US Does anyone reball a ps3?

Discussion in 'The ASSEMblergames Marketplace' started by kreg, Feb 2, 2019.

  1. kreg

    kreg Spirited Member

    Joined:
    Feb 22, 2009
    Messages:
    144
    Likes Received:
    24
    Im looking for a ps3 motherboard repair. I could just as easily replace it, but... I'm attached to it and wondering if it can be saved :). Looking for anyone I can hire. Its from a CECHL01

    Thank you
     
  2. ItsMeMario

    ItsMeMario Gutsy Member

    Joined:
    Feb 11, 2014
    Messages:
    486
    Likes Received:
    166
    Since I have seen Louis Rossmann talking about reballing on youtube back in 2014, I have a whole new opinion about reballing :

     
    modrobert and Pikmin like this.
  3. bowser22

    bowser22 .

    Joined:
    Mar 23, 2012
    Messages:
    204
    Likes Received:
    132
    @kreg , gocybershopping in Canada does these repairs and I recently talked to them, they warrant their work for 6 months as well.
     
    Last edited: Feb 3, 2019
  4. benny679

    benny679 Active Member

    Joined:
    Jul 14, 2013
    Messages:
    40
    Likes Received:
    11
    I can't say I agree with that, the way he explained it was very unscientific and I couldn't find any sources. We know tin whiskers do form with lead free solder... WIth the xbox 360s, the failure point actually was the solder balls. Occasionally e74 would be caused by a GPU with microfractures.

    The part where he said to heat the board and see if it fixes it can be the result of board flex.
     
    bowser22 likes this.
  5. nonosto

    nonosto Intrepid Member

    Joined:
    Jan 5, 2012
    Messages:
    698
    Likes Received:
    54
  6. segasonicfan

    segasonicfan Robust Member

    Joined:
    Sep 23, 2011
    Messages:
    262
    Likes Received:
    40
    Great video, I didn't realize there were such significant differences with VDP reballing. I will say though that I've had luck installing custom fan control software after reballing. I think it can work as a long-term fix so long as you're not doing intense graphics work and there are new measures to control the heat in place beyond OEM.

    Basically I agree with Louis Rossman. Gotta keep customers happy long-term if you want to stay in business (and of course do the right thing).
     
  7. TriMesh

    TriMesh Site Supporter 2013-2017

    Joined:
    Jul 3, 2008
    Messages:
    2,355
    Likes Received:
    778
    I think he's full of shit. I have worked doing failure analysis on some very complex PCBAs and the number of defects I've seen caused by 2nd-level interconnect failures can be counted on the fingers of one hand. Over the same time period, I have seen hundreds of defects caused by cracked solder balls on the primary interconnect - and there is no guesswork involved in this because you can clearly see the fractures on an X-ray.

    In my opinion, the reason he's seeing these failures is that he's damaging the 2nd-level interconnect during rework - either by uneven heating of the chip, excessive temperature ramp rates or through popcorning (since I've never seen him bake a board before working on it).

    So on one level he's right - HE shouldn't attempt to reball chips because he lacks the skills to do it correctly, but trying to claim this is because the primary initial failure mode is 2nd level interconnect failure is just bullshit.
     
  8. ItsMeMario

    ItsMeMario Gutsy Member

    Joined:
    Feb 11, 2014
    Messages:
    486
    Likes Received:
    166
    @TriMesh

    My impression also was, that hes talking a bit more about those "little computer shops" and/or "we reball on ebay" people here, that are just using a hot air gun...and/or simple youtube soldering skills, which is indeed NOT a professional way to fix someones laptop/console. It works, but the customers will come again.
     
    Mord.Fustang likes this.
  9. TriMesh

    TriMesh Site Supporter 2013-2017

    Joined:
    Jul 3, 2008
    Messages:
    2,355
    Likes Received:
    778
    I wouldn't be disagreeing with him at all if he had said that - using a heat gun is indeed a very good way to damage the 2nd level interconnect in the chip because it heats things up way too fast and unevenly. What I was taking exception to was the assertion that the 2nd level interconnect was a predominant initial failure mode - because it's simply not true. Just ask anyone at all that has worked in failure analysis with this stuff.
     
  10. Nully

    Nully Dauntless Member

    Joined:
    Nov 23, 2010
    Messages:
    740
    Likes Received:
    103
    You're not alone, he was called out and ridiculed on many forums etc for that video.
     
  11. zzattack

    zzattack Spirited Member

    Joined:
    Nov 19, 2014
    Messages:
    161
    Likes Received:
    62
    I don't know the specific reason behind BGA chips failing in MacBooks, but the phenomenon of BGAs simply being bad and eventually failing, yet possibly having their lifespan somewhat extended due to heating up the chip is definitely a thing.
    So whether or not his claims are bullshit comes down to what applies to the chips in which's context he makes his claim.
    There's 2 reasons to reball:
    1) to replace lead free solder
    2) to magically revive chips where soldering isn't the problem

    Now 1) is totally valid and as we all know has resurrected thousands of game consoles. As for 2), yes indeed sometimes prolongs the usability of some devices, but cannot remedy the fact that other BGAs simply degrade due to other faults.

    Here's an example of a series of chips where the manufacturer admits to premature malfunctioning: http://processors.wiki.ti.com/index...x/1x/0x_IO_Buffer_Premature_Aging_Assessment#

    Yet, these chips often regain their function for several months after applying heat to them. That effect is what I guess Rossman refers to as 'reballing is bullshit'.
     
  12. benny679

    benny679 Active Member

    Joined:
    Jul 14, 2013
    Messages:
    40
    Likes Received:
    11
    Not to derail this thread even more, and I'm not an expert on the subject... my degree is in computer science, not electrical engineering, but I've done many repairs on the 360 consoles.

    To elaborate on what I said above, the Xbox 360 was plagued by not only the EPA standards at the time but the design of the console, and I imagine the PS3 suffered a similar fate. The Xbox 360's exterior design was finalized before the actual hardware was decided on, and thus the engineers had to work around that. If anyone here has ever owned a PowerMac G5 (I own a dozen of them), you know the lengths Apple went to to try to keep them cool. The heatsinks are massive, and they eventually moved to liquid cooling. Granted, the 360 has an advantage since the lithography is smaller, and the GPU more often failed, but its worth noting. Also the GPU had a smaller heatsink...

    Anyway, very rarely have I encountered an actual chip failure repairing these machines. It is true that the E74 occasionally will be the result of a defective GPU, but you have to consider the history of the console. More often than not, the console was already "repaired" using a paint stripping gun and machine screws, likely with not even enough heat to melt lead free solder. Don't forget the towel trick, or jamming the fans with nails, or stuffing pennies under the RAM chips... I could go on.

    To clear up what this was actually doing, and what Louis is likely talking about, was warping the shit out of the board. Take out any Xenon or Zephyr 360 board, which IIRC was a 90nm process versus the later 65nm, and they will be warped. Heat it enough and the board will warp enough to make contact with whatever solder joints were broken from thermal cycling and warp. The other method of course was people using machine screws to squeeze the board to the chip to make contact. This may have improved heat dissipation a bit, but warped the boards even more since nobody ever thought to use a backplate.

    If anyone in this thread has a lead solder reballed 360, feel free to chime in. Has it failed again within a few months? Didn't think so...
     
  13. benny679

    benny679 Active Member

    Joined:
    Jul 14, 2013
    Messages:
    40
    Likes Received:
    11
    My first thought was that he was working on machines that were previously reflowed or reballed with bad temperature profiles, but its possible. Was that bit at the beginning of the video of Macs having a timer to fail a joke or was he actually serious? If he was serious then he seriously is delusional.
     
  14. jjona5

    jjona5 Spirited Member

    Joined:
    Feb 18, 2015
    Messages:
    128
    Likes Received:
    4
    I had my ps3 phat backwards compatible professionally reballed(never opened prior) both chips with lead and have had zero issues since. I forget the company I used but they were big at the time and did lots of different consoles. I have probably used it a few hundred hours since it was fixed. I also installed the aftermarket fan that used to be sold and put it a later gen power supply that was supposed to run cooler.
     
    benny679 likes this.
  15. TriMesh

    TriMesh Site Supporter 2013-2017

    Joined:
    Jul 3, 2008
    Messages:
    2,355
    Likes Received:
    778
    It's a lot more complicated than that - the first thing is that solder ball failure on BGA packages is something that's been known about for a long time - well before anyone was using RoHS solder. It was only considered a significant problem in some applications though (one notable one was in car electronics - if you're in a cold climate the engine compartment can undergo very large temperature swings).

    It was a combination of things that made it such a major issue with these consoles - the first (and the one everyone tends to talk about) was the use of RoHS solder, which is definitely less ductile and more brittle than the SnPb solders - but there have been a large number of products that use it without any issues. To see why the early 360s were so bad, you need to look at the entire design and see how all the risk factors come together.

    The next thing is the use of fcBGA packaging - these have two sets of solder balls - one ("Primary interconnect") are the ones you see on the underside of the package and the second ("2nd level interconnect") are the ones that attach the die to the substrate. One non-obvious thing about this construction is that although the substrate is a form of PCB material the resin is carefully selected to match the TCE of the die as closely as possible to minimize stresses on the (much finer) 2nd level interconnect - but this also increases stresses on the primary interconnect.

    Additionally, these are relatively high-power chips and dissipate significant heat, which obviously tends to cause significant self-heating. In an attempt to mitigate this, the chips are heavily clock-gated (I.E. they turn on the sections that are not in use right now) - this has the positive result of reducing power consumption and heat output, but also the negative result of causing the chip to undergo many more heating and cooling cycles than would be typical in a non clock-gated design (which just gets hot when you turn it on and stays that way until you turn it off again). This imposes a much larger number of thermal cycles on the interconnect.

    Finally, these are consumer products intended for use in a domestic environment, and the thermal controls are designed with an emphasis on reducing fan noise over maintaining low temperatures. The poor design of the 360 heatsinks didn't help here either.

    Yes, replacing the solder balls with the (more ductile) SnPb types will certainly increase the fatigue life of the interconnect - but my feeling is that it will eventually fail anyway simply because the thermal design is bad.

    Yeah, I just took exception to his claim that 2nd level interconnect failure was the primary failure mode, because it's just bullshit. In reality it's sufficiently rare on PCBAs that haven't been messed with that if you see it there is tendency to suspect that the part either had an inherent defect or something bad happened to the board, like being dropped.

    I think the comment about a warranty expiration timer is almost certainly a joke - it's a running joke in the electronics business anyway.
     
    Nully likes this.
  16. modrobert

    modrobert Rising Member

    Joined:
    Jul 23, 2005
    Messages:
    68
    Likes Received:
    21
    If you look at the warranty times given for reflowing vs reballing for PS3 consoles, usually it's one 1 month (or none) for reflow and 3 months for reballing, gives you an idea what to expect, and also proves there is more to it than just the soldering alone.
     
    Nully likes this.
  17. benny679

    benny679 Active Member

    Joined:
    Jul 14, 2013
    Messages:
    40
    Likes Received:
    11
    I did state that about the thermal issues IBM was having with PowerPC in general later on. Its part of the reason that failures were mitigated in later consoles, they went from a 90nm process to a 65nm process and eventually 45nm with the slim consoles. Japser, Corona, Trinity, and even Falcon failures are rare, despite the phat consoles sharing essentially the same design as the earlier models. Thermal cycles definitely play a role but those boards also warped really bad if you inspect them.

    It sounds like the MacBooks tend to run very hot because Apple doesn't want to take away from the aesthetics of the computer and add real vents to it or include many fans. It would make sense that the primary interconnect would fail, but for whatever reason everyone seems to insist that the chips themselves are bad. I haven't seen any real evidence.
     
  18. segasonicfan

    segasonicfan Robust Member

    Joined:
    Sep 23, 2011
    Messages:
    262
    Likes Received:
    40
    Not many people talk about the heatsink design failure of the 360. I was absolutely blown the F away at how bad it was designed. The metal plate underneath the GPU and CPU is actually thermally isolated from the heat sink over those chips... So heat doesnt actually get to travel onto the other side of the board and make contact with all the thick shielding that surrounds the console. It just sits there on a little island of thermal insulation with the hope that the fans can pull enough of it away before the bga's cook. And we all know how that turned out. Seriously the most idiotic heat sink design I've ever seen.
     

Share This Page