PS3 Frankenstein PHAT PS3: CECHA with 40nm RSX

[RIP-Felix]

sophistry

 

[Victor Hugo Alvarez]

sophistry

Me? Naah, ohh hardly i find it an pure Logical argument, theres kind of 2 possibilities now, put pressure only across all Rsx fiberglass but not in the die and if work then will be proof that theres no bump failure detect in that SPECIFIC System Rsx, if the system wont work and if try a second TEST Observation but now with Pressure on Silicon and the System Works that will be proof that, that system has bumps failure cause we try Firstly pressing its big Bgas and it didn't work But now with Pressure on die and the bumps now the console works.

All an absolutely logical teoretically observation, well we can have another effects being generated or fixing another part of the gpu, but the main problem sounds absolutely to be fixable by That logical theory which is Even simple but im the only and first suggesting Such test to detect bump failure by That kind of test.

Edit: Of course we could and most consider all things just like the Vrams in Rsx that use some sorts of Bgas But as I said before, the primary logic in on focus on Rsx Die and its Bumps and absolutely if we press all Rsx fiberglass and not the die of Rsx in a ps3 with Rsx related ylod by syscom diagnos or whatever and if that system start to work again THAT WILL BE PROOF OR EVIDENCE THAT, THAT SPECIFIC PS3 NEVER HAD BUMP FAILURE, now the simple part which is called EMPIRICAL TEST is Really so Easy Easy Easy Everybody can do to, just copy the test that I created! And repeat to see final absolut evidence of how much system would work or not with that test and bla bla bla.... empírical.....

 

[Dariush Aghaee]

Here.

Incredible! Awesome
But I always had a question. Way the products before that generation and years don't suffer the problem? Any fat PS2 for example is working like there is no life limit at all? I know there are some dead PS2s but the number is almost zero.

Better design of Cell? Well maybe in the Cells with Black Plastic between the IHS and its Body of Fiberglass, that one would Have better Copper wire inside of its fiberglass but only that, cause its thickness its ultra thin like half the or less the thickness of standart Cell fiberglass, but for that Cell only find in just some of the ps3 Ss I believe that one was obvious as seens it was made smaller fiberglass to cut production Prices on ps3.

Yes. Better Tg balance, IHS design, etc. when their knowledge of making processors improves, the cost of producing lowers but the improvement in the science should lead to more reliable production. I agree $ony just looks for more benefit but they don't make processors themselves and I really don't believe the actual producer steps back for lowering cost if stepping backwards makes no more production cost.

 

[Victor Hugo Alvarez]

sophistry

Bro well i'm Not trying to disagree with your research but I really have to say as an real scientific analisys that Great Majority of People when they Remove the Nec Tokins they do Not Use Heat, they use physical force, and for what I see theres lots and lots and lots of ps3s that have Tokins removed by physical force and Tantalums installed and they come back to life, I watch like hundreds of those vídeos, brazilians are Crazy to try repair ps3 like that and the ps3s do work!!! Also I think you forget the way of just Add more tantalums without even touching Nec Tokins and ps3 works, Example I had one cok-002 that when I Buy that ps3 and try Play dmc 4 it just ylod on Main Screen menu of Devil may Cry 4 then I installed 2 or 3 tantalums on Rsx line supply right close on the nec tokins and ps3 never more turned off in games, but then I lend that ps3 of mine to my uncle and he said the ps3 died and was ylod just in a specific part (that one you first take the molotovs and saw the creature inside the magazines underground somewhere) and right after my uncle told me I knew It, well The Last of Us is one of those games made by Naughty Dogs Who heavely Uses the Cell Processor 6 Spes Cores so of Course I just go and I installed more 2 or 3 tantalums on Cell BroadBand supply of Energy and hell Yea the PS3 stop yloding.

 

[Victor Hugo Alvarez]

sophistry

Felix You said The Nec Tokins only work at 67c on Coks, now do you believe 67c is enough to make that Burned Parts inside the Nec Tokins??? I dont think so!

 

[Pacorretaco]

That is an illistration. The voids and cracks from electromigration are microscipic and easily reconnected from slight changes in pressure.

I had an explosion effect and used it for visual purposes. It's not tlike they are burning to a crisp. Besides, even if some did there are many others that are on the verge of working and just haning on the edge of working. That's always the case, because the FlexIO can't maintail it's calibration if the bumps (or BGA for that matter) are marginal and constantly shifting impeedance during operation. When it gets just past the tippi g point, it begins failing. And then even miniscule changes in mounting pressure or temperature can cause it to work temporarily. Which can explain why people report being able to start the console after multiple attempts or heating it up first. Or a reflow, or completly unnecessary tokin replacment, or towel trick, or pressure test, or smacking the SOB upside the head a couple of times. The connection is precarious.

This is tricky but it can also be one of the important differences.

Slight changes in pressure affecting "the bumps"?
I disagree.
Why? Precisely because of your favorite thing, the underfill... You need a LOT of pressure (like a famous "bolt mod") to possibly affect "the bumps" since the area is smaller and the underfill is there protecting them. Would need to overcome its strength. Not something that can be done with bare hands, lightly flexing the board/hitting the machine... Reassembly...etc.

If smacking the top of the machine is changing the behavior...
You dont have a "precarious bump"... You have a precarious BGA that should be reballed. Many technicians will agree as we have seen this before, is not rare at all...

 

[Pacorretaco]

Felix You said The Nec Tokins only work at 67c on Coks, now do you believe 67c is enough to make that Burned Parts inside the Nec Tokins??? I dont think so!

All can fail...
Anything on the board can fail. We have found bad tokins. Not many on these old boards, but they are a thing. They can fail. (And not just because they turned brown...)
The big old BGA can fail... We have found that too.
Sometimes however the chip has internal damage. This is a fact. Why and how often each thing happens though? Well thats the real question.
Some people blame "the bumps". And mayyybe they could fail too. We just havent found "the broken bump" yet.

That was my original question which apparently not everyone understood. Where is "the broken bump"? Has anybody found it? How? How does it look like and how does it behave?

I proposed an experiment that has to do with the reply above:

To find "the broken bump": (no need 50k dollar X-rays)

Take an expertly reballed PS3 that failed again, suspicious of "broken bumps". Then give it extreme pressure and see if it works.

If it works with the pressure, it should be the "broken bumps" since it cant be the big BGA anymore.

This has been found repeatedly and reliably in the Xbox 360 with the famous "bolt mods" and the papers support it. But never on a PS3.
I hope it makes sense.

 

[RIP-Felix]

I think he is confusing electromigration with power surge. Very different things.

Electromigration is normal. Doesnt result from a fault. It's just an electrical phenomenon that has to be accounted for in the design of the processor.

Only a power surge would have the ability to burn out bumps and processor to a crisp, like he was talking about.

 

[RIP-Felix]

Slight changes in pressure affecting "the bumps"?
I disagree.
Why? Precisely because of your favorite thing, the underfill... You need a LOT of pressure (like a famous "bolt mod") to possibly affect "the bumps" since the area is smaller and the underfill is there protecting them. Would need to overcome its strength. Not something that can be done with bare hands, lightly flexing the board/hitting the machine... Reassembly...etc.

If smacking the top of the machine is changing the behavior...
You dont have a "precarious bump"... You have a precarious BGA that should be reballed. Many technicians will agree as we have seen this before, is not rare at all...

Peope seem to have this idea in their head that solids are hard, unbendable, not at all maleble. And that is a known, demostrably incorrect perception. It's only because of our scale. At our size, human beings don't preceive the physical world the way it actually works. So we see things from a biased point of view.

If we were scaled up, say to the size of a sky scraper, when we walked the ground would have visable waves, ripples. Because the ground at that scale doesn't behave like a solid, it behaves like a viscous fluid. People have reported seeing visible undulations of the ground during earthquakes for example. They describe it as a shocking sight, to realize their understanding of the world and place in it is so incorrect (and vulnerable). People have been litterally thrown off the ground by these ground waves, and injured falling back down.

My point is on the larger scale solids behave like fluids. It's because they do bend, compress, droop, sag, and drip. They are just far more viscous. And on the microscopic these small variations are magnified. If it bends an impreceptable amount when you press on the chip, heat it up, or change mounting pressure, that can be all that's needed to connect a broken bump. Especially one that was teetering on the edge.

So even though it feels like you're not bending the solid unferfill, it's actull bending and compressing just like a fluid, on the microscopic level. These cracks are that thin.

You are applying your preception of how physics behaves on our scale to the microscopic scale of these cracks. Which is not how it actually works at that scale. They don't behave the way it makes sense at at our scale. So when you use a "solid's don't bend or compress" argument, you are doing so from an scale that doesn't apply to this situation.

And if you think this is wierd, it onky get's wierder the smalle you go. When you get into quantum effects (like we're talking about increasingly smaller processors) you really have to abandon logic and rely of the absurd explanation quantum physics has come up with to explain their experiments. Defying logic on our scale nearly completly.

Again, point is you are using a plausable sounding, yet fallious argument. = sophistry

Pseudoscience loves sophistry. It doesn't stand up to scrutany, but the target audiance is not the serious scientific community (who will tear it apart for the trash it is), but rather the general public who are desperate to believe what they want to, truth be damed!

Also, trolls love to use such arguments to create a sense of debate where none exists. It appears to onlookers that "the experts" dont agree, therefore they can believe whichever explanation they like better.

 

[Pacorretaco]

Peope seem to have this idea in their head that solids are hard, unbendable, not at all maleble. And that is a known, demostrably incorrect perception. It's only because of our scale. At our size, human beings don't preceive the physical world the way it actually works. So we see things from a biased point of view.

If we were scaled up, say to the size of a sky scraper, when we walked the ground would have visable waves, ripples. Because the ground at that scale doesn't behave like a solid, it behaves like a viscous fluid. People have reported seeing visible undulations of the ground during earthquakes for example. They describe it as a shocking sight, to realize their understanding of the world and place in it is so incorrect (and vulnerable). People have been litterally thrown off the ground by these ground waves, and injured falling back down.

My point is on the larger scale solids behave like fluids. It's because they do bend, compress, droop, sag, and drip. They are just far more viscous. And on the microscopic these small variations are magnified. If it bends an impreceptable amount when you press on the chip, heat it up, or change mounting pressure, that can be all that's needed to connect a broken bump. Especially one that was teetering on the edge.

So even though it feels like you're not bending the solid unferfill, it's actull bending and compressing just like a fluid, on the microscopic level. These cracks are that thin.

You are applying your preception of how physics behaves on our scale to the microscopic scale of these cracks. Which is not how it actually works at that scale. They don't behave the way it makes sense at at our scale. So when you use a "solid's don't bend or compress" argument, you are doing so from an scale that doesn't apply to this situation.

And if you think this is wierd, it onky get's wierder the smalle you go. When you get into quantum effects (like we're talking about increasingly smaller processors) you really have to abandon logic and rely of the absurd explanation quantum physics has come up with to explain their experiments. Defying logic on our scale nearly completly.

Again, point is you are using a plausable sounding, yet fallious argument. = sophistry

Pseudoscience loves sophistry. It doesn't stand up to scrutany, but the target audiance is not the serious scientific community (who will tear it apart for the trash it is), but rather the general public who are desperate to believe what they want to, truth be damed!

Also, trolls love to use such arguments to create a sense of debate where none exists. It appears to onlookers that "the experts" dont agree, therefore they can believe whichever explanation they like better.

Oh yeah, the board was a liquid all this time. But dont worry about that, no argument there!
What you might worry about is using words such as "sophistry" and "pseudoscience", which somebody else could rightfully apply back to you, hehehe.
I know your intentions are good, but theres no need to overcomplicate things so much with technical voodoo, so much so that the normal person feel overwhelmed and exploit the fact that nobody will have the time to research so much stuff themselves.

The board may be liquid and all, but the underfill is there to provide support.
If the machine respond to light pressure, flexing or movement... Even simple assembly and reassembly...
It legitimately points to the big BGA. Which is bigger and does not have any underfill, making it actually susceptible to these physical movements or "waves".

"Bolt mods" on the other hand, are indeed capable of compressing the solid underfill "as a fluid", like you say... And that way possibly affect the bumps at least temporarily.
Not something that can be easily done with bare hands.

This is supported by the real world... And real technicians with real experience in these machines.
After such tests have lifted the chip, then found dodgy pads, dark, non-wet oxidized or sometimes broken. Not "broken bumps" because the reballing was enough in these cases.

Again, I could be wrong like anyone else but cant be wrong in simply asking for some evidence. If you think that there can be machines that respond to pressure but is "because of the bumps"...
Should be easy test. Make sure the big old BGA is ok by reballing, then see if it could still affected by a similar problem that respond to that pressure.

 

[Victor Hugo Alvarez]

I think he is confusing electromigration with power surge. Very different things.

Electromigration is normal. Doesnt result from a fault. It's just an electrical phenomenon that has to be accounted for in the design of the processor.

Only a power surge would have the ability to burn out bumps and processor to a crisp, like he was talking about.

Felix then you think slight or soft pressure can affect the Bumps!!? Alright that means that all pressure mods applied on processors should bring them back to life, so you have been refutated Felix cause no Puting an médium pressure will not bring processors back to life, you have to use alot of Pressure, an pressure like you believe like Moving the board around vs bolt pressure on processors is another whole universe history.

 

[RIP-Felix]

Slight changes in pressure affecting "the bumps"?
I disagree.
Why? Precisely because of your favorite thing, the underfill... You need a LOT of pressure (like a famous "bolt mod")

If it bends an impreceptable amount when you press on the chip, heat it up, or change mounting pressure, that can be all that's needed to connect a broken bump. Especially one that was teetering on the edge.

So even though it feels like you're not bending the solid underfill, it's actull bending and compressing just like a fluid, on the microscopic level. These cracks are that thin.

"Bolt mods" on the other hand, are indeed capable of compressing the solid underfill "as a fluid", like you say... And that way possibly affect the bumps at least temporarily.
Not something that can be easily done with bare hands.

You keep saying that as if by doing so it will somehow convine me. I have explained my reasoning. I believe it is possable.

This is a rediculous game of "yes it is." "No it's not." "yes it is"...repeating endlessly.

What we have here is an old fashon impass.

I don't see the point in further argument. IMO it is just as reasonable to expect you to provde proof the bumps do not cause thermomechanical reconnection. But I've not heard single purposed test that would do so. And I have repeatedly stated without 3D xray, I cannot provide proof of the bumps to your satisfaction. In neither case does the lack of proof support either position.

Where is "the broken bump"? Has anybody found it? How? How does it look like and how does it behave?

I proposed an experiment that has to do with the reply above:

The theory for bump defects is sound. It has already been proven to be an issue for most, if not all Nvidia chipsets during bumpgate. It is not my burdon to prove bumps. It lies on us to disprove it applies to the RSX. And I haven't heard anything that does.

The opposite is not possible. You keep attacking this point because you know we cant prove it's the bumps and your're hoping that will look like you're making a point. The only point your making is that we're all too poor to have a representative sample of RSX's sent to a 3D xray for silicon failure analysis. It doesn't mean it isn't there.

 

[Victor Hugo Alvarez]

I ALREADY TOLD HOW TO DETECT BUMP FAILURE IN AN RSX GPU, SO CAN ANY BODY PAY ATENTION OR AT LEAST LIKE MY COMMENT, JUST DO WHAT I SAID ABOUT PRESSURE ON RSX VRAMS, I DON'T HAVE ANY BOARD NOW WITH RSX BGA/BUMP ISSUE SO I CANT TEST IT.

 

[Victor Hugo Alvarez]

@RIP-Felix if you wanna absolutely Scientific tests then you shall know you don't use the words "break" or "burn out" when you trying to say that the bumps got Cracked.

 

[RIP-Felix]

I ALREADY TOLD HOW TO DETECT BUMP FAILURE IN AN RSX GPU, SO CAN ANY BODY PAY ATENTION OR AT LEAST LIKE MY COMMENT, JUST DO WHAT I SAID ABOUT PRESSURE ON RSX VRAMS, I DON'T HAVE ANY BOARD NOW WITH RSX BGA/BUMP ISSUE SO I CANT TEST IT.

I head what you said. Let me reiterate. It wont rule out the bumps.

1. You need to try it on many ps3's with the same problem
2. You need to delid

It changes the pressure equation. First, the pressure of the leaf spring that is compressing the die will be gone. Decompressive load. Second you have to apply alot of pressure to the interposer to delid in the first place. Changing the in situ balance of pressures that have led to the precarious connection. Both of which are a plasuable competing explantion for either result. If it reconnects, could be bumps reconnected due to slight changes in pressure due to the delid or lack of pressure on the die. If it doesn't work, then you have to assume BGA's are not it and it supports the bumps hypothesis. Which assumes that the BGA cracks would respond to the pressure test, which they don't always. This is something I don't make assumption on. If the pressure test works it confirems either a BGA or Bump defecr. If it doesn't work, it rules out nothing. There can still be either.

So that test wouldn't yield any useful results.

 

[RIP-Felix]

@RIP-Felix if you wanna absolutely Scientific tests then you shall know you don't use the words "break" or "burn out" when you trying to say that the bumps got Cracked.

Those are common terms that people understand and give a good enough approximation for the idea. Viewers/readers require scientists to use common terms and illustrations to make an otherwise difficult to understand topic and easy one. In my experiance only people trying to hide their mistakes and culpability present the issue using tehcnical terms, in the hopes people wont understand.To bamboozle them.

"Break" can be a crack or sheared pad resulting from cumulative strain and deformation. "Burn out" refers to voiding and cracks that result from the foce of the current passing through. BTW that same force can wear away at the dielectric that separates the nanoscale traces of the die itself, leading to shorts. And heating the chip above the dielectric's Curie point can heal the dielectric and restore the separation temporarily. So reballs can cause that situation to appear resolved when they are not.

There is also the possability that the heat from a reball can melt the 5-10% tin content in the lead bumps to partially make a connection inside the bump, which is not strong. Kinda like gluing a crack. Once the syress resumes, stress will exploit the crack and propagate along it, breaking the "glue."

Point is, there are lots of plausable explanations for why reballs appear to work and can vary in how long they last. And if I covered all of them in the video, it would have been 2 hours long. It was already long enough.

 

[Octal450]

This is tricky but it can also be one of the important differences.

Slight changes in pressure affecting "the bumps"?
I disagree.
Why? Precisely because of your favorite thing, the underfill... You need a LOT of pressure (like a famous "bolt mod") to possibly affect "the bumps" since the area is smaller and the underfill is there protecting them. Would need to overcome its strength. Not something that can be done with bare hands, lightly flexing the board/hitting the machine... Reassembly...etc.

If smacking the top of the machine is changing the behavior...
You dont have a "precarious bump"... You have a precarious BGA that should be reballed. Many technicians will agree as we have seen this before, is not rare at all...

This is spot on. I have used this exactly to successfully diagnose the issues with very high reliability (95%+).

Kind Regards,
Josh

 

[Pacorretaco]

I don't see the point in further argument. IMO it is just as reasonable to expect you to provde proof the bumps do not cause thermomechanical reconnection. But I've not heard single purposed test that would do so. And I have repeatedly stated without 3D xray, I cannot provide proof of the bumps to your satisfaction. In neither case does the lack of proof support either position.

The theory for bump defects is sound. It has already been proven to be an issue for most, if not all Nvidia chipsets during bumpgate. It is not my burdon to prove bumps. It lies on us to disprove it applies to the RSX. And I haven't heard anything that does.

The opposite is not possible. You keep attacking this point because you know we cant prove it's the bumps and your're hoping that will look like you're making a point. The only point your making is that we're all too poor to have a representative sample of RSX's sent to a 3D xray for silicon failure analysis. It doesn't mean it isn't there.

Felix I get it, nobody has time to argue.
But maybe our case is a bit different? After all it was you who made long videos and all, which I like too.

I did not ask for "proof". I just asked for some "evidence" AT LEAST, which I would still like to see... Otherwise are just incomplete stories, sorry...

And maybe the test that @Victor Hugo Alvarez suggested was incomplete because it was missing a step but...
There was another proposed test that did not involve expensive X-rays. Should not that difficult, especially according to the stories.
That was my original question which I will quote again:

To find "the broken bump": (no need 50k dollar X-rays)

Take an expertly reballed PS3 that failed again, suspicious of "broken bumps". Then give it extreme pressure and see if it works.

If it works with the pressure, it should be the "broken bumps" since it cant be the big BGA anymore.

This has been found repeatedly and reliably in the Xbox 360 with the famous "bolt mods" and the papers support it. But never on a PS3.
I hope it makes sense.

 

[truemaster]

ok guys looks like we have 2 different opinions about the 90nm gpu my self felix and dead end strongly believes that the bga is for replace. but the otherside of people that believe the reball is a permant fix they have the right to believe so its not a crime. but heres a neutral and logical answer if yours ps3 has a 3034 (gpu error) what would you do? reball the chip? or change it with a working one? and since there is a way to solder back a newer made gpu that produce less heat, cosumes less power and the phat heatsing can keep it twice as cooler compared to slim why not change it with 40 nm? for a peace of mind? reball the same gpu can be a two side blade, and replacement with a known working one or newer one would be my choice and that not only for ps3 but on anything with bga issues as long the said bga is not married to motherboard as the cpu of ps3-xbox 360. reaaly now continue to argue is pointless lets accept that we have different opinions and be done with it

 

[Pacorretaco]

ok guys looks like we have 2 different opinions about the 90nm gpu my self felix and dead end strongly believes that the bga is for replace. but the otherside of people that believe the reball is a permant fix they have the right to believe so its not a crime. but heres a neutral and logical answer if yours ps3 has a 3034 (gpu error) what would you do? reball the chip? or change it with a working one? and since there is a way to solder back a newer made gpu that produce less heat, cosumes less power and the phat heatsing can keep it twice as cooler compared to slim why not change it with 40 nm? for a peace of mind? reball the same gpu can be a two side blade, and replacement with a known working one or newer one would be my choice and that not only for ps3 but on anything with bga issues as long the said bga is not married to motherboard as the cpu of ps3-xbox 360. reaaly now continue to argue is pointless lets accept that we have different opinions and be done with it

I appreciate that you try to make peace (theres already peace by the way), but I need to reply fast because I dont want people to get confused like last time.

That may or may not be "neutral" or "logical" but is certainly not an "answer".
Because it has nothing to do with the discussion. Never had... People were always free to do what they want or what they think is "best" or "more practical" for them.

But that "peace of mind" may be due to fear of something that is not that well understood...
Peace of mind didnt which work so well for you in particular, when your board failed shortly after anyway... After the the 40nm RSX replacement from Deadend.