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

[Edwired]

Trouble started when people automatically jump to reflowing then sell it faulty when the damage is too far gone. But fair play to him to try a different way of fixing ylod is related to nec tokin

 

[Rezendes]

Hello, currently attempting this fix on my YLOD CECHA01 fat ps3 with no luck. I have tried replacing the power supply with a known good psu as well. I removed 4 of the NEC capacitors on one side of the board (side without the cpu/gpu) and placed 16 of the tantalum 470uF 6v diagonally as the guide suggests. No change in the instant ylod. I am trying to figure this out so right now I have all the tantalums removed and did my best to clean up the pads. the Positive and Negative pads are shorted from all points, all pads have continuity, is this supposed to be this way? Remember, I still have the other 4 NEC caps on the other side of the board (with cpu/gpu). Is there a way I can test the still present NEC caps on the other side? I have no more tantalums to replace those with currently, I will likely order more when I can.

 

[jacobsson]

A hobbyist fixing stuff, love his channel so I thought I'd share:

 

[wrx884]

A hobbyist fixing stuff, love his channel so I thought I'd share:

lol just watching this now i always find his stuff entertaining haha. wheres squeept? oh wait he threatened to leave again didnt he? LOL hes going to be pulling out what hair hes got left when he sees this hahahaha

Yes, that is it. I have some spare ps3s. Is any ps3 cell good to replace that?

yeah unfortunately shes toast, it can be fixed but its a little involved as snake mentioned, unless u got the tools and rework machines to repair it urself, it wouldnt be viable to get repaired as just in the labour u will freak out with the costs lol. put this aside as a test board it'll be handy for something at least.

M

I expect Steve from TronicsFix to work on this topic too, but unfortunately as far as i am aware, he doesn't work with PlayStation3s, as he owns a repair shop, is not really a good business to repair consoles that for him aren't worth selling,but he might be generous enough to try it.

Sent from my G8141 using Tapatalk

Id be surprised if he bothers, from memory he doesn't own a rework station either and gets an outsider to do any reballing etc... i think hes just set up to do smaller easier fixes and what not, but he does play with My Mate Vince a bit so it could happen.

 

[Eld]

[
you accidentally did not mix up the polarity when soldering?

Ughh! Yes you're right, how embarassing. I incorrectly assumed the big line was for negative I will have to re-solder them all!

I wonder why did the fix appear to work then, if the capacitors are all reversed?

 

[Naked_Snake1995]

lol just watching this now i always find his stuff entertaining haha. wheres squeept? oh wait he threatened to leave again didnt he? LOL hes going to be pulling out what hair hes got left when he sees this hahahaha



yeah unfortunately shes toast, it can be fixed but its a little involved as snake mentioned, unless u got the tools and rework machines to repair it urself, it wouldnt be viable to get repaired as just in the labour u will freak out with the costs lol. put this aside as a test board it'll be handy for something at least.



Id be surprised if he bothers, from memory he doesn't own a rework station either and gets an outsider to do any reballing etc... i think hes just set up to do smaller easier fixes and what not, but he does play with My Mate Vince a bit so it could happen.

Same thoughs, as almost no rework station is required to do so, just something to pry them off and solder them.

Steve actually responded to MyMateVince, stating that he found the fix rather interesting and never knew that could be the issue, so can we expect this to go mainstream, MyMateVince covered it, DisplacedGamers covered it, RetroRGB covered it, now let's way for TronicsFix to cover it [emoji23]

Is nice how word spreads out so quickly, i never thought this thread would be highly viewed and followed.

PS: No diss thowards Squeept, although we started on the wrong foot here, he actually helped us with his equipment, which put this thread on the right tracks, so i must thank him personally, i know we kinda didn't agree on our methods, but hey opinions have to be respected.

Sent from my G8141 using Tapatalk

 

[wrx884]

Same thoughs, as almost no rework station is required to do so, just something to pry them off and solder them.

Steve actually responded to MyMateVince, stating that he found the fix rather interesting and never knew that could be the issue, so can we expect this to go mainstream, MyMateVince covered it, DisplacedGamers covered it, RetroRGB covered it, now let's way for TronicsFix to cover it

Is nice how word spreads out so quickly, i never thought this thread would be highly viewed and followed.

PS: No diss thowards Squeept, although we started on the wrong foot here, he actually helped us with his equipment, which put this thread on the right tracks, so i must thank him personally, i know we kinda didn't agree on our methods, but hey opinions have to be respected.

Sent from my G8141 using Tapatalk

Well u kinda do need the rework station so u can remove the faulty cap/s to test them properly or remove the BGA's, other wise its still a guessing game at this stage still. Someone like Steve i doubt would waste too much time on guessing, i mean shit i still havent had any luck with them yet and squeept claims hes done 30 odd over the last couple years i think i read so those odds just there is hard to pick, kinda frustrating too lol. I am starting to notice more of a pattern on the G-P models tho that are popping up as fixes so maybe thats my issue not enough of the later fats being attempted.

ur right the more people that can expose it, it will only help us all in the long run so the more exposure the better.

Hahaha dw im sure squeept will read that as "playful" i mean its only hair right? hhahaha

 

[jacobsson]

@wrx884 AFAIK a CELL is only good paired with its original SYSCON and NANDS, but I guess if you're able to replace a CELL the other chips should be a breeze. I might be totally wrong about this though

 

[wrx884]

@wrx884 AFAIK a CELL is only good paired with its original SYSCON and NANDS, but I guess if you're able to replace a CELL the other chips should be a breeze. I might be totally wrong about this though

yeah pretty much, only down fall is if ur working on an unknown YLOD board already, ur more than likely going to need to do the GPU too or any other hidden issues that might be causing it but yeah the swap isnt hard if u got the equipment. The Cell isnt as easy to keep flat when re attaching to the mother board they are much thinner than the GPU's

 

[Maroon Storm]

My straight question. Can I use 3(2 one side 1 on the other side)+3+3+3 470uF tantrum on slim 2xxxx? I have only 470s. At the moment on the bottom part 4+4 (this where nothing on that side) tantrums are placed, on the other side 3+3 tantrums are placed parallel with a thick cable soldered to the ground. This slim doesn't want to start with this setup. Maybe am I just soldered wrong something? Because once it has started itself when I had used 3+3 parallel on both side but after a couple of on/off cycle. Thank you for the help!

it can be fixed but its a little involved as snake mentioned

Basiclly that CELL unfixable. Cool. I'm so dissapointed. Today I'd opened a water damaged phat. I cleaned it with a lot of IPA. But still it has just blinked red after the green. I don't think I'm going to solder tantalums on that one.

RetroRGB

Which video is that? I want to watch his video about this. I've seen all the other ones.

I wonder why did the fix appear to work then, if the capacitors are all reversed?

This was almost happened with me. But still this is interesting why is that working?

 

[Naked_Snake1995]

My straight question. Can I use 3(2 one side 1 on the other side)+3+3+3 470uF tantrum on slim 2xxxx? I have only 470s. At the moment on the bottom part 4+4 (this where nothing on that side) tantrums are placed, on the other side 3+3 tantrums are placed parallel with a thick cable soldered to the ground. This slim doesn't want to start with this setup. Maybe am I just soldered wrong something? Because once it has started itself when I had used 3+3 parallel on both side but after a couple of on/off cycle. Thank you for the help!


Basiclly that CELL unfixable. Cool. I'm so dissapointed. Today I'd opened a water damaged phat. I cleaned it with a lot of IPA. But still it has just blinked red after the green. I don't think I'm going to solder tantalums on that one.



Which video is that? I want to watch his video about this. I've seen all the other ones.



This was almost happened with me. But still this is interesting why is that working?

You can use a 2x1 470uF configuration, making a total of 3 per NEC on a 2000 Slim, there is nothing wrong with that, ive seen that before, do with what you have at the moment, and if that isn't enough, do a 2x2 configuration.

Regarding the RetroRGB coverage, its only a weekly news podcast coverage, like back in September, i think it was podcast #75 or #76, he was tempted to try it out, although his schedule is busy, so he didn't announce when he was going to try it out,but he did mention he is curious about it, and to try on a few YLOD systems, just to see the result.

Sent from my G8141 using Tapatalk

 

[xxxmarioxxx]

Hello, I'm planning to try this fix on a YLOD CECHA01 motherboard, but I still have some quesitons,

It seems that old NEC capacitors are very difficult to remove, even with heat guns. Heat gun will melt the plastic shell but internals of the capacitor are still soldered to the motherboard. Can anyone suggest the best way to remove these capacitors?

I found these SMD tantalum capacitors on RS:
Panasonic Tantalum Capacitor 1500μF 2.5V dc Polymer Solid ±20% Tolerance POSCAP TPE Series
Can I use two of these to replace two NEC capacitors on the back of the motherboard?

Thanks.

 

[Beanerman]

Well, my system went back to ylod again. Was running fine, then suddenly just shut down.
Any thoughts?
Sent from my SM-N950U using Tapatalk

 

[Beanerman]

Success! I replaced 1nec, rsx, just like tutorial. CECH-G
4- 470uf tantaliums.
Covered them with electric tape.
Put system back together, and Boom, no YLOD. I would've used my own finished pics, but I got so excited, I just put it back together before I realized, LOL.
Thank you Naked_Snake!

Sent from my SM-N950U using Tapatalk

Above post

Sent from my SM-N950U using Tapatalk

 

[SiegHart93]

Above post

Sent from my SM-N950U using Tapatalk

perhaps u need to replace another nec tokin on the rsx ! i recommend u to replace 1 nec on cell too

 

[Beanerman]

Ok, replaced another on rsx, and 1 on cpu side, still 2.5 to 3 seconds ylod. I guess 1 more on cpu and see what happens.

Sent from my SM-N950U using Tapatalk

 

[Edwired]

I should be expecting the ps3 either this week or next week due to pitney bowes postage been slow lately

 

[Moltiplier]

Hi all, I've been watching this thread for some time, and it's been an interesting discussion to watch. For people's curiosity, I wanted to share some of the info about the NEC/Tokin Proadlizer capacitors that I've found looking into the issue.

These are just an assortment of literature and datasheets about the capacitors with the first one being what I think is the most interesting. It discusses how the capacitors are built, and for comparison, I've linked the Wikipedia article discussing polymer electrolytic caps.

https://www.nec.com/en/global/techrep/journal/g06/n05/pdf/t060514.pdf
https://en.wikipedia.org/wiki/Polymer_capacitor
https://www.nec.com/en/global/techrep/journal/g09/n01/pdf/090116.pdf
https://file-databank.netdna-ssl.com/wp-content/uploads/2018/09/NEC-TOKIN-0E907-DATASHEET.pdf
http://static6.arrow.com/aropdfconversion/36d0ad746e44626c36256180d88868ee6cc4e2d5/wrcatalog.pdf
http://datasheet.elcodis.com/pdf/13/95/139530/awr-2r5srb122mf25s.pdf

For the issue of voltage on the chips, the schematic states the nominal voltages being 1.0V for the Cell and 1.2V for the RSX. I don't know if these are the exact voltages, and they may have ranges that they operate. I think that various 90nm processes could do upward of 1.5V. That 1.6V figure seems wrong to me and could be a measurement error.

On the issue of ESR, remember that ESR reduces when you put capacitors in parallel. You don't need 1mOhm ESR caps when replacing one NEC/Tokin caps with 4 tantalum or aluminum polymer, but if you want to match ESR, then if you replace 1 cap, you need 4 capacitors with 4mOhm ESR.

Also, a small change in ESR isn't going to raise the voltage to 1.6V. In fact, @squeept that spectrum looks wrong for the GPU power supply because those appear to be square waveforms. This is using a buck converter meaning that the square wave will go through an inductor on its way to the load. The ripple voltage across the load should be a triangle waveform rather than the square waveform shown in the image.

I'd suggest double-checking how you have your oscilloscope connected to the PS3 because it looks almost like you have it on the wrong side of the inductor. Of course, it could be something else, but a square wave shouldn't be coming from the power supply if the probe is connected to the capacitors.

As for what could cause theses caps to fail, I'm wondering if these caps could fail with an intermittent short. If anyone knows of a Toshiba laptop with the issues described for bad NEC/Tokin, removing them might be the best bet for getting something to give a concrete answer to faults in these caps.

I suggest these since a sizable number had issues, and as far as I am aware, they haven't been shown to have potential other causes such as the PS3 and RSX. They may help in pinpointing problems with the PS3 caps. The laptop may have to be destroyed in the process though to not expose the cap to heat.

As for whether these are relevant, I disagree with the sentiment that the failures in Toshiba laptops are somehow irrelevant to potential failures in the PS3. The use cases are the same with a few differences circuit designs. I find it hard to believe that a Toshiba laptop would push these caps beyond 5Arms, which I'm guessing that the 900uF cap in the laptop has a similar spec to the WR series caps.

Toshiba A300 schematic: https://www.informaticanapoli.it/download/SCHEMIELETTRICI/TOSHIBA/Toshiba Satellite a300_a305_inventec_potomac_10s_pt10s_rev_x01_sch.pdf

@Beanerman Have you checked for shorting between the power rail and ground?

Mod edit: edited formatting for easier readability

 

[Beanerman]

Everything seems to be holding a current, but again, I only have a simple multimeter to check such things.
Thanks @Moltiplier

Sent from my SM-N950U using Tapatalk

 

[Naked_Snake1995]

Hi all, I've been watching this thread for some time, and it's been an interesting discussion to watch. For people's curiosity, I wanted to share some of the info about the NEC/Tokin Proadlizer capacitors that I've found looking into the issue.

These are just an assortment of literature and datasheets about the capacitors with the first one being what I think is the most interesting. It discusses how the capacitors are built, and for comparison, I've linked the Wikipedia article discussing polymer electrolytic caps.

https://www.nec.com/en/global/techrep/journal/g06/n05/pdf/t060514.pdf
https://en.wikipedia.org/wiki/Polymer_capacitor
https://www.nec.com/en/global/techrep/journal/g09/n01/pdf/090116.pdf
https://file-databank.netdna-ssl.com/wp-content/uploads/2018/09/NEC-TOKIN-0E907-DATASHEET.pdf
http://static6.arrow.com/aropdfconversion/36d0ad746e44626c36256180d88868ee6cc4e2d5/wrcatalog.pdf
http://datasheet.elcodis.com/pdf/13/95/139530/awr-2r5srb122mf25s.pdf

For the issue of voltage on the chips, the schematic states the nominal voltages being 1.0V for the Cell and 1.2V for the RSX. I don't know if these are the exact voltages, and they may have ranges that they operate. I think that various 90nm processes could do upward of 1.5V. That 1.6V figure seems wrong to me and could be a measurement error.

On the issue of ESR, remember that ESR reduces when you put capacitors in parallel. You don't need 1mOhm ESR caps when replacing one NEC/Tokin caps with 4 tantalum or aluminum polymer, but if you want to match ESR, then if you replace 1 cap, you need 4 capacitors with 4mOhm ESR.

Also, a small change in ESR isn't going to raise the voltage to 1.6V. In fact, @squeept that spectrum looks wrong for the GPU power supply because those appear to be square waveforms. This is using a buck converter meaning that the square wave will go through an inductor on its way to the load. The ripple voltage across the load should be a triangle waveform rather than the square waveform shown in the image.

I'd suggest double-checking how you have your oscilloscope connected to the PS3 because it looks almost like you have it on the wrong side of the inductor. Of course, it could be something else, but a square wave shouldn't be coming from the power supply if the probe is connected to the capacitors.

As for what could cause theses caps to fail, I'm wondering if these caps could fail with an intermittent short. If anyone knows of a Toshiba laptop with the issues described for bad NEC/Tokin, removing them might be the best bet for getting something to give a concrete answer to faults in these caps.

I suggest these since a sizable number had issues, and as far as I am aware, they haven't been shown to have potential other causes such as the PS3 and RSX. They may help in pinpointing problems with the PS3 caps. The laptop may have to be destroyed in the process though to not expose the cap to heat.

As for whether these are relevant, I disagree with the sentiment that the failures in Toshiba laptops are somehow irrelevant to potential failures in the PS3. The use cases are the same with a few differences circuit designs. I find it hard to believe that a Toshiba laptop would push these caps beyond 5Arms, which I'm guessing that the 900uF cap in the laptop has a similar spec to the WR series caps.

Toshiba A300 schematic: https://www.informaticanapoli.it/download/SCHEMIELETTRICI/TOSHIBA/Toshiba Satellite a300_a305_inventec_potomac_10s_pt10s_rev_x01_sch.pdf

@Beanerman Have you checked for shorting between the power rail and ground?

Mod edit: edited formatting for easier readability

Finally someone who comes with solid information,and with plenty of homework done, not to mention finally,but finally links the TOSHIBAs as a potential failure counterpart with the PlayStation3s, i dont know how many times ive mentioned that already,and there is a good reason why,but i always go contradicted, at least someone speaks on common sense.

The same TOSHIBA A300 schematic you provide,was the same Laptop ive owned from 2008 to 2011, it died due to a NEC failure, of course i didnt knew at the time, nor nobody knew it could link to the NECs themselfs, everyone just assumes the CPU or the GPU was fried, mind you that this was the time when GPU and CPUs on TOSHIBAs werent soldered,but rather PGAs and PCI Links, so that wasnt a CPU or GPU issue, later on i discovered about the NECs, around the end of 2011,but by them i have sold that laptop,but from videos sure as hell it worked alright, even though i never made a link with the PlayStation3, because this was unknown at the time.

One little neat trick ive found with the laptop, was if you heated the back,just long enough,which i did, the laptop would work again flawlessly, sounds like a familiar Hair-Dryer fix from the PlayStation3? Same principle are applied, if both have the same component failure,even though the PS3 circuit is designed differently.

About the 90nm Voltage, i belive ive read somewhere but cant recall, that the CELL runs at 1.5V´ish, and the RSX the same or a bit lower, i know 1.5V sounds a lot by todays standarts,but by 2005 standarts was pretty much normal for a 90nm chip, for comparasions sake, a Core 2 Duo first Gen,which arrived the same time with the PS3, ran at 1.3V´ish, a Pentium D, around the same CELL mark, so i see 1.5V being pretty common on Chips from that era.