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

An update from my last post:

I took a food break after making my post and before cleaning up the HH01, I wanted to see if the YLOD triggered.

It did.

To figure out why, I remembered a post (I think from Squeept) that said you can tell if the YLOD is due to needing a reball by putting some weight on the motherboard. So I did, I put maybe about 10lbs of weight from my thumb pressing down on the clamp where the RSX is and booted the PS3. It booted perfectly into the XMB. I then came off the pressure of the motherboard and within seconds the YLOD triggered.

I tried it again, but this time I kept the weight on there until I started feeling some heat coming off the motherboard clamp. After about 5 minutes I started feeling some temp in my thumb and let go. I booted the only game that was installed, which was Flow, and I was able to play it for 30 minutes before shutting it down for the day.

I think this was the reason why i didn't get a YLOD after disassembling the console to clean out the dust bunnies - the stress relief from the motherboard being unmounted allowed the BGA on the RSX to "relax" and make a new connection upon reassembling it to the point where I could boot it up and leave it on.

So definitely it seems the only absolute reason why this HH01 SKU gave the YLOD - is because the RSX needs a reball.

Im not sure where to go from here because I COULD do a reflow as I have a hot air station, but I dont have any way to currently rig up a way to do a proper reflow to hold the board flat as possible. And Im not going to actually do this as a side-hustle because the investment into the necessary tooling and time and more practice units needed to be consistent at doing this job exceeds what Im wiling to spend on this project.


I did come across Booter's youtube channel and he seems to be the guy where I'll send my personal HE01 to do the servicing mine needs (aside from perhaps the frankenstein mod)


So this is currently where I'm at. Im confident in reading the syscon, perhaps a reflow if I can find a way to affix the motherboard to a mounting plate of sorts to elevate it and keep it flat. But outside of that, I think i'll just stick with reading syscon.


One shower-through I had with reading syscon: After entering Internal mode... is there a way to get out of external mode, like instead of using "CXRF" I would go back to using "CXR" and just use the few minor commands available in external mode just to extract the error log? Or does that even matter? Would updating the official firmware have any effect on this? Would it brick the console?
 
Hi there,
Just some info that might be useful for statistics.
PS 3 slim CECH-2004B bought with 2-10s YLOD (usually around 7-8s). Syscon log indicated 801002.
I extracted two NECs and soldered 6 x TPSD337K006R0045 next to RSX
I was quite confident that these are 4.5 mOhms, but as I am looking at the package label it says 0R045, which suggests something else...

Anyways, the console after thermal paste replacement, dusting and putting together, worked like a charm. I played some games, did several power cycles and so far so good.


> errlog
00000000
# CODE CLOCK
# A0801002 2003A0B2
# A0801002 200384C5
# A0801002 200384B0
# A0801002 20038494
# A0801002 20038476
# A0801002 20034858
# A0801002 20034844
# A0801002 20034831
# A0801002 20034805
# A0801002 200347CB
# A0801002 20034784
# A0801002 20034748
# A0801002 20034701
# A0801002 200346F3
# A0801002 20034679
# A0801002 2001EF50
# A0801002 2001EDE2
# A0801002 2001ED0C
# A0801002 2001DE9A
# A0801002 2001DE8B
# A0801002 2001DE00
# A0801002 2001DDF1
# A0801002 1C41B806
# A0801002 1C41B7EF
# A0801002 1C41B7D4
# A0801002 1C38E525
# A0801002 1C38E518
# A0801002 1C38E50B
# A0801002 1C38E500
# A0801002 1C11348D
# A0801002 1BA28E32
 
Hey psx community! I have just received an old COK-002 board fully stripped of NEC/Tokins. At this point, seems like using syscon is useless until I install some replacement capacitors there to further diagnose.
Spend last 2 days soaking all the information from this forum, Felix's YT videos and I am aware that the most probable scenario is faulty 90nm RSX gpu, but I wouldn't like to draw any conclusions without actually receiving an error that points to that.
At this point, I am blind, or better said overwhelmed with a quantity of information I gathered.
Wanted to first simply buy 32 * 320uF 2,5V with 9mOhm to install them, but then I found out about these PS3 Tanatalizers and some recommendations to install MLCCs in addition for high frequencies...

Is it worth it to go with MLCCs, or should I just go ahead with my original plan to buy 32 Panasonic ETPE330MA9GB, install them and continue with syscon?
 
Hook up uart interface and dump historical logs - it might be that some other error than tolkin's related are still stored in a memory - it costs nothing. It is better to have this knowledge prior to any further investments. Regarding which way to choose regarding caps, I will not help much.
 
Hi, I got PS3 fat DIA-002, but with YLOD. I connect PS3 to FT232, and with SYSCON reader find this:
ERR 00: 00000000 A0A02031 FFFFFFFF
ERR 01: 00000000 A0A02031 FFFFFFFF
ERR 02: 00000000 A0A02031 FFFFFFFF
ERR 03: 00000000 A0232102 FFFFFFFF
ERR 04: 00000000 A0A02031 FFFFFFFF
ERR 05: 00000000 A0A02031 FFFFFFFF
ERR 06: 00000000 A0201B02 FFFFFFFF
ERR 07: 00000000 A0A02031 FFFFFFFF
ERR 08: 00000000 A0201B02 FFFFFFFF
ERR 09: 00000000 A0A02031 FFFFFFFF
ERR 10: 00000000 A0201B02 FFFFFFFF
ERR 11: 00000000 A0A02031 FFFFFFFF
ERR 12: 00000000 A0201B02 FFFFFFFF
ERR 13: 00000000 A0201B02 FFFFFFFF
ERR 14: 00000000 A0201B02 FFFFFFFF
ERR 15: 00000000 A0A02031 FFFFFFFF
ERR 16: 00000000 A0201B02 FFFFFFFF
ERR 17: 00000000 A0A02031 FFFFFFFF
ERR 18: 00000000 A0201B02 FFFFFFFF
ERR 19: 00000000 A0A02031 FFFFFFFF

Can you help me please. Is it a problem with broken Temperature Monitor?
 
Good morning. I have a ps3 CECHA00 with 60 days of use. CFW EVILNAT 4.89.3. Never opened and clean. it starts after about 15 start attempts. After it started, it works perfectly. When I turn it off via the XMB menu, it beeps 3 times then turns off then the red LED flashes. could it be the capacitors? Thank you for your answers.

Edit: CPU 52° RSX 43° FAN 34% after 2h idle on XMB.
 
Hello everybody.

What is the beginning of an error in the log, is it ERR 00: 00000000 A0403034 FFFFFFFF or ERR 19: 00000000 A0202120 FFFFFFFF?

ERR 00 OR ERR 19?


-===================================
ERR 00: 00000000 A0403034 FFFFFFFF
ERR 01: 00000000 A0403034 FFFFFFFF
ERR 02: 00000000 A0403034 FFFFFFFF
ERR 03: 00000000 A0403034 FFFFFFFF
ERR 04: 00000000 A0403034 FFFFFFFF
ERR 05: 00000000 A0403034 FFFFFFFF
ERR 06: 00000000 A0403034 FFFFFFFF
ERR 07: 00000000 A0403034 FFFFFFFF
ERR 08: 00000000 A0403034 FFFFFFFF
ERR 09: 00000000 A0403034 FFFFFFFF
ERR 10: 00000000 A0202120 FFFFFFFF
ERR 11: 00000000 A0202120 FFFFFFFF
ERR 12: 00000000 A0202120 FFFFFFFF
ERR 13: 00000000 A0202120 FFFFFFFF
ERR 14: 00000000 A0202120 FFFFFFFF
ERR 15: 00000000 A0202120 FFFFFFFF
ERR 16: 00000000 A0202120 FFFFFFFF
ERR 17: 00000000 A0202120 FFFFFFFF
ERR 18: 00000000 A0213013 FFFFFFFF
ERR 19: 00000000 A0202120 FFFFFFFF
====================================
 
Hi all, wanted to detail my experience here with YLOD PS3 bought from ebay. CECHH01 40GB MGS Special edition.

I just wanted a PS3 to try softmodding or homebrew or whatever and get a bargain on a quick fix...

The listing had this description:

"THIS UNIT WILL SOMETIME ACUALLY WORKS WITH NO PROBLEM BY THIS I MEAN IT WILL OPERATE AS IT SHOULD.
BUT WHEN THIS CONSOLE WONT ACT RIGHT THIS WILL LET GAME GO IN AND PLAY WITH NO PROBLEMS BUT SOMETIMES THIS UNIT WILL TAKE THE GAME AND THE LIGHT WILL BLINK AND GO TO RED TO ME I BELEIVE THAT THE FAN HAS AN ISSUE CAUSE I CAN HEAR IT SPIN AND THEN IT STOPS THEN THE RED LIGHT WILL COME ON."

When I got the unit it was YLOD immediately. Tried it a few times within 2 days, always YLOD after 3+ seconds. I wasn't aware of what I was getting myself into. Lots of research and a quick snizz at the syscon with the handy as heck usb serial adapter I bought for peering into random electronics showed A0801002 from line 1 to 20 minus err on line 11. After lots of research decided to give RIP-Felix's Tantalizer a go.

Tried to get PCBWay to make the boards but the engineer kicked back the 0.6b gerbers saying "We need the holes/slots to be 0.45mm inside of the board outline, or it cannot be made.
The size of the board is enlarged a bit in the fabrication file." I took a squizz at the modified gerbers but honestly I couldn't make out half of what I was looking at forget trying to eyeball a pcb size difference. I decided whats another $30 on top of this and had the boards made with the PCBWay modified gerbers.

To remove the tokins I used a half gorilla half shaky hands surgeon approach, breaking the tokin apart but not ripping stuff off the traces. To remove the shrapnel from traces I used lots of flux, solder, tweezers and a pinecil with stock tip at 375c. I lifted one trace on the "outside" section so yeah, less gorilla. I cleaned the traces with the Louis Rossmann method of solder, wick, clean, flux, solder, wick, clean. I soldered the Tantalizer boards on as best I could then added the 3 large 470uf caps from Tantalizer guide. Sorry, couldn't afford the extra 4 special caps per board =/. I also need to mention I only replaced the 4 tokins for the RSX and not the CELL.

When ripping up the tokins the ones on the side opposite the cell and rsx looked fine but the 2 living next to the RSX looked like they were burnt on the inside around the edges of what I assume is the foil pack. Removing the tokin lid before shredding it might give you an indication of the tokins current status? The burnt caps are in the below image. The good ones looked the same without the brown around the edges.
avcs5nq.jpg


Here is an image of the final product in shaky hands potato-o-vision

CbYspNQ.jpg


After replacing the 4 RSX tokins, I did a quick test and it didnt YLOD so I put it back together, connected it to the TV and boy was I excited to see...

v2fbpIo.jpg


Oh, ... anyone have a PS3 controller they want to lend me?! (kidding)

I can't say for certain that the Tantalizer fixed it since I cant fully test it right now without a controller but before replacing the tokins all it ever did was YLOD the few times I turned it on. In any case I would like to give a huge thanks to RIP-Felix and everyone who has made PS3 diagnosis more than a guess and check affair. Actual diagnosis is so much better than "stick it in the oven dude it totally works bruh!" like every YLOD fix from a quick web search.

Felix, if you want the PCBWay modified gerbers please let me know and I would be happy to send you some of the extra PCBs I had made, with extra caps, at my expense.

TL;DR: YLOD ebay special. Syscon read A0801002 so I replaced the 4 RSX tokins with Tantalizers. Tokins near RSX looked like they were burnt around the edges on inside. Appears to be fixed.
 
Glad you got it working. Nice work diagnosing the problem!

You should keep the tantalizers in case you ever need them again. Never know. I have plenty of prototypes I never used. So I don't need them.

PCB manufactuerers complaim about the file sometimes and other times they dont. Luck of the draw whether you'll get a tect look at them and be like, "WTF is this?" Then proclaim they need to modify it to get it to work. I'm working on a revision to avoid these kinds of issues. v1.0 will hopefully solve that.

The scortching around tokin paripherals is normal. It doesn't seem to be an indicator of their overall health, as we've seen severly scorched "looking" caps that still worked. And confirmed bad ones that didnt have any at all.
 
Glad you got it working. Nice work diagnosing the problem!

You should keep the tantalizers in case you ever need them again. Never know. I have plenty of prototypes I never used. So I don't need them.

PCB manufactuerers complaim about the file sometimes and other times they dont. Luck of the draw whether you'll get a tect look at them and be like, "WTF is this?" Then proclaim they need to modify it to get it to work. I'm working on a revision to avoid these kinds of issues. v1.0 will hopefully solve that.

The scortching around tokin paripherals is normal. It doesn't seem to be an indicator of their overall health, as we've seen severly scorched "looking" caps that still worked. And confirmed bad ones that didnt have any at all.

Thanks! After watching your video I feel lucky it actually seems to be the tokins and not the rsx itself. The unit was entirely clogged with dust and I bet the previous owner really did think it might be a problem with the fan because I bet it was LOUD! Maybe it cooked the caps. That is interesting that cooked looking cap internals is not an indicator of their status. I would have expected for sure it would be a sign of at least marginality. Of course once you pop the top you are going to have to replace it anyway.

I'd like to point out at least for me having only access to a regular soldering iron and no hot air rework I was not able to reflow the entirety of the negative rails under the tants. I wopped as much heat and solder as I was comfortable with at the outer connection I could get at but I am positive it did not reflow the entire 2 surfaces together underneath the board. It could be my technique sucks but I wanted to point that out.

I ordered a controller and plan to thoroughly put it through its paces if its truly functional. If anything else goes wrong I promise to leave a note for further data collection.
 
Of course once you pop the top you are going to have to replace it anyway.

Actually, if you don't damage the contacts, it'll be fine. It's just a resin case to prevent shorts. A rather radical strategy to cool the tokins is to remove this case and use thermal pads directly touching the capacitor, so that it more efficiently conducts heat to the RF shield, extending their lifespan.

Fun fact I left out of my documentery to save time.

I'd like to point out at least for me having only access to a regular soldering iron and no hot air rework I was not able to reflow the entirety of the negative rails under the tants. I wopped as much heat and solder as I was comfortable with at the outer connection I could get at but I am positive it did not reflow the entire 2 surfaces together underneath the board. It could be my technique sucks but I wanted to point that out.
That's fine. It's not required to get the entire rail to flow. I designed it that way so that they could be reflowed on the board like a tokin and that would give a very low impeedance path to ground and the positive rails. But it's more than good enough just making 4 spot connections at the corners. It's safer too, since you don't have to dump a bunch of heat into the area next to the processors (putting the BGA at risk).

Based on the seller's notes I suspect that the CPU needs delidded. Prepare yourself.
 
To remove the tokins I used a half gorilla half shaky hands surgeon approach, breaking the tokin apart but not ripping stuff off the traces. To remove the shrapnel from traces I used lots of flux, solder, tweezers and a pinecil with stock tip at 375c. I lifted one trace on the "outside" section so yeah, less gorilla. I cleaned the traces with the Louis Rossmann method of solder, wick, clean, flux, solder, wick, clean. I soldered the Tantalizer boards on as best I could then added the 3 large 470uf caps from Tantalizer guide. Sorry, couldn't afford the extra 4 special caps per board =/. I also need to mention I only replaced the 4 tokins for the RSX and not the CELL.

If you have a preheater, this process is much simpler and safer - its a case of just pre baking the board at 100c for at least 20 minutes, then gradually crank up the preheater to 150c, then use a hot air gun at 350c air speed 5-6, wave around the tokens and they just slide off without any damage.

PS3 boards suck heat like there is no tommorrow, so getting the prebake (up to temps) process is the key to all of this, and no one mentions this - otherwise you get burnt boards or warping effect going on.
 
PCB manufactuerers complaim about the file sometimes and other times they dont. Luck of the draw whether you'll get a tect look at them and be like, "WTF is this?" Then proclaim they need to modify it to get it to work. I'm working on a revision to avoid these kinds of issues. v1.0 will hopefully solve that.

My first batch of these pcb's are on the way - PCBway complains about the size of the holes, saying it needs at least 0.25mm, so in the end adjusted the design a bit.

Thanks for the design RIPfelix! - Would be great to share on pcbway on a working model of this pcb?
 
PCBway complains about the size of the holes, saying it needs at least 0.25mm, so in the end adjusted the design a bit.
This makes no sense anyway, since there are no holes except the slots, which are larger than 0.25mm. So IDK what they're on about. Sounds like complaining.
 
This makes no sense anyway, since there are no holes except the slots, which are larger than 0.25mm. So IDK what they're on about. Sounds like complaining.

Yeh, i adjusted it enough for them not to complain anymore, will see how it works out when i recieve them.
 
I have a spot in TLOU that will trigger a YLOD on two of my systems. It is 100% repeatable which makes it a great test. If anybody has diagnosed a known bad cap system, it would be cool to see if you can load my save for this particular spot and see if it triggers YLOD. In both my cases it was 1002 fault
 
Hello everybody.

What is the beginning of an error in the log, is it ERR 00: 00000000 A0403034 FFFFFFFF or ERR 19: 00000000 A0202120 FFFFFFFF?

ERR 00 OR ERR 19?


-===================================
ERR 00: 00000000 A0403034 FFFFFFFF
ERR 01: 00000000 A0403034 FFFFFFFF
ERR 02: 00000000 A0403034 FFFFFFFF
ERR 03: 00000000 A0403034 FFFFFFFF
ERR 04: 00000000 A0403034 FFFFFFFF
ERR 05: 00000000 A0403034 FFFFFFFF
ERR 06: 00000000 A0403034 FFFFFFFF
ERR 07: 00000000 A0403034 FFFFFFFF
ERR 08: 00000000 A0403034 FFFFFFFF
ERR 09: 00000000 A0403034 FFFFFFFF
ERR 10: 00000000 A0202120 FFFFFFFF
ERR 11: 00000000 A0202120 FFFFFFFF
ERR 12: 00000000 A0202120 FFFFFFFF
ERR 13: 00000000 A0202120 FFFFFFFF
ERR 14: 00000000 A0202120 FFFFFFFF
ERR 15: 00000000 A0202120 FFFFFFFF
ERR 16: 00000000 A0202120 FFFFFFFF
ERR 17: 00000000 A0202120 FFFFFFFF
ERR 18: 00000000 A0213013 FFFFFFFF
ERR 19: 00000000 A0202120 FFFFFFFF
====================================

00
 
Back
Top