PS3 Fault finding YLOD with the SYSCON - First steps and Error reporting

How did you get rid of the excess balls? Suction pickup tool?

This is a good idea to reduce thermal cycles on the IC! With the board on the preheater it should make flowing the balls easier too...less top heat. I was thinking that using a reflow hot plate for the IC would make balling the chip easier. No hot air to blow the balls together.
 
Yes that syringe is made for pick excess balls.
I leave direct heat stencil on board. I see many people complaining about wrapping stencil, pushing with any longer metal piece will keep in place and heating in small areas, when moving to another area will keep stencil on right point because of already soldered balls getting cold. In my case wrap is happening on pcb method in middle only, so I increased bottom from 120 to 140. Top reading 120 now less wrap but I don't want to see bottom of board very rusted so I go quick. Anyway less heat stress will get to ic, tedious probably for ppl who are confident doing it with non direct heat stencil. I tried that method long time ago and it is a pain for me as blind person. I want to work more under microscope from now.
Edit
An don't worry about placing ic, if you place it you will fill with both hands if push down gently on all rams and try moving on horizontal direction it won't move, that's right position.
I will reply more after another jsd board. If this goes well on both, I'm thinking of going like this and test also lead free balls.
Once again logic explains why flux has to be cleaned/wash underneath of any bga: with time that ic getting hot will activitate back flux and again ylod/blod error. Once board is tested and and starts fine, board is clean with ipa and pcb cleaning flux (water based),dry 4 hours is best at 120 in well measured oven.
 
Last edited:
Hi everyone
Just picked up a CECHA COK-001 off ebay with YLOD. System had never been opened prior to my doing. Ran syscon and these are my error codes:
ERR 00: 00000000 A0801200 0B48868E
ERR 01: 00000000 A0801200 FFFFFFFF
ERR 02: 00000000 A0801001 FFFFFFFF
ERR 03: 00000000 A0801200 FFFFFFFF
ERR 04: 00000000 A0801200 FFFFFFFF
ERR 05: 00000000 A0801200 FFFFFFFF
ERR 06: 00000000 A0801200 FFFFFFFF
ERR 07: 00000000 A0801200 0B4886B7
ERR 08: 00000000 A0801200 0B48869D
ERR 09: 00000000 A0801200 0B48868E
ERR 10: 00000000 A0801001 0B488688
ERR 11: 00000000 A0801200 0B48868C
ERR 12: 00000000 A0801001 0B48868E
ERR 13: 00000000 A0801001 0B488681
ERR 14: 00000000 A0902203 FFFFFFFF
ERR 15: 00000000 A0902203 0B488685
ERR 16: 00000000 A0801001 0B488684
ERR 17: 00000000 A0801001 FFFFFFFF
ERR 18: 00000000 A0801001 0B488687
ERR 19: 00000000 A0801200 0B488686
Tried looking up the codes myself without much success. Could someone please help me decipher these codes and see if this console is fixable?
 
funtyler said:
Hi everyone
Just picked up a CECHA COK-001 off ebay with YLOD. System had never been opened prior to my doing. Ran syscon and these are my error codes:
ERR 00: 00000000 A0801200 0B48868E
ERR 01: 00000000 A0801200 FFFFFFFF
ERR 02: 00000000 A0801001 FFFFFFFF
ERR 03: 00000000 A0801200 FFFFFFFF
ERR 04: 00000000 A0801200 FFFFFFFF
ERR 05: 00000000 A0801200 FFFFFFFF
ERR 06: 00000000 A0801200 FFFFFFFF
ERR 07: 00000000 A0801200 0B4886B7
ERR 08: 00000000 A0801200 0B48869D
ERR 09: 00000000 A0801200 0B48868E
ERR 10: 00000000 A0801001 0B488688
ERR 11: 00000000 A0801200 0B48868C
ERR 12: 00000000 A0801001 0B48868E
ERR 13: 00000000 A0801001 0B488681
ERR 14: 00000000 A0902203 FFFFFFFF
ERR 15: 00000000 A0902203 0B488685
ERR 16: 00000000 A0801001 0B488684
ERR 17: 00000000 A0801001 FFFFFFFF
ERR 18: 00000000 A0801001 0B488687
ERR 19: 00000000 A0801200 0B488686
Tried looking up the codes myself without much success. Could someone please help me decipher these codes and see if this console is fixable?
Click to expand...

As far as I know it's "just" a thermal error.
Try cleaning the motherboard from dust, etc.
And remove (completely remove) any thermal paste used on the RSX and CELL processors (and the heatsink too).
Then apply new thermal paste on the RSX and CELL, close everything, and try turning it on again.
Good luck! ~
 
Yes, error 1200 is a CPU overheat. Very common error. It needs to be dealt with before the heat causes a BGA defect.

Open up, replace thermal paste. If you are capable, the CPU really needs to be delidded and have the old dried out paste between the heat spreader and die replaced. On almost every console I've delidded the CPU dies paste was all dried up and bad enough to cause overheating regardless of replacing the paste between the heatsink and IHS.

But if you're not confident you can do it safely (its easy to destroy traces and brick the console without a good delidding tool for the cpu) then just clean and replace the paste. It may work with good enough results. You'll know if the fan ramps up three or more steps under load that it needs delidding. Its normal to ramp up past the first fan setting, to just audible level. That'll happen a few minutes in at idle. In game it usually ramps another level, kinda noisy not too bad. Thats normal too. But if it ramps above that its suspect and if it goes another level above that, its in need of repaste.
 
Thank you for the quick replies. I should also add when I turn on the PS3 it starts normally for about 10-15 seconds (doesn't quite get to the xmb). After the 15 seconds, the fan spins up to jet engine level and the console shuts off. I guess, I have never heard or seen a system overheat that quick. I am fairly comfortable delidding the CPU and GPU; however, I wanted to see if there was anything else to check before delidding.
 
RIP-Felix said:
Can you provide any history to the conole? There is a 3034 in the log, before the latest 1002's. That could mean it was reflowed/reballed and later developed tokin issues. However, we would need to see the work history on the console. Can you take a picture? I there any flux residue on the RSX? Have the tokins been replaced? If the previous person tried a tantalum mod and used bad caps or technique, it could explain those errors. The 3034 concerns me.
Click to expand...

PS3 was never opened, No flux residue on RSX, no tokins been replaced. Will try replacing the tokins myself later this weekend

Here's some pictures
https://imgur.com/a/9Rj3Mnh

UDc9FBZ.jpg

5mHkc6z.jpg
 
Last edited:
funtyler said:
Thank you for the quick replies. I should also add when I turn on the PS3 it starts normally for about 10-15 seconds (doesn't quite get to the xmb). After the 15 seconds, the fan spins up to jet engine level and the console shuts off. I guess, I have never heard or seen a system overheat that quick. I am fairly comfortable delidding the CPU and GPU; however, I wanted to see if there was anything else to check before delidding.
Click to expand...
Yup, a CPU without a HS can overheat in seconds. These things are a furnace! Will easily heat a bedroom.
 
MicrowaveEgg said:
PS3 was never opened, No flux residue on RSX, no tokins been replaced. Will try replacing the tokins myself later this weekend

Here's some pictures
https://imgur.com/a/9Rj3Mnh

UDc9FBZ.jpg

5mHkc6z.jpg
Click to expand...

I don't see any flux residue. If it was reballed or reflowed, they cleaned it well. If a pro did it, there should be a QA sticker. Don't see one.

It looks pretty stock inside to me. Did you notice any signs of previous disassembly? The screws that hold the card reader in are often forgotten. Screws should have that fresh CRACK of having never been unscrewed. Flex cable holding BR and card reader may be creased more on consoles that have been opened multiple times. There is a piece of foam tape on the HS that usually is stuck a bit to the RF shield that will be broken if the HS has been removed before. And of course the white old thermal paste and warranty stickier. Those are the things I note on first disassembly just to be sure the console has or has not been opened before.

The reason I'm wondering is because of that 3034 in your error log. While yes, you currently have a 1002, which we associate with bad tokins, that doesn't mean a BGA defect can't cause it also. If the console had been opened and worked on, that might explain the disappearance of the 3034 and appearance of a 1002. But from what I'm seeing in your pics, that appears to be pretty stock. And concerns me that you still have a BGA defect.

If there is an RSX issue and you replace tokins you may experiance a false positive, where the console works and you feel elated for a few weeks before the YLOD returns with a 3034 in the log. Hope for the best, just be prepared for that possability.
 
Not always. 3034 is usually an issue with the RSX, often a BGA defect, but can also be the RSXRam, die bumps, or even CPU. Reballing is the right way forward for repair, but doesn't always work. Reflowing is inferior, and probably shouldn't be considered repair. More like a band aid on wound that needs stitches. It might stop the bleeding and close the wound temporarily, but it won't hold.
 
...but 3034 is not going to be fixed with tokin replacement. The heat from soldering in tantalum usually makes people think it works (warps the board enough for microscopic solder ball crack to mechanically reconnect). When the strain relaxes in a few thermal cycles (usually less than 2 weeks) the 3034 returns. Happened alot before we started using the SYSCON for diagnosis. Lots of false repairs.
 
RIP-Felix said:
I don't see any flux residue. If it was reballed or reflowed, they cleaned it well. If a pro did it, there should be a QA sticker. Don't see one.

It looks pretty stock inside to me. Did you notice any signs of previous disassembly? The screws that hold the card reader in are often forgotten. Screws should have that fresh CRACK of having never been unscrewed. Flex cable holding BR and card reader may be creased more on consoles that have been opened multiple times. There is a piece of foam tape on the HS that usually is stuck a bit to the RF shield that will be broken if the HS has been removed before. And of course the white old thermal paste and warranty stickier. Those are the things I note on first disassembly just to be sure the console has or has not been opened before.

The reason I'm wondering is because of that 3034 in your error log. While yes, you currently have a 1002, which we associate with bad tokins, that doesn't mean a BGA defect can't cause it also. If the console had been opened and worked on, that might explain the disappearance of the 3034 and appearance of a 1002. But from what I'm seeing in your pics, that appears to be pretty stock. And concerns me that you still have a BGA defect.

If there is an RSX issue and you replace tokins you may experiance a false positive, where the console works and you feel elated for a few weeks before the YLOD returns with a 3034 in the log. Hope for the best, just be prepared for that possability.
Click to expand...

No I forgot to post that still had the warranty seal in tact and everything. Doesn't look like anyone opened anything, The tape and other things look untampered with. In fact the Cell thermal paste is still wet which surprised me.

For the 3034s, it appeared after the 1002 errors, is it possible its the other way around? Could bad caps cause the 3034 errors? Just wondering since 3034s only showed up in two error logs.

I also have 2 other unopened PS3s and like 7 other opened PS3s and maybe more. Will dump those syscon logs after I'm done recapping this one.
 
My advice as always when ppl listen 3034 is something was explained for cell and 44xx is something on imput of rsx.
This guy sorted with both reball, only rsx did not work. It may work for few months with one ic reball but I will always suggest reball both.
9ee1cb979880df8b989e78d53a27f11b.jpg
 
Last edited:
I'm still struggling to improve my reballing process. Honestly yes, if you could could get repeatable reballing results, go ahead and do both CPU/RSX. No sense putting the console through unnecessary reflow cycles later.

But I refute the whole 'data errors indicate RSX, lack of them indicate CPU' hypothesis. I frankensteined a console with only a 3034 and it still works. So the CPU was fine. The issue was the RSX. We're still in the dark with the meaning of a lot of SYSCON codes. And the associations we have made (3034 = RSX issues requiring reball, 1002 = RSX filter/caps) are not causal. We still don't know exactly which voltage, resistance, communication delay, etc that triggers the code. And because of that, the list of problems that can cause an error it likely much larger than we've identified.
 
Yes Felix true, uart debugging is helping like car diagnostic, it is around that defective part point problem, it is helping very much but most people think to rely only on one data error without taking any further measurements.
I've started to see this to often. Not only normal users with only a few knowledge but some people who do repairs they try accelerated processes and fails most often because they don't start in step by step. I wish to take ppl money only to reball one ic, after 3 months another ic. I wish and hate working 8 hours on one board but in the end nothing is failing, except those " special glod" were I can't explain only one ram of rsx missing from schematic.
Edit
Another funny solution today went good in ps4 area.
1c895af07c7352e39ca0aecaf901f2d5.jpg

To quate an very important old user @BwE am I right?
 
Last edited:
Haha...yeah, the video's on YouTube are the sloppy way. Well, most of them. @DeadEnd's frankenstein video is one exception. And there are some decent DIY reballing videos, but none specifically about the PS3 or YLOD. All of the videos about the YLOD and Reflowing are INCREDABLY misleading. Sensationalists looking for views instead of technicians telling truth.

People want a simple solution to complex problems. YouTubers and politicians tell you what you want to hear. Rarely is it that simple!
 
This is indeed all about truth! You will improve with time your reball. One most important thing is keeping practicing, at one point it will be simple and easy. Just make your way step by step, until now all your suggestions are right. I've been quated from a user here for doing Frankenstein mod, based in USA . I will help him because he could not find someone doing it there? Isn't nobody ready to make this mod there? I quate Louis Rossmann to see his fees, he will probably swear me eventually. If someone is sending something here then return cost is more than work fees. Any good suggestions for USA reball both? Or at least rsx 40nm, then with time may reball cell if really needed.
 
You must log in or register to reply here.

Similar threads

Back
Top