PS3 (Research/Experimental) - NEC/TOKIN Capacitors Replacement - YLOD

[raidriar]

Send him a PM, he doesn't check the forums regularly. Oh, be sure to include your errorlogs, proving you have the requisite 1002 errors. Otherwise he'll assume you're another cultist.

I do not have 1002 only 1001 and it's extremely random when it happens

 

[RIP-Felix]

I do not have 1002 only 1001 and it's extremely random when it happens

He might agree to do it as a test. I know he has an oscilloscope and can confirm the voltage ripple. I don't think he's really into troubleshooting/repair services. He's more of a buy/fix/sell guy. You'll have to ask him.

1001 is not closely correlated to bad tokins, like 1002 is. 1001 is still mysterious. It happens with alot of other errors and sometimes spontaneously. We don't have good evidence to conclude it has to do with the tokins. And IF bad CPU tokins can cause 1001s, then the only way to know for sure is to probe them with an O-scope. It could be one of the first warning signs, however. So I'm curious to see what your ripple looks like under load.

When it comes to 1001, I have a suspicion about the input capacitors to the CPU's PWM Buck controller, but that's completely speculative.

 

[raidriar]

He might agree to do it as a test. I know he has an oscilloscope and can confirm the voltage ripple. I don't think he's really into troubleshooting/repair services. He's more of a buy/fix/sell guy. You'll have to ask him.

1001 is not closely correlated to bad tokins, like 1002 is. 1001 is still mysterious. It happens with alot of other errors and sometimes spontaneously. We don't have good evidence to conclude it has to do with the tokins. And IF bad CPU tokins can cause 1001s, then the only way to know for sure is to probe them with an O-scope. It could be one of the first warning signs, however. So I'm curious to see what your ripple looks like under load.

When it comes to 1001, I have a suspicion about the input capacitors to the CPU's PWM Buck controller, but that's completely speculative.

If you are in the USA you are welcome to take a crack at it. You would think if it were load related, as soon as a CPU spike came through, it would kill the system. I've done it in TLOU and Incharted 2, both of which have nice spikes but it doesn't trip. It's completely random. Sometimes it is during rendering, sometimes it is during cutscene playback. Could it be it does this because the board starts to flex after it heats up and then loses contact with the PCB? This unit did not have a warranty sticker and has a 430 day total runtime on it

 

[RIP-Felix]

If you are in the USA you are welcome to take a crack at it. You would think if it were load related, as soon as a CPU spike came through, it would kill the system. I've done it in TLOU and Incharted 2, both of which have nice spikes but it doesn't trip. It's completely random. Sometimes it is during rendering, sometimes it is during cutscene playback. Could it be it does this because the board starts to flex after it heats up and then loses contact with the PCB? This unit did not have a warranty sticker and has a 430 day total runtime on it

Well, I can't rule out BGA wear, but you don't have the tell tale 3034/4xxx errors yet. So I wouldn't worry about it until it happens.

A random, intermittent YLOD was the use case for this mod in the first place. That's the type of YLOD the OP originally reccomended this fix for. They're pretty rare, but have had good success rate. I suggest first attempting tantalum piggybacks...

Just scrape away some soldermask next to the tokins and install a couple of tantalums on each side, maybe 1 or 2 to each CPU and GPU. You don't have to do as many as in the pic above. There's a decent chance this will fix your issue, for now at least.

I'm not keen on taking on modwork right now. I'm currently in more of a learning phase. I've been watching EEVblog videos, learning about Op Amps, VCOs, PLLs, PWM controllers, etc. When I say learning I mean blank staring at the screen hoping a light turns on. I've been reading the datasheets, cross-refrencing schematics, cataloging error codes...all in an attempt to better understand the power on sequence and what can go wrong in the process. It's basically a crash course in electrical engineering. I literally had to cool it this last week due to eyestrain.

It's starting to pay off though. This week I fixed another microscope at work. I thought it had a bad LED driver (a simple dimmer circuit consisting of a LM2575 DC/DC converter, LM358 OP Amp, a potentiometer and Min/Max trim pots). I reverse engineered the whole circuit before finding out that the wires connecting to the power rocker switch were dodgy...haha! I soldered them and it worked. Wasted a bunch of time barking up the wrong tree, but feel proud I was able to figure it out without paying thousands to a repair service. In case you're wondering, we put off this sort of work until there's down time. This week was slow, so it was a good time to putter away at the bad scopes (make this hobby useful to my employer). It has already saved us over $10,000 in maintenance/replacement costs. I usually check for loose wires first, so IDK why I missed that. Whetever, it works now.

It still feels like I have a long ways to go before gettiing how all of this works together to make the PS3 tick. But I feel like I understand the basics powering the board, which is alot more complicated than you would think. And the devil's in the details. Honestly I think at this point I'm going to have to start making assignments with the math to force myself to gain an intimate understanding of what's going on. Without running the numbers and watching the effect it's hard to grasp basic concepts and retain that knowledge. It's just some of the concepts are redicouous to wrap my head around. I've read and watched videos multiple times and still can't tell you what a pole on a Bode plot is. Then there's identities and calculus...CALCULUS!!! It's at the point I'm asking myself, "WTF am I doing?"

I have a problem!

 

[raidriar]

Well, I can't rule out BGA wear, but you don't have the tell tale 3034/4xxx errors yet. So I wouldn't worry about it until it happens.

A random, intermittent YLOD was the use case for this mod in the first place. That's the type of YLOD the OP originally reccomended this fix for. They're pretty rare, but have had good success rate. I suggest first attempting tantalum piggybacks...

Just scrape away some soldermask next to the tokins and install a couple of tantalums on each side, maybe 1 or 2 to each CPU and GPU. You don't have to do as many as in the pic above. There's a decent chance this will fix your issue, for now at least.

I'm not keen on taking on modwork right now. I'm currently in more of a learning phase. I've been watching EEVblog videos, learning about Op Amps, VCOs, PLLs, PWM controllers, etc. When I say learning I mean blank staring at the screen hoping a light turns on. I've been reading the datasheets, cross-refrencing schematics, cataloging error codes...all in an attempt to better understand the power on sequence and what can go wrong in the process. It's basically a crash course in electrical engineering. I literally had to cool it this last week due to eyestrain.

It's starting to pay off though. This week I fixed another microscope at work. I thought it had a bad LED driver (a simple dimmer circuit consisting of a LM2575 DC/DC converter, LM358 OP Amp, a potentiometer and Min/Max trim pots). I reverse engineered the whole circuit before finding out that the wires connecting to the power rocker switch were dodgy...haha! I soldered them and it worked. Wasted a bunch of time barking up the wrong tree, but feel proud I was able to figure it out without paying thousands to a repair service. In case you're wondering, we put off this sort of work until there's down time. This week was slow, so it was a good time to putter away at the bad scopes (make this hobby useful to my employer). It has already saved us over $10,000 in maintenance/replacement costs. I usually check for loose wires first, so IDK why I missed that. Whetever, it works now.

It still feels like I have a long ways to go before gettiing how all of this works together to make the PS3 tick. But I feel like I understand the basics powering the board, which is alot more complicated than you would think. And the devil's in the details. Honestly I think at this point I'm going to have to start making assignments with the math to force myself to gain an intimate understanding of what's going on. Without running the numbers and watching the effect it's hard to grasp basic concepts and retain that knowledge. It's just some of the concepts are redicouous to wrap my head around. I've read and watched videos multiple times and still can't tell you what a pole on a Bode plot is. Then there's identities and calculus...CALCULUS!!! It's at the point I'm asking myself, "WTF am I doing?"

I have a problem!

I'm not skilled enough to attempt this unfortunately.

 

[RIP-Felix]

I'm not skilled enough to attempt this unfortunately.

Practice makes perfect. I believe in you!

 

[raidriar]

Practice makes perfect. I believe in you!

COK001 is too valuable for me to ruin. I have a scrap board to practice on but I still wouldn't be comfortable doing it on a working board

 

[RIP-Felix]

COK001 is too valuable for me to ruin. I have a scrap board to practice on but I still wouldn't be comfortable doing it on a working board

If that 1001 is related to tokins degradation it'll get worse over time and then you'll have nothing to lose when it eventually gets unbearable. Especially if you keep practicing until then, gaining confidence.

 

[moptop219]

What happened? Why did you remove the RSX? I thought you only had 1002 and 3004 on this console. That means you should keep going with the tokin replacments.

Did you get a 3034 in a post I missed?

Cooked the Syscon, then damaged a few other bits. So decided to practice removing the RSX on it.

 

[moptop219]

3013

BE_SPI DI/DO ERROR

CELL not communicating to syscon via SPI (1.2V MC2_VDDIO and 1.2V BE_VCS no output) = Possible shorts on the line, check C4001 and trailing caps. Possible dead CPU?

Another user had one on a CPU he damaged while deliding.

Can't measure the voltage on VDDIO, console doesn't run long enough. Checked all the original caps C4001 4005 4010 and the measurements seem fine. Can't find anything shorted.

 

[RIP-Felix]

Can't measure the voltage on VDDIO, console doesn't run long enough. Checked all the original caps C4001 4005 4010 and the measurements seem fine. Can't find anything shorted.

Can you list the voltages you checked and were good?

I am working on a flowChart that shows the voltage pathway, starting with the PSU and branching out to all the system voltages. The main ones are 12V_MAIN, 5V_MAIN, 5V_EVER, and 3.3V_MAIN. From those the rest are derived. So knowing what's good and what's not, eliminates the branches the problem can be on.

A step #21 also narrows things down some. 00-11 = Power Sequence Init, Basic Voltage/Clock Initialization. It's here I would expect an error with system voltages. 20-22 = CELL Initialization. That's when the CPU is first powered up and checked. So my guess i the problem is related to this...

I have a vague lead based on supposition and hunch.

On page 2/24 in the schematic there is a signal tree of +1.2V_MC2_VDDIO with a reference to the "P_L_BYPASS." My mind went to the PLL, which makes me think clock.

IC6012 (JL6063) --> +1.2V_MC2_VDDIO = BE_SPI, CHKSTP, JTAG, TBEN, & P_L_BYPASS
It gets powered by (IC6003/Q6601) --> 3.3V_MISC, so that needs to be present (which it should be, because the errors occurred at step #80 = Power On state. The Power On Sequence had completed successfully without error and +3.3V_MISC is one of the first checks). I thought I'd just mention it for the sake of completeness.

On page 18/24, IC5004 produces/monitors the reference clocks for the SB/CPU/RSX FlexIO Core (+1.2V_RC_VDDIO).

That's where the evidence dries up and previouse experiance comes in. @vyktormvmpay25 has said these errors tend to be and issue with the CPU's BGA. Since they gennerally point toward the CPU's side of the FlexIO, I'd agree that's a plausible conclusion. Remember the CPU's BGA can go bad too.

I'd check the voltages leading to IC5004 & IC6012. Check their respective enable/control Signal states (when do they go low). I'd measure the clock and see if the frequencies match expected (whatever that is. IDK yet). For example, the phase difference between the reference clock and PLL should be locked. But the error says they're not. That makes me suspect the feedback loop or loop filter.

I'm still super uneducated on PLLs and it's one of the MANY EE topics I've been researching lately.

All of this is part of my research project. I've already defined system voltages and created a voltage flowChart for the main ICs in the chain. Next up is to do the same for the Clock generators and Signal transduction pathways. Getting closer to presenting my findings - transcriptions really, all the information is already in the schematics. It's just hard to follow. But I'm not there yet. Especially in the area of clock generators or PLL.

For example, in that video he mentions the importance of the PLL filter in maintaining stability. I suspect PLL instability is a major suspect for producing your errors. That's why the refrence to "P_L_BYPASS" peaked my interest. It could be an area with deteriorating filtering caps (like the tokins were for VDDC). But again, My understanding of this is very preliminary and raw. And I'm grasping at straws.

 

[moptop219]

Can you list the voltages you checked and were good?

Sorry the capacitor values were good, not the voltages.

 

[DeadEnd]

Sorry if I'm interrupting something, but this video is important enough to be mentioned here as well. Here Microsoft is explaining the issue with their GPU. Go to around 19 minute. So I doubt it was much different for Sony's case.

 

[RIP-Felix]

Sorry if I'm interrupting something, but this video is important enough to be mentioned here as well. Here Microsoft is explaining the issue with their GPU. Go to around 19 minute. So I doubt it was much different for Sony's case.

I would like to see a video on Bumpgate itself. There's very little information about how widespread the issue was and what devices are affected. And the specific measures taken to prevent it, if in fact they have been. Was it eliminated? Or does it still affect FCBGA today, albeit mitigated to an "acceptable" margin. They mention it in passing, but I'm keenly interested to see it expanded into it's own video.

But let's not pretend the BGA is not a problem. We see BGA defects when lifting chips all the time, like black oxidized pads that have been exposed to the air for a long time and will not be fixed without a reball. Especially around the corners and edges, where the deformation stress is greatest.

 

[DeadEnd]

I would like to see a video on Bumpgate itself. There's very little information about how widespread the issue was and what devices are affected. And the specific measures taken to prevent it, if in fact they have been. Was it eliminated? Or does it still affect FCBGA today, albeit mitigated to an "acceptable" margin. They mention it in passing, but I'm keenly interested to see it expanded into it's own video.

But let's not pretend the BGA is not a problem. We see BGA defects when lifting chips all the time, like black oxidized pads that have been exposed to the air for a long time and will not be fixed without a reball. Especially around the corners and edges, where the deformation stress is greatest.

But there is information , plenty of devices were affected. A lot of laptop graphics cards for instance. All the links are here https://www.badcaps.net/forum/showthread.php?s=eeb96ac8863d4fe73243a4254155809a&t=94383

My take is that they have eventually eliminated it by hiring better packaging engineers, applying better underfill and improving bumpgate material.

Basically issue is a combination of thermal dissipation of the chip with hotspot, generating mechanical stress due to thermal expansion, wrong underfill material and wrong bump material.
It's not the solder balls between the chip and the board.

Chips affected are mostly from 2006-2008.
MCP and GPUs are affected, both in laptops and desktops.
Some articles list G86, G86A2, G84, C51, G72, G72M, G73, G72A3, MCP67 and NV42. But these aren't the only ones, some G9x and other MCPs were affected too.
That's some later GeForce 6000, almost all GeForce 7000, almost all GeForce 8000, some early GeForce 9000. Affected MCPs are MCP5x (northbridge only, not southbridge), MCP6x and early MCP7x.

 

[RIP-Felix]

But there is information , plenty of devices were affected. A lot of laptop graphics cards for instance. All the links are here https://www.badcaps.net/forum/showthread.php?s=eeb96ac8863d4fe73243a4254155809a&t=94383

That's just talking about Nvidea. I'm talking about how widespread the issue was in the industry, outside GPU's.

Even if they did mitigate bumpgate with better underfill, they can not eliminate it. There's still an interface between materials with different CTE's (Coefficient of Thermal Expansion). FCBGA does not have sufficient strain relief. The shearing, compression, and expansion forces are still there. They could go back to wiring bonding (PBGA, a wire can bend), but bumps offer higher pin density and lower manufacturing complexity = cheaper. So industry is fine with the decreased reliability, so long as it doesn't rise to the level of failures seen with the XBOX 360 and necessitate a recall.

Go back and watch that 360 video around the 4:15 mark. They said something like 40-50% of launch consoles manufactured before release were failing on the QA test bench. That's right off the line! Not consoles that went to customers and were thermal cycled until the bumps failed. No, these were factory soldering defects! So clearly there was an issue with the factory's FCBGA process, like reflow profiles, flux, ROHS solder chemistry, etc. The ones that didn't fail on the test bench went out to customers, where they quickly failed during normal use by their intended audiance. Just because they got past QA doesn't mean they had a good strong solder bond.

So there is a spectrum here. In the early days the bonds were weak and would fail 90% of the time in the first 3 years of use (a clearly defective and unacceptable product). Fast forward to today where the industry has perfected the process and we still see GPU failures on FCBGA within 5-10 years. When is FCBGA acceptable? Is 3 years the limit? 5? 10? 20? How long should a company be liable for repair or replace? The question is, "at what point to do you consider FCBGA to be an unreliable technology?"

It's not just me asking! Automotive, military, and space industries have standard testing procedures for electronics. Electronic manufacturers even grade their products to meet these standards. These industries require reliability for high stakes applications. If you're relying on a micro-controller to guide a satellite or ICBM, you can't afford a 5% failure rate! Can you imagine if 1 out of every 20 rockets had to be aborted because the guidance computer thought it would be a good idea to change the flight plan? And what if that rocket was carrying a nuclear payload? Now you just spread radiation over god knows who! Clearly for those industries, the reliability of the system is of paramount importance.

Let's say FCBGA is at the 10 year reliability mark today. Imagine your PS5 and XBOX Series X only lasts 10 years. Are you okay with spending $500 for that level of reliability? What if it only lasts 5 years?

Personally, I still enjoy my Dreamcast, OG XBOX, PS2, and GameCube. But those consoles would be long dead by now if they only had a 10 year reliability. I would say there's no problem, but only if the gaming industry provides a way for me to keep playing the games I paid for. I need Backwards compatibility on new consoles to replace the unreliable console they sold me (PS3). SONY did this with the launch PS3, but left us High and Dry with the PS4 (great console on it's own right, don't get me wrong). I'm pleased to see that PS5 has maintained BC with PS4 and announced the limited forward porting of PS3 titles available for purchase on PSN. Same with PS1 and 2 titles. It seems they finally understand the need for all of PlayStation being on one platform - the one for which new working hardware exists and that SONY can profit from. Bring PlayStation back to SONY. A job they inadvertently outsourced to the emulators by not providing a competing product.

MS has been better about this since the 360 debacle. It scared them strait! And the XBONE forced them to look inward to their roots. If they want to be here to stay, they have to cater to gamers! Gamers want reliable hardware to play THEIR GAME LIBRARY. The operative words implying personal ownership of video games. None of this digital only or TV BS! This generation's sales are proving this point. PS5/XBSX digital only versions are being sold in much fewer numbers than their Physical Disc Versions, despite being $100 cheaper. Ownership is important to gamers! Partly because we want to play all the games we bought for the OG, 360, one and seriex X. Backwards compatibility provides an instant library of games. But we need a disc drive to access that! Otherwise we have to buy a port. So the 100 cheaper console isn't really cheaper! The back catalog is expected to be part of the platform we're buying into! Ports are fine, even if we have to pay for it again, but it's a big pill only swallowed when easier developers come to the table with enhancements and achievements. The Halo Master Chief Collection is what I'm talking about! Come to the table, don't suck, and yeah...I'll consider paying again for it.

MS has their finger on the pulse of gamers this generation. BC and Gamepass are the killer features selling the Series X. Nintendo and SONY better take notice.

However, the entire console industry better screw their heads on strait and create forward portable platforms, because the Hardware limitations of FCBGA is going to undermine this trust moving forwards. New hardware must be able to replace failing PS5/XBSX's. My PS6 and NEXT-BOX better play what the current generation can. Otherwise, I'm not okay with FCBGA's being in the console. If you're going to put an unreliable technology in my $500 game console, you better commit to forward porting it's library!

The clear and present danger is the PC, which is perfectly poised to own this market. Microsoft's master plan, if you will. If gamers are tied to the HW, they're wasting their money. It'd be better spent on a PC, where they get to keep all their games and upgrade at will. That's where this is heading.

PC has already removed the biggest barrier to entry for console gamers - the keyboard and mouse. Remember when I said I'd consider buying the Halo Master Chief collection? Well, I did...for PC! The reason is because I can connect my wired USB XB360 controller and play a better version on PC. That and I got a big discount from...I forget if it was steam, ubisoft connect, GOG galaxy, EPIC games, origin, humble bundle...one of the many ways PC gamers get coupon codes, sales, and deep price breaks that console gamers don't. But I felt safer making that digital only transaction because I wasn't tied to console hardware. I feel like my digital games library on PC are more permanent, because traditionally they have been. PC has built that trust with gamers. Consoles have lost it! And now cross-play online allows PC/Console gamers to compete online together. What incentive do console gamers have to buy a console? PC is just as easy to jump into now. All the barriers to entry, that kept console gamers from becoming PC gamers have been systematically dismantled by PC. Soon gamers will realize consoles are a bad value proposition.

And at the end of this Microsoft will be in a perfect position to benifit from the death of the video game console. SONY will have no chance to compete with windows. This is a giant middle finger to Nintendo who refused to entertain MS's buyout proposal. I think that Nintendo has played the console game best. They correctly identified what makes console better then PC - convenience. The switch is the embodiment of convenience. Merging on the go with the big screen is the step forward the console needed to compete with the direction MS is taking. SONY on the other hand has chosen the perilous strategy to try and compete with MS, a battle they can't win. At the end of the day, I think SONY is the next SEGA.

Wow, that took a turn. Started talking about bumpgate and ended with consolegate. Whatever, I like to rant. Go ahead, tell me it's a personal failing. I'm aware and don't care.

 

[DeadEnd]

That's just talking about Nvidea. I'm talking about how widespread the issue was in the industry, outside GPU's.

Even if they did mitigate bumpgate with better underfill, they can not eliminate it. There's still an interface between materials with different CTE's (Coefficient of Thermal Expansion). FCBGA does not have sufficient strain relief. The shearing, compression, and expansion forces are still there. They could go back to wiring bonding (PBGA, a wire can bend), but bumps offer higher pin density and lower manufacturing complexity = cheaper. So industry is fine with the decreased reliability, so long as it doesn't rise to the level of failures seen with the XBOX 360 and necessitate a recall.

Go back and watch that 360 video around the 4:15 mark. They said something like 40-50% of launch consoles manufactured before release were failing on the QA test bench. That's right off the line! Not consoles that went to customers and were thermal cycled until the bumps failed. No, these were factory soldering defects! So clearly there was an issue with the factory's FCBGA process, like reflow profiles, flux, ROHS solder chemistry, etc. The ones that didn't fail on the test bench went out to customers, where they quickly failed during normal use by their intended audiance. Just because they got past QA doesn't mean they had a good strong solder bond.

So there is a spectrum here. In the early days the bonds were weak and would fail 90% of the time in the first 3 years of use (a clearly defective and unacceptable product). Fast forward to today where the industry has perfected the process and we still see GPU failures on FCBGA within 5-10 years. When is FCBGA acceptable? Is 3 years the limit? 5? 10? 20? How long should a company be liable for repair or replace? The question is, "at what point to do you consider FCBGA to be an unreliable technology?"

It's not just me asking! Automotive, military, and space industries have standard testing procedures for electronics. Electronic manufacturers even grade their products to meet these standards. These industries require reliability for high stakes applications. If you're relying on a micro-controller to guide a satellite or ICBM, you can't afford a 5% failure rate! Can you imagine if 1 out of every 20 rockets had to be aborted because the guidance computer thought it would be a good idea to change the flight plan? And what if that rocket was carrying a nuclear payload? Now you just spread radiation over god knows who! Clearly for those industries, the reliability of the system is of paramount importance.

Let's say FCBGA is at the 10 year reliability mark today. Imagine your PS5 and XBOX Series X only lasts 10 years. Are you okay with spending $500 for that level of reliability? What if it only lasts 5 years?

Personally, I still enjoy my Dreamcast, OG XBOX, PS2, and GameCube. But those consoles would be long dead by now if they only had a 10 year reliability. I would say there's no problem, but only if the gaming industry provides a way for me to keep playing the games I paid for. I need Backwards compatibility on new consoles to replace the unreliable console they sold me (PS3). SONY did this with the launch PS3, but left us High and Dry with the PS4 (great console on it's own right, don't get me wrong). I'm pleased to see that PS5 has maintained BC with PS4 and announced the limited forward porting of PS3 titles available for purchase on PSN. Same with PS1 and 2 titles. It seems they finally understand the need for all of PlayStation being on one platform - the one for which new working hardware exists and that SONY can profit from. Bring PlayStation back to SONY. A job they inadvertently outsourced to the emulators by not providing a competing product.

MS has been better about this since the 360 debacle. It scared them strait! And the XBONE forced them to look inward to their roots. If they want to be here to stay, they have to cater to gamers! Gamers want reliable hardware to play THEIR GAME LIBRARY. The operative words implying personal ownership of video games. None of this digital only or TV BS! This generation's sales are proving this point. PS5/XBSX digital only versions are being sold in much fewer numbers than their Physical Disc Versions, despite being $100 cheaper. Ownership is important to gamers! Partly because we want to play all the games we bought for the OG, 360, one and seriex X. Backwards compatibility provides an instant library of games. But we need a disc drive to access that! Otherwise we have to buy a port. So the 100 cheaper console isn't really cheaper! The back catalog is expected to be part of the platform we're buying into! Ports are fine, even if we have to pay for it again, but it's a big pill only swallowed when easier developers come to the table with enhancements and achievements. The Halo Master Chief Collection is what I'm talking about! Come to the table, don't suck, and yeah...I'll consider paying again for it.

MS has their finger on the pulse of gamers this generation. BC and Gamepass are the killer features selling the Series X. Nintendo and SONY better take notice.

However, the entire console industry better screw their heads on strait and create forward portable platforms, because the Hardware limitations of FCBGA is going to undermine this trust moving forwards. New hardware must be able to replace failing PS5/XBSX's. My PS6 and NEXT-BOX better play what the current generation can. Otherwise, I'm not okay with FCBGA's being in the console. If you're going to put an unreliable technology in my $500 game console, you better commit to forward porting it's library!

The clear and present danger is the PC, which is perfectly poised to own this market. Microsoft's master plan, if you will. If gamers are tied to the HW, they're wasting their money. It'd be better spent on a PC, where they get to keep all their games and upgrade at will. That's where this is heading.

PC has already removed the biggest barrier to entry for console gamers - the keyboard and mouse. Remember when I said I'd consider buying the Halo Master Chief collection? Well, I did...for PC! The reason is because I can connect my wired USB XB360 controller and play a better version on PC. That and I got a big discount from...I forget if it was steam, ubisoft connect, GOG galaxy, EPIC games, origin, humble bundle...one of the many ways PC gamers get coupon codes, sales, and deep price breaks that console gamers don't. But I felt safer making that digital only transaction because I wasn't tied to console hardware. I feel like my digital games library on PC are more permanent, because traditionally they have been. PC has built that trust with gamers. Consoles have lost it! And now cross-play online allows PC/Console gamers to compete online together. What incentive do console gamers have to buy a console? PC is just as easy to jump into now. All the barriers to entry, that kept console gamers from becoming PC gamers have been systematically dismantled by PC. Soon gamers will realize consoles are a bad value proposition.

And at the end of this Microsoft will be in a perfect position to benifit from the death of the video game console. SONY will have no chance to compete with windows. This is a giant middle finger to Nintendo who refused to entertain MS's buyout proposal. I think that Nintendo has played the console game best. They correctly identified what makes console better then PC - convenience. The switch is the embodiment of convenience. Merging on the go with the big screen is the step forward the console needed to compete with the direction MS is taking. SONY on the other hand has chosen the perilous strategy to try and compete with MS, a battle they can't win. At the end of the day, I think SONY is the next SEGA.

Wow, that took a turn. Started talking about bumpgate and ended with consolegate. Whatever, I like to rant. Go ahead, tell me it's a personal failing. I'm aware and don't care.

Honestly 10 years of reliability isn't bad. But I'd take it if it was 10 years minimum and after that all bets are off.

Spoiler: PC rant

I think you are being a little bit of pro PC here. This isn't the place to talk about that. I personally turn off cross play, because I don't need mouse/keyboard players to aim at me with perfect reaction and precision. I think Sony consoles are not going to be that easy to kill off. I personally don't like PC gaming because I use PC for work, and I have to stare at a PC screen at work also. It's just not enjoyable for me to come home and sit in a chair, and then end up staring at yet another tiny PC monitor for gaming. I don't want to be bothered to hook up wires to my TV either. I just plop down on a sofa and turn on the console... It's something simple. It's also simple enough for most people, unlike PC. There is extra work on computers. You need to update drivers for instance. Some games might not run properly. Controller support isn't perfect. Sometimes when I hooked up Xbox one controller to a PC, there was something off about the movements. It didn't have the same fluency as consoles. That just gave me headaches... Windows could glitch. So there are a lot of extra small things that could happen with a PC. With consoles, you are in a way protected from that.

 

[RIP-Felix]

Honestly 10 years of reliability isn't bad. But I'd take it if it was 10 years minimum and after that all bets are off.

Spoiler: PC rant

I think you are being a little bit of pro PC here. This isn't the place to talk about that. I personally turn off cross play, because I don't need mouse/keyboard players to aim at me with perfect reaction and precision. I think Sony consoles are not going to be that easy to kill off. I personally don't like PC gaming because I use PC for work, and I have to stare at a PC screen at work also. It's just not enjoyable for me to come home and sit in a chair, and then end up staring at yet another tiny PC monitor for gaming. I don't want to be bothered to hook up wires to my TV either. I just plop down on a sofa and turn on the console... It's something simple. It's also simple enough for most people, unlike PC. There is extra work on computers. You need to update drivers for instance. Some games might not run properly. Controller support isn't perfect. Sometimes when I hooked up Xbox one controller to a PC, there was something off about the movements. It didn't have the same fluency as consoles. That just gave me headaches... Windows could glitch. So there are a lot of extra small things that could happen with a PC. With consoles, you are in a way protected from that.

Spoiler: More PC Stuff

True, cross-play with people using a Mouse in FPS is not fair. But they can build in controls to exclude mouse users if you don't want to play with them. I feel it prudent to point out that there are 3rd party mice of consoles. So there are mice users on console too.

In any case, PC gaming is way easier now than it was 10 years ago. I was and am a staunch console guy. But I've been converted to PC as well. It's truely an awesome platform! I greatly appreciate MS commitment to releasing games to both series x and PC, so gamers can choose how to experience them. Steam updates automatically. Nvidea Game ready drivers prompt you when there's an update and all you have to do is click a couple of buttons to DL and Install. Sure windows can be buggy, but at least I can choose when my PC installs updates, not have to wait 10minutes every time I boot up my console!

About controllers, PC gamers have options console gamers don't. Preferr the DualShock 4? There's a solution for that. Sure it involves more setup, but you don't even get the option on console. You're stuck with the controllers given. I get the advantage of curated experience, where all of the guesswork is taken away. That's why I use a wired XB360 controller. USB is low lag and the 360 controller is windows de-facto standard. It's automatically recognized and setup without fuss. And I have an assortment of adapters to allow me to use OG retro controllers for emulators. Although they do take setting up to get working first.

About PC monitors, I don't use them except at work. I use an HTPC in my entertainment system hooked up to my HDTV. I custom built it with an IR receiver and programmed a Logitech Harmony remote. All I have to do is press "HTPC" on my remote and it starts the HTPC, TV, and changes to the correct HDMI input. I also use a front-end (GameEX) that launches on startup. It list all my emulators and PC games in one awesome interface. I can launch them with the press of a button. It also has hooks for KODI or Plex, to do the same for movies/music/TV. It took plenty of setup to get working, sure. But it beats the pants off anything console's offer. In almost every way imaginable, the PC and emulation community have improved the experience beyond what consoles offer.

The only thing they do better is having it all setup out of the box.

But my point was that MS is the only one who stands to benefit from PC gaming stealing market share from consoles.

 

[mekmitasdigoat]

Well, I can't rule out BGA wear, but you don't have the tell tale 3034/4xxx errors yet. So I wouldn't worry about it until it happens.

A random, intermittent YLOD was the use case for this mod in the first place. That's the type of YLOD the OP originally reccomended this fix for. They're pretty rare, but have had good success rate. I suggest first attempting tantalum piggybacks...

Just scrape away some soldermask next to the tokins and install a couple of tantalums on each side, maybe 1 or 2 to each CPU and GPU. You don't have to do as many as in the pic above. There's a decent chance this will fix your issue, for now at least.

I'm not keen on taking on modwork right now. I'm currently in more of a learning phase. I've been watching EEVblog videos, learning about Op Amps, VCOs, PLLs, PWM controllers, etc. When I say learning I mean blank staring at the screen hoping a light turns on. I've been reading the datasheets, cross-refrencing schematics, cataloging error codes...all in an attempt to better understand the power on sequence and what can go wrong in the process. It's basically a crash course in electrical engineering. I literally had to cool it this last week due to eyestrain.

It's starting to pay off though. This week I fixed another microscope at work. I thought it had a bad LED driver (a simple dimmer circuit consisting of a LM2575 DC/DC converter, LM358 OP Amp, a potentiometer and Min/Max trim pots). I reverse engineered the whole circuit before finding out that the wires connecting to the power rocker switch were dodgy...haha! I soldered them and it worked. Wasted a bunch of time barking up the wrong tree, but feel proud I was able to figure it out without paying thousands to a repair service. In case you're wondering, we put off this sort of work until there's down time. This week was slow, so it was a good time to putter away at the bad scopes (make this hobby useful to my employer). It has already saved us over $10,000 in maintenance/replacement costs. I usually check for loose wires first, so IDK why I missed that. Whetever, it works now.

It still feels like I have a long ways to go before gettiing how all of this works together to make the PS3 tick. But I feel like I understand the basics powering the board, which is alot more complicated than you would think. And the devil's in the details. Honestly I think at this point I'm going to have to start making assignments with the math to force myself to gain an intimate understanding of what's going on. Without running the numbers and watching the effect it's hard to grasp basic concepts and retain that knowledge. It's just some of the concepts are redicouous to wrap my head around. I've read and watched videos multiple times and still can't tell you what a pole on a Bode plot is. Then there's identities and calculus...CALCULUS!!! It's at the point I'm asking myself, "WTF am I doing?"

I have a problem!

Hey Felix, I have a CECHA suffers from the stress YLOD(shuts off under high loads) with the confirmed 801002 error code. I'm going to try the piggyback way to fix it.
I'm wondering what is the part number of these caps you used for TOKIN cap piggyback? Are they 470uF 2.5v 7mΩ Panasonic 2R5TPE470M7 TaPol caps? Also if I remembered correctly, in this thread somewhere you (maybe others, sorry I forget) mentioned it's better to use AlPol as they are same material as stock TOKINs? Does it matter actually?

Thanks.

 

[Pooka]

Spoiler: tl;dr pc is wasted potential

Speaking of consolegate, PC is basically too much wasted potential. The platform can be good but I need one thousand accounts on one thousand services and to download one thousand blatant DRM launchers to play one or two singleplayer games. Can't even play them straight away, must first launch the Epic Steam Store and stay online or use an extremely unreliable offline "mode" that can break randomly over time (the naming alone implies it's second class trash).

And since one of these godforsaken launchers have launched (heh) an attack against physical PC games, I can find barely any physical PC games anywhere, not even used, but PS3 used games are so cheap and common it's not even funny anymore. I can buy games for a fraction of their ON-SALE PRICE on digital PC stores, and just pop them into a console and play right away. No DRM, no accounts, and I get a box with a manual as a bonus.

Not to mention even when you have a dual monitor setup and try to alt-tab out of the game to do work, it minimizes anyway so it's actually painful to try and get a game working side-by-side with a workflow of sorts. So even the PC's greatest strength, multi-tasking, is hit or miss when you try to play games. I know the PC can do 4196p360fps and mod games to high hell, but if I have to jump through so many hoops imposed by useless launchers that act as DRM and have no ability to buy used games, then maybe, I think, the PC platform is just a sad showcase of wasted potential. It could be so great if it wasn't for all the shackles.

Admittedly though, the PC is very excellent when it comes to backwards compatibility, and given the ranting about launchers, GOG is at least a decent enough storefront to mandate NO launchers or DRM for singleplayer games, so there's something at least. On the other hand, DeadEnd makes a good point there, I'm also less inclined to play games on the same PC I worked on for several hours already.

That's just my two cents, I like reading about the technical aspects of the PS3, so keep on the good work there!