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

[elsam]

So I've bought 2 broken PS3 phat for about $ 40 from local facebook marketplace:
- CECHG - 2007 (65nm CELL, 90nm RSX, moterhboard: SEM-001)
- CECHJ - 2008 (65nm RSX, 65nm RSX, motherboard: DIA-002)

Both have YLOD, the first (2007) has instant ylod and the other (2008) has a delay of 3-4 seconds before ylod.
I decided to fix the 2008 version first by replacing 2 "108 NEC TOKIN" on the rsx side (yes this phat model uses 0E 108 on rsx) with 6 tantalum 470uf 6v on the bottom side of the motherboard where there is an RSX and CELL processor but still delayed YLOD.

After that I decided to replace 2 "128 NEC TOKIN" on the CELL side with 6 pieces of 470uf 6v tantalum but still delayed YLOD, so i decided to replace a few more NEC TOKIN on the other side of the motherboard, but because replacing process is painful, i decided to use piggy-back method on the RSX side by adding two tantalum capacitor 470, but guess what happens, it's now instant YLOD lol.

So I removed the two 470 capacitors and decided to replace the NEC TOKIN like I did on the other side of the motherboard,
but when I destroy the shell of NEC TOKIN, I see this NEC TOKIN capacitor looks like charred, is it because I use the piggy back method with a voltage value of 6 volts or it was already burnt when I bought it, I don't know, I continued to replace it with three 470 tantalums.

My calculations:
3 x 0E 108 = 3000uf replaced by 470 x 9 = 4230uf

So I think it's more than enough to keep the last NEC TOKIN on the RSX side without changing it because I don't have any jumper cables either, and after I turned it on the instant YLOD was gone and it returned to be delayed YLOD.

After that I decided to check the last NEC TOKIN on the RSX side to see if it burned or not, and when I opened the shell, this capacitor looked charred like the previous capacitor, so I took it off and tried turning on the console without jumper wires, and instant YLOD again.

My question:
1. Will the piggy-back method using a capacitor with a higher voltage value damage the NEC TOKIN capacitor ?
2. Do I have to replace the 2 remaining NEC TOKIN in CELL in order for the console to boot up? Because what I read, you only need to replace the NEC TOKIN on the rsx side to fix the YLOD problem (or is not?)
3. Since I'm a very noob about jumpers, I want to ask what kind of jumper cable do you guys use? Where can I get it? Are there specific specifications you should look for? Because I couldn't find any specific discussion on this topic.

Thanks.

 

[MYS]

So I've bought 2 broken PS3 phat for about $ 40 from local facebook marketplace:
- CECHG - 2007 (65nm CELL, 90nm RSX, moterhboard: SEM-001)
- CECHJ - 2008 (65nm RSX, 65nm RSX, motherboard: DIA-002)

Both have YLOD, the first (2007) has instant ylod and the other (2008) has a delay of 3-4 seconds before ylod.
I decided to fix the 2008 version first by replacing 2 "108 NEC TOKIN" on the rsx side (yes this phat model uses 0E 108 on rsx) with 6 tantalum 470uf 6v on the bottom side of the motherboard where there is an RSX and CELL processor but still delayed YLOD.

After that I decided to replace 2 "128 NEC TOKIN" on the CELL side with 6 pieces of 470uf 6v tantalum but still delayed YLOD, so i decided to replace a few more NEC TOKIN on the other side of the motherboard, but because replacing process is painful, i decided to use piggy-back method on the RSX side by adding two tantalum capacitor 470, but guess what happens, it's now instant YLOD lol.

So I removed the two 470 capacitors and decided to replace the NEC TOKIN like I did on the other side of the motherboard,
but when I destroy the shell of NEC TOKIN, I see this NEC TOKIN capacitor looks like charred, is it because I use the piggy back method with a voltage value of 6 volts or it was already burnt when I bought it, I don't know, I continued to replace it with three 470 tantalums.

My calculations:
3 x 0E 108 = 3000uf replaced by 470 x 9 = 4230uf

So I think it's more than enough to keep the last NEC TOKIN on the RSX side without changing it because I don't have any jumper cables either, and after I turned it on the instant YLOD was gone and it returned to be delayed YLOD.

After that I decided to check the last NEC TOKIN on the RSX side to see if it burned or not, and when I opened the shell, this capacitor looked charred like the previous capacitor, so I took it off and tried turning on the console without jumper wires, and instant YLOD again.

My question:
1. Will the piggy-back method using a capacitor with a higher voltage value damage the NEC TOKIN capacitor ?
2. Do I have to replace the 2 remaining NEC TOKIN in CELL in order for the console to boot up? Because what I read, you only need to replace the NEC TOKIN on the rsx side to fix the YLOD problem (or is not?)
3. Since I'm a very noob about jumpers, I want to ask what kind of jumper cable do you guys use? Where can I get it? Are there specific specifications you should look for? Because I couldn't find any specific discussion on this topic.

Thanks.

I think (based on wat i had done so far on 4 PS3), it as higher chance of working if you replace back using NEC Token. Direct tantalum cap might not work on these newer version of PS3 as the current might not achieve what is required by the processor.

Just buy some Token off aliexpress and some solder paste. Using a heat gun, the NEC token can be easily reflow ("solder") onto the board.

 

[slamz369]

im wondering is it possible to only replace 4 of the NEC's on the opposite to the chips side (because its easier to get to) with 20 or 22 470uf to cover all NEC's just incase?
maybe 10 each with 200uf to fill out the rest?

ive ordered some 470uf's and will be testing it when they come, but wondering if its a good idea or not?

 

[MYS]

im wondering is it possible to only replace 4 of the NEC's on the opposite to the chips side (because its easier to get to) with 20 or 22 470uf to cover all NEC's just incase?
maybe 10 each with 200uf to fill out the rest?

ive ordered some 470uf's and will be testing it when they come, but wondering if its a good idea or not?

I think will depend on your model's processor. For my CECHH (SEM mainboard), I ad tried replacing tantalum (tried different combination of full 470uf, full 330uf, mix 470uf and 330uf) on one side but still YLOD. I remove all Token and try again with different combination of tantalum but still didn't work. But replacing back the Token bought the system back.

But didn't have this issue with my much older backward compatible PS3. These system came back to life after changing 1 side of the Token using 470uf. But after it work, i replace both side and it is still working till today. But based on the CECHH result, i think i gonna find time to replace back the Token incase the processor gets "burn" out.

 

[elsam]

I think (based on wat i had done so far on 4 PS3), it as higher chance of working if you replace back using NEC Token. Direct tantalum cap might not work on these newer version of PS3 as the current might not achieve what is required by the processor.
Just buy some Token off aliexpress and some solder paste. Using a heat gun, the NEC token can be easily reflow ("solder") onto the board.

Hi, thanks for the reply

Using the Nec Tokin again, which is the source of the YLOD problem, is likely to only bring trouble back in the future. But I'll try it later when everything else fails.

I think will depend on your model's processor. For my CECHH (SEM mainboard), I ad tried replacing tantalum (tried different combination of full 470uf, full 330uf, mix 470uf and 330uf) on one side but still YLOD. I remove all Token and try again with different combination of tantalum but still didn't work. But replacing back the Token bought the system back.

But didn't have this issue with my much older backward compatible PS3. These system came back to life after changing 1 side of the Token using 470uf. But after it work, i replace both side and it is still working till today. But based on the CECHH result, i think i gonna find time to replace back the Token incase the processor gets "burn" out.

Btw, Just curious about the tantalum capacitor you are using, is it a Panasonic 2R5TPE470M9 ?

Because from what I've read in this thread, choosing a right capacitor that is close to the specifications of the Nec Tokin can help eliminate YLOD.

and also..

Hiya, from my experience when i replaced the 8 x old NECs, with 8 x "new" NECs, at first i wanted to test with just half to see, and i still had YLOD, it wasn't until i had installed all 8 x "new" NECs that the PS3 worked again.

When i measured the "new" NECs they were like 1400uF to 1600uF (approx)... so really each processor is getting 6000uF. When i did my tests with the Tantalums + MLCCs (as can be seen in my post above), i tired 1st with only (approx) 4800uF on each processor and i had YLOD, only when i installed 7000uF+ (16 x 470uF) per processor, did the PS3 work again, anything less wouldn't work at all, or would work only 30% of the time, the more capacitors = the more it worked.

Tantalum capacitors genereally have a 10 - 20% tolerance, but if they come from China, like some i played around with, oh man - some 470uFs (20%) were actually reading at 600 - 700uF+ !! So unless the capacitors are of high quality and sourced from authentic places, then aiming for 4800uF per processor won't be enough with cheap caps. So aim for 6000 - 7000uF per processor and that should help even with cheap chinese capacitors.

But... even with quality capacitors, i would still aim for 6000uF per processor.

So @aminiani, if you replace only 2 NECs with Tantalums, and the other 2 NECs that you leave are the faulty ones, then you will likely only have approx' 3760uF (8 x 470uF) in total per processor, that's not enough, it's like another 3500uF too short to work, as mentioned above.

Also, if the internal bridging wire of the NECs you left on the board is faulty, then this too will result in YLOD with your present attempt / set-up. Because potentially no current is getting across to the processors without correct bridge wires. But then adding a bridge wire with a full Tantalum replacement can be problematic. Please see my photos above from my latest post, and how the capacitors are only on the bridge wire, otherwise adding a "raw" bridge wire in parallel to the tantalum capacitors can potentially cause some, or most of the current to get re-routed, and not get filtered properly by the capacitors, before then arriving at the processors.

So in short, total capacitance of 6000uF - 7000uF+ per processor, lower end if using high quality capacitors, and higher end if using cheap chinese capacitors. If a person ends up replacing all the NECs with Tantalums, then they are forced to consider the bridging wire(s), and how they will go about installing them correctly to not get YLOD.

i am interesting to test this.. So i am planning to order this panasonic capacitor on aliexpress.

 

[MYS]

Hi, thanks for the reply

Btw, Just curious about the tantalum capacitor you are using, is it a Panasonic 2R5TPE470M9 ?

Because from what I've read in this thread, choosing a right capacitor that is close to the specifications of the Nec Tokin can help eliminate YLOD.

and also..

.

Yup I am using this 2R5TPE470M9.

 

[Helinton Lunardi]

Hello, i was traying to put eletrolic capacitors, unsucessful some times it even turn on but not for long, so, i bought some tantalum capacitors amd replace the 4 nec in the up board, now it works in menu, but when i try to run a game or a app like netflix it just keeps black screen, i put 3 tantalum capacitors for each nec, 128 and 108, should i put more? Like 4 in the nec 128 area? Or keep chaging the ones in the other side of the board?

 

[RIP-Felix]

PS3 #1 - Felix's First Post...Aww
(...continued from SHUMPS forum.)​

I apologize in advance, this is going to be a long post:

The TL;DR:

• Anyone who says this is a 100% fix is wrong.
  
• Reballing is not a con.
  
• The NEC/TOKINs do fail.
  
• Low ESR of Tantalum caps matters.
  
• The current across the caps is no joke. More jumpers = safer.
  
• Total capacatitanc might mater, but dosen't "appear" to.
  
• Learn to properly Delid your CPU, make a tool, and do it before you turn your PS3 on again!!!
  
• Jailbreak and install webMAN for tighter fan control.

I have been investigating this repair myself for over a year. You can read about my adventure here. I am a hobbyist, not an electrical engineer! However, I have been diving deep into EE videos and technical documentation to try and understand what is really going on. To summarize my findings and experience:

PS3 #1 (CECHA01) has been through hell and is currently dead with a hard YLOD. However, before that the tant cap fix worked multiple times, and failed quickly afterwards due to various soldering mistakes (cold solder joints, not enough jumpers, flux/jumper wires causing bridging, not deliding the CPU IHS which was overheating and complicating things, etc.) In the process I learned how to solder properly to the motherboard and about the RLC circuit the NEC/TOKIN caps are part of.

• From carefully reading the schematic, this is an RLC power filter/decoupling circuit designed to provide derippled, decoupled, a clean +1.0v (CPU) and +1.2v (RSX). Here is a simplified schematic I made to make it easier to read, but it's not 100% accurate.
  
• The general rule is that using more capacitance if fine (as most of us have been assuming is the case). However, in RLC circuits the capacitance is important if you want to Pass or Stop a specific band (frequency range). Therefore it's possible that SONY engineers chose 4800uF for a reason. The RLC filter is needed because of the DC-DC switching voltage regulators that step the +12v main to 1-1.2v the chips drink. These regulators introduce noise that must be removed. It's most efficient to do so as close to the chip as possible, which explains the proximity of the NEC/TOKIN and the ceramics (CE in the pic above) to the chips. The RLC circuit SONY chose to employ does this very efficiently, but because of the inductor, it introduces resonance. Basically, capacitance could be important. I have not confirmed if the resonant frequency obtained by using 4800uF capacitance is important or not, and not sure how to. However, to remove the variable it is advisable to just match a replacement spec for spec.
  
• The amount of current that flows across these caps is large. I found this out the hard way (see pic below). A single AWG22 solid core conductor was inadequate and shorted! I also had trouble with shorting when they laid flat, because they ran over the GND rails. They would heat up, carbonize and short. So I had to go over the top, such that it was impossible to short. Keep this in mind. Below are pics from PS3#1, which didn't survive all the failed attempts and shorts.

• Something odd happened with PS3#1 above after all these repair attempts. The YLOD disappeared! Great, but artifacting would freeze the console within a few minutes of startup. That indicated an RSX issue that probably requires reball to fix. I attempted a reflow and it resulted in a GLOD! I attempted it again with more heat and proper flux, but the board warped on me and resulted in a hard YLOD. I intend to try a reflow and/or reball of both the CPU and RSX, but given how much the board warped, I doubt PS3#1 will survive. I'm still waiting on supplies and may revist it later, but my focus shifted to PS3#2.
  

PS3 #1 Continued here...

PS3 #2 (A01)​

• ESR is an important consideration in filter circuits. The lower the better. 4x NEC/TOKINs have a combined ESR of 0.375mOhms, so that's the target you should be looking to match. I found that 18x 270uF, 2.5v, 6mOhm Tantalum capacitors have the same total capacitance and a lower ESR. Therefore they are ideal for a CECHA01 as 1:1 replacement. I also chose these for their size, so that they are easier to solder in place, look clean, and can have the OG thermal tape replaced afterwards. I used hot air (380C) in my left hand to preheat the area while soldering the caps in using my iron (340C) in the right. I found it easier to use Kapton tape to hold the Cap in place. I installed 8x of these in place of one NEC/TOKIN for each the CPU and RSX on PS3 #2 (CECHA01), which was a hard YLOD. This console had not been previously worked on and appeared to be stock inside. I decided to leave 3x NEC/TOKINs in place, so I would not need to add jumpers. I delided the CPU/RSX and applied MX-4 TC. This fixed the YLOD and has been very stable about 24 hours of on-time. I jail broke it so I could install webMAN mod and see the temps (67C CPU, 58C RSX, Fan 33%, set point 68C). It's well controlled now and I'm confident that it's going to last, but that remains to be seen.

• Please note that the total capacitance that fixed PS3#2 is not known because I left the OG caps in place. 2160uF of added capacitance was adequate for this console, but if I were to replace all the NEC/TOKINs I'd have used 18 total per chip to match the OG spec. I didn't need to and adding jumpers is potentially problematic. Many people have has success going over 4800uF total capacitance, so it "apparently" doesn't matter.
• Lastly, you can't uses continuity buzzer to check your work. The normial resistance between +/GND rails is below the threshhold for most multimeters. I have measured it between 2.5-3.5ohms when the caps are properly installed. Be sure to thoroughly clean the area of flux after install, as I have noticed flux residues can decrease that resistance and cause a short.

I want to offer a hypothesis about the YLOD progression, based off my experience and that of others I've read:

1. After 2 years, SONY's cheap thermal grease begins to dry. An air gap forms in-between the IHS and the Die on the CPU and/or RSX. Since this is under the heat spreader, it's not easy to get to. Most people do not replace it. People usually only replace the thermal compound between the IHS and the Heatsink, but that's only half of the solution. It's helps lower temps, but doesn't prevent overheating. BOTH CPU and RSX need to be delided!
   
2. You start to notice the fan gets noisier. The SYSCON will pretty much allow the CPU/RSX to get to about 75C before the fan ramps into high gear. Above 70C is bad, so SONY's default fan control scheme is terrible. They obviously prioritized sound over longevity. The only way to change this is to Jailbreak the PS3. And it needs to be done. But if you don't change the thermal paste to restore thermal contact between the IHS and the die, then the chips overheat and the fan tries it's best to remove the heat. Of course this is a loosing battle because the chip is becoming increasingly insulated. I noticed this on my working PS3. After a few minutes the fan would ramp up to a noisy level. It still worked, so I assumed it was just the Hot running PS3 we all know and love. But that's not it. It's your fair warning of overheating! I delided the RSX, because it was easy, and not the CPU, because it was hard. I just delided it this week and the thermal greases was completely dried up. There was no direct contact anymore! This needs to be replace on every PS3 that hasn't had it done!
   
3. The excess heat places unnecessary strain on the NEC/TOKINs causing them to fail faster. They die and cause a YLOD. Or thermal cycling at temperatures above 75C places unnecessary stress on the BGA and causes fatigue fractures leading to a YLOD, before the Caps fail. Based on two consoles I've personally replace caps on, which is too small a sample size to draw definitive conclusions, I would say that the caps fail first more often.
   
4. People send their console in for a reball, which repairs the BGA. If the cause of the YLOD was the caps, the heat from the process often restore function to the NEC/TOKINS temporarily before they die again. This could explain some of the reballed consoles that YLOD again within a matter of months. Some repair shops don't remove the IHS during the reball. First this means it takes longer for them to heat up and be removed, which is harder on the chips and motherboard. Second, they did not fix what was causing the chips too overheat in the first place. So they will again overheat and cause another YLOD soon after the reball. Lastly, the heat that shorten the life of the NEC/TOKINS may not have killed them before the solder balls cracked and caused the YLOD. A reball my fix the console and it will run until the NEC/TOKINS finally die (which they are prone to anyway). This accounts for the rest.
   
5. Fed up, people sell their PS3 cheap on e-bay.
6. We install a tantalum caps to repair the consoles that do not have damaged solder balls. Deliding is still necessary to prevent excess heat and thermal stress that would eventually cause BGA damage. Combining these with a reball may provide a longer lasting repair, but placing that much heat on chips will shorten their life. So a reball should be avoided. The best way is to delid and cool the chips as soon as temps start to rise out of control. And a jailbreak with webMAN mod to monitor temps and tighten up the fan controls will help immensely.

PS3 #2 Continued here...

 

[cuttrogue]

Felix, great write up, but see my post: https://www.psx-place.com/threads/a-final-and-definitive-answer-for-yearly-ps3-yold-problems.31303/

The BGA is not the problem in the majority of the cases, reballing is a waste of resources. The problem is not the chip, it is the die itself.

 

[RIP-Felix]

You're right of course. There's nothing you can really do to fix a die failure on the Flip-chip BGA. The solder balls between die and substrate are to numerous and tiny to replace without automation. That's why I say it's better not to reball/reflow if it can be avoided. The heat is likely to hasten the failure of the die itself, and that's really the end. Of course that will happen anyway, it's just a matter of time. It's a reality of FCBGA technology that the industry has adopted. They'd rather make something that dies in 10 years forcing an "upgrade" than make something they have to compete with for 25-30 years.

I much prefer the days of passive heatsinks and thermal pads.

 

[pasdebip]

Hello
I've been following this post for a while and I'm wondering why multiply tantalum rather than taking 2 tantalum one 1000µF and the other 470µF while keeping the same type of case and low ESR?
This will limit bad welds.

 

[RIP-Felix]

Working with an IR preheater or hot air to preheat the area will limit bad welds.

To answer your question, the smaller the capacitance the easier it is to find sufficiently low ESR caps. Low ESR is important for better efficiency and heat performance. 3.5mm caps don't require removing solder mask or soldering at an angle, but you can't find them above 330uF or perhaps 470uF. Also, ESR follows the Parallel Resistance Equation only for capacitors of equal value.

NEC/TOIKINs for example:
= 1 / [(1/R)+(1/R)...]
= 1 / [(1/1.5mΩ)+(1/1.5mΩ)...]
= 1/(4/1.5mΩ)
= 4800uF 0.375mΩ

So for example an equivalent ESR using 270uF 6mΩ ESR would be:
= 1/(18/6mΩ)
= 4860uF 0.333mΩ

You can't use the Parallel Resistance Equation like this unless you use identical caps. ESR can be calculated for dissimilar caps, but it gets more complicated. It's not just easier to use the same capacitors, but they will perform the same and last the same amount of time (on average).

 

[YLODslave]

Hi, I'm sure this gets asked a lot but I'm running out of options for my CECHA01 repair. When I got it, it already had instant YLOD (I turn it on, the fan revs up, then about a second later it beeps and starts flashing red). Because the fan was turning on I ruled out a faulty PSU and decided to replace the two NEC capacitors on the back of the motherboard under the RSX. I replaced them with tantalum capacitors described in the guide. When I tried to turn the ps3 on again I had the same instant YLOD. I made sure everything was connected correctly and that there were no bridges in my Tantalums (Although it is kind of a sloppy job). Is there anything else I can do troubleshoot before I give up completely? Thanks

 

[MYS]

Hi, I'm sure this gets asked a lot but I'm running out of options for my CECHA01 repair. When I got it, it already had instant YLOD (I turn it on, the fan revs up, then about a second later it beeps and starts flashing red). Because the fan was turning on I ruled out a faulty PSU and decided to replace the two NEC capacitors on the back of the motherboard under the RSX. I replaced them with tantalum capacitors described in the guide. When I tried to turn the ps3 on again I had the same instant YLOD. I made sure everything was connected correctly and that there were no bridges in my Tantalums (Although it is kind of a sloppy job). Is there anything else I can do troubleshoot before I give up completely? Thanks

I think for CECHA01 (same as one of my unit), just replace all NEC token and make sure you have at least 4 thick wire (AWG maybe 14 or 16) to link them up. It should work if the failure is due to NEC Token.

 

[YLODslave]

I think for CECHA01 (same as one of my unit), just replace all NEC token and make sure you have at least 4 thick wire (AWG maybe 14 or 16) to link them up. It should work if the failure is due to NEC Token.

Is there a way to test if it's the NEC Tokens before I start buying and replacing more? I read that the instant YLOD can be anything

 

[squidfox]

I also have a console that has the YLOD problem. Removed all the NECs but I did not have enough tantalums so I tried with 1/3 of what is needed but no luck. Then I found two 4700uF electrolytic capacitors and soldered them and added a small wire to connect the two + sides (it was the cut off leg of the cap, about a milimeter thick). After turning the console on still YLOD.

So do you think that with proper connecting wires it would be possible to use the electrolytic capacitors or do they have a very different ESR? I don't have a meter to test it.

Oh and also, if you replace the NECs with the tantalums, please don't try to turn on the console untill the + sides have been connected as otherwise the voltage sense on the VRM controller won't see any voltage and it will increase the PWM to the mosfet controller and the voltage will increse until the caps go out in a glorious firework

 

[MYS]

Is there a way to test if it's the NEC Tokens before I start buying and replacing more? I read that the instant YLOD can be anything

Nope, unless you are able to remove the NEC token without damaging them and a meter to measure it's reading. It is more of a try and error thing, but chance of solving YLOD is high.

 

[RIP-Felix]

So do you think that with proper connecting wires it would be possible to use the electrolytic capacitors or do they have a very different ESR? I don't have a meter to test it.

No, electrolytice have a different profile than tantalums. The NEC/TOKINs are even better than the Tantalums, but Tants are the closest alternative. Tantalum can resist the heat and will last much longer than electrolytics when placed so close to a processor. Lastly, electrolytics have an ESR that is an order of magnitude higher.

You can use an ESR meter to check if Electrolytic caps are Okay, which read ESR accuratly down to about 0.2Ω (200mΩ). However, you need to be able to read 0.0015Ω (1.5mΩ) to check the NEC/TOKINs. This cannot be done with a simple meter. It would require a network analyzer or oscilloscope tricks to pull off. Also, it can't be done in curcuit. The act of removing them intact requires heat that would temporarily restore their capacitance and lower the ESR, giving you the false idea that they are fine when they were not before you heated them. They will fail again soon.

A potential test to see if your NEC/TOKINs are at fault is to heat them to about 180C with hot air. Do so carefully and don't heat for too long - a few minutes at most. That "can" temporarily restore their capacitance and lower the ESR enough for the system to pass the boot check (YLOD). 180C isn't enough to melt the solder, so it's not reflowing them. It can however cause the board to flex enough to restore connection of cracked solder balls, making it difficult to know for certain that's not what happened. This is why you don't want to over do it with the heat. If it works, the YLOD will return soon, but you'll have a pretty good indicator that the caps are at fault. It could still be the caps if this doesn't work and it could still be the BGA. It's not 100% diagnostic.

Note: the process of removing and soldering new caps in heats the remaining NEC/TOKINs, if done properly with hot air & flux to help the solder wet properly (assuming you leave some of them). Many of the failed attempts are likely due to cold solder joints. That's almost impossible to prevent without also using hot air to preheat the area. This can also temporarily restore function to the NEC/TOKINs, potentially giving you the false assumption that the caps you installed fixed the issue. Replacing all the NEC/TOKINs and bridging is the only way to know for certain. And the BGA can and will still fail eventually!

 

[YLODslave]

No, electrolytice have a different profile than tantalums. The NEC/TOKINs are even better than the Tantalums, but Tants are the closest alternative. Tantalum can resist the heat and will last much longer than electrolytics when placed so close to a processor. Lastly, electrolytics have an ESR that is an order of magnitude higher.

You can use an ESR meter to check if Electrolytic caps are Okay, which read ESR accuratly down to about 0.2Ω (200mΩ). However, you need to be able to read 0.0015Ω (1.5mΩ) to check the NEC/TOKINs. This cannot be done with a simple meter. It would require a network analyzer or oscilloscope tricks to pull off. Also, it can't be done in curcuit. The act of removing them intact requires heat that would temporarily restore their capacitance and lower the ESR, giving you the false idea that they are fine when they were not before you heated them. They will fail again soon.

A potential test to see if your NEC/TOKINs are at fault is to heat them to about 180C with hot air. Do so carefully and don't heat for too long - a few minutes at most. That "can" temporarily restore their capacitance and lower the ESR enough for the system to pass the boot check (YLOD). 180C isn't enough to melt the solder, so it's not reflowing them. It can however cause the board to flex enough to restore connection of cracked solder balls, making it difficult to know for certain that's not what happened. This is why you don't want to over do it with the heat. If it works, the YLOD will return soon, but you'll have a pretty good indicator that the caps are at fault. It could still be the caps if this doesn't work and it could still be the BGA. It's not 100% diagnostic.

Note: the process of removing and soldering new caps in heats the remaining NEC/TOKINs, if done properly with hot air & flux to help the solder wet properly (assuming you leave some of them). Many of the failed attempts are likely due to cold solder joints. That's almost impossible to prevent without also using hot air to preheat the area. This can also temporarily restore function to the NEC/TOKINs, potentially giving you the false assumption that the caps you installed fixed the issue. Replacing all the NEC/TOKINs and bridging is the only way to know for certain. And the BGA can and will still fail eventually!

So the BGA will just fail eventually and there's no way to stop it? That's kind of discouraging, lol. I'm trying to find 18 awg wire in my area for the bridging since I replaced the NEC tokins on the underside next to the RSX, I don't have high hopes so I'll probably be back. I've seen a lot of people saying that you should always delid, is that true? Thanks

 

[squidfox]

Thank you for the detailed info. Yes the electrolytic capacitors that I have have an ESR of 100mOhm which is way higher than the NECs. Since I already removed all the NECs from the board I can not test if soldering these on with al the NECs intact would help to determine if the caps are the source of YLOD. It would e convenient to use them for fault finding as they have a large capacitance and easy to solder. Just that the PS3 can not be fully assembled for testing as the caps are large