in this way.
I’d say this is real, done it a few times myself. The solder used at the manufacturers was likely lead free whereas the processors can usually be bought RoHS/Non RoHS which explains the long removal/quick resolder.
Luckily I now use a Jovy for most of my BGA work which speeds things up a lot.
I can do it in less than 2 minutes from start to finish. It’s called having the right tools. Hot air rework station and a IR rework station. (both bought off ebay.)
The dude needs to buy some real tools.
I’ve reballed xbox gpus with a heatstation before, I worked it out to taking about 50 seconds to liquify the solder.
Reballing, yeah it only takes about a minute for the balls to liquify, but to then reflow that reballed cpu takes more than a minute, and the ramp up and ramp down time should take 5 – 6 minutes to do.
I own a hot air station, an IR preheater, and a decent reflow oven… a real reflow oven, not one of those home made/hacked ones.
Overall, I would never try this with a heatgun.
sorry if i’m picking on you. you make alot of points that repeat in other posts.
60/40 is all around better stuff, unless you plan on eating/breathing it.
i didn’t just drop it on there. you can see the beam from the 10w led flashlight i used to look under the chip. the dimples in the board from the solder mask also make it easier to allign.
barbs? that’s the point of cleaning the board. you can see me checking the smoothness of the board to make sure all that rohs solder is gone. i usualy put some flux in the solder braid to make it wick better. the qtips arent a problem if the board is properly clean. the chip is from a junk ibm x41 motherboard reballed.
small boards are less susceptable to flex. i heated it from below to keep from melting all the fragile connectors stickers etc on the top of the board. you can varry the heat just by moving the heatgun up/down from the board.
if it didn’t take some skill, or knowledge it wouldn’t be a hack it would be “idiot blows up board with heatgun” or “well manered man uses proper tools, to replace chip, with same chip!(yay)” didn’t expect this to even end up on hackaday.
the tools for this cost me $25, including the ungar variable temp “wood burner”. in total the whole thing would have cost $85, including the eee pc not shown. the lcd is from another eee.
the song is foregone destruction, by Michiel van den Bos.
When the 4gb onboard flash of my Eee900 died I bought a really cheap 2nd hand Eee701 and desoldered the 4 flash chips and then soldered them into the Eee900, it sadly didn’t work (the flash control chip was designed for a different chip manufacturer I think) so I transfereed the chips back into the 701 and they still work to this day in that machine.
Then I bought a 2nd hand Eee900 :)
I thought that was some intense soldering but replacing the CPU, that’s some serious solder work!
> It doesn’t look like he added any solder after >the cleaning process.
>Perhaps he’s relying on the small amount left on >the tinned pads of the board?
Writing articles about how hard BGA is to work with without understanding how BGA works?
I’m not calling ballsheets on this, ’cause been there done that. Just not in this careless way. This guy certainly has luck, anyone working with BGAs know that gunning a PCB with hot air from behind is BAD idea if it’s not firmly seated to prevent bending from thermal stress, let alone in that position. It’s a miracle his board survived and doesn’t ended like a boomerang. And well, there is many issues (broken pads, oxidization, popcorn effect, etc) to care for when doing BGA work at home…
A proven DIY-method is the “electric stove + heat gun + thermocouple + chronometer” combo, with patience and practice some very decent results can be achieved. I’ve reworked and swapped many GPUs in this way, with great success to date (knockin’ wood). Though not for profit, I just enjoy repairing dead things.
Solder paste and screening is something evidently most people here don’t know about..
Also the proper way is heat gun over the chip on a stencil, paste with screen stencil, seat in stencil heat gun..
foundries and factories use stencils, screens, and ovens..
Its evident some people dont know what a bga is and how its supplied….
I admire the guts of this guy. And yes, maybe he got lucky, and yes, I agree that ‘this is not the way to do it’, but that’s part of what ‘hacking’ is, isn’t it? I haven’t heard him say that this is the right way to do it.
Also, I haven’t seen him testing it ;)
Until I saw the second video…
ok, I will ask the question noone else has asked… how did he do this with a heat gun without stripping all the parts off the other side of the board? Not mention that just wantonly heating the board like that would also throw off other parts even on the top side. The slightest vibration would knock those tiny parts all over the place. There is a reason we use heat guns to quickly strip a board of parts for reclaiming them.
And I may have to call BS on this as well. The processor was definitely skewed still. I understand that surface tension of the solder would pull the BGA into place, but it was still skewed even after he claimed to be done.
I have successfully done a RAM chip once before using a hot air reflow station. But it required a ton of prep work including lots of kapton tape, removing parts that I knew were going to go flying and reinstalling them after, making tin foil masks, shields, heat sinks, etc… I also had to make a deflector out of the tinfoil to redirect the air under the BGA part (I actually found a nozzle that does this just recently.)
Obviously the pin density on a RAM chip is not as great as on this processor.
Watch this fella and your jaws will drop guaranteed:
[youtube http://www.youtube.com/watch?v=u3oSQIURiZ0?rel=0&w=420&h=315%5D
That kid is good :)
He had time to swap circuit board at 6:38…
If it is real, it is very impressive, and lucky. I miss the days when laptops came with sockets for the CPUs; it wasn’t that long ago.
My Advent 7113 laptop started out with a Celeron M 1.73GHz CPU but was easily upgraded to a C2D 2GHz thanks to the socketed CPU. Lots of AMD laptops have socketed CPUs as well.