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

[DeCode_762]

Hello everyone,

A few weeks ago I bought a FAT CECHG01 from a thrift store and found it GLOD'd when I first tried it out.
Read up on troubleshooting and here's my findings:

ofst[ 64]:err_code:0xffffffff, clock:0x1778fdb7  2012/06/23 22:20:39
ofst[ 68]:err_code:0xffffffff, clock:0x177944ad  2012/06/24 03:23:25
ofst[ 72]:err_code:0xa0801701, clock:0x178b509a  2012/07/07 19:55:06
ofst[ 76]:err_code:0xa0801001, clock:0x17cfc784  2012/08/28 18:16:36
ofst[ 80]:err_code:0xa0801001, clock:0x17cfcc7b  2012/08/28 18:37:47
ofst[ 84]:err_code:0xa0801002, clock:0x17cfce02  2012/08/28 18:44:18
ofst[ 88]:err_code:0xa0801701, clock:0x17cfcff5  2012/08/28 18:52:37
ofst[ 92]:err_code:0xa0801001, clock:0x14aa7e5d  2010/12/26 22:29:49
ofst[ 96]:err_code:0xa08014ff, clock:0x17dedeee  2012/09/09 05:00:30
ofst[100]:err_code:0xa0801701, clock:0x14ac4c19  2010/12/28 07:19:53
ofst[104]:err_code:0xa0801701, clock:0x17e24eb6  2012/09/11 19:34:14
ofst[108]:err_code:0xa0801701, clock:0x17e392c7  2012/09/12 18:36:55
ofst[112]:err_code:0xa0802022, clock:0x17eb9fcb  2012/09/18 21:10:35
ofst[116]:err_code:0xa0802022, clock:0x17eb9fcc  2012/09/18 21:10:36
ofst[120]:err_code:0xa0802022, clock:0x17eba165  2012/09/18 21:17:25
ofst[124]:err_code:0xa0802022, clock:0x17eba166  2012/09/18 21:17:26
ofst[  0]:err_code:0xa0802022, clock:0x17ec06ec  2012/09/19 04:30:36
ofst[  4]:err_code:0xa0802022, clock:0x17ec06ed  2012/09/19 04:30:37
ofst[  8]:err_code:0xa0802022, clock:0x17ec07cc  2012/09/19 04:34:20
ofst[ 12]:err_code:0xa0802022, clock:0x17ec07ce  2012/09/19 04:34:22
ofst[ 16]:err_code:0xa0801701, clock:0x17efb750  2012/09/21 23:40:00
ofst[ 20]:err_code:0xa0801001, clock:0x17f01522  2012/09/22 06:20:18
ofst[ 24]:err_code:0xa08014ff, clock:0x17f0d7f9  2012/09/22 20:11:37
ofst[ 28]:err_code:0xa0801701, clock:0x17fe1b95  2012/10/02 21:39:33
ofst[ 32]:err_code:0xa0801701, clock:0x17ff31af  2012/10/03 17:26:07
ofst[ 36]:err_code:0xa08014ff, clock:0x18010928  2012/10/05 02:57:44
ofst[ 40]:err_code:0xa0801701, clock:0x1807a89f  2012/10/10 03:31:43
ofst[ 44]:err_code:0xa0801701, clock:0x180849eb  2012/10/10 14:59:55
ofst[ 48]:err_code:0xa0801701, clock:0x181cfb95  2012/10/26 07:43:17
ofst[ 52]:err_code:0xa0603040, clock:0x19c6763e  2013/09/14 01:19:58
ofst[ 56]:err_code:0xa0603040, clock:0x19c67658  2013/09/14 01:20:24
ofst[ 60]:err_code:0xa0603040, clock:0x19c67660  2013/09/14 01:20:32

Bringup : 1280 times
Shutdown: 1138 times
Power-on: 73day 16hour 49min 12sec

I think the console died about 10 years ago and was never powered-on since.
Leaving the console on never results in a forced shutdown or error, it just stays GLOD'd indefinitely.
I'm very new to PS3 troubleshooting, but I did do some reading and would like to know if you guys think I'm on the right track.

I'm thinking the earlier 1001 errors were just improper shutdowns.
The 1002 error might be faulty TOKINs.
The fix for the 2022 would be replacement of the AV OUT IC.
As for the 3040 error, I'm wondering if this would be fixed by the TOKINs being replaced.
I'm hoping that the RSX doesn't require re-balling, as I don't have the tools to do that kind of work :'(

Any advice would be greatly appreciated!

Cheers

 

[No_One]

So a little update after doing some manual scanning of the board I found this gash close to the 12v power supply connector.

Initially I wrote it off as welp that's it this is dead time to look for a new board. but after stewing it over I convinced myself that it was worth a try to solder a jumper wire for an attempted trace repair. After all it such a small break in the trace how hard could it be...

At this point I'm just hopping I didn't brick anything trying to use a solder tip and jumper wire that was way to big for this. It's back to the drawing board and I still am unaware if this is even related to my 3001 (12v power failure) error. but might as well post an update so my "progress" can be tracked.

 

[Polynian]

Hi i got one cechg04 already delidded with wire. İt has 3 types of ylod i get the errorlog and it showed bunch of 1002 and 1200 and İt gives the 1200 when the consoles fan turn at max speed for a sec and shuts and it gives the 1002 when the fan turns so slow for a sec and shuts , i searched it and people say 1002 can be tokins and just for a try i heat the tokins for a test and all the 1002 1200 errors are gone now it just gives 3032 everytime. Btw this is my second time trying the heating the tokin method, the first time 3032 error last for 3 days and it turned back to 1002 1200. ( Now its still 3032 it can turn back to 1xxx any time) İ assume if i change the tokins everything except 3032 is going to get fixed but what the hell is 3032 error means can you guys help please. (Btw i get 2203 south bridge error for 4 or 5 times but it never came back after that)

 

[RIP-Felix]

Alot of the time 3001 is caused by not plugging the 12v connector in when testing SYSCON errorlogs. The lack of times stamps is normal after unplugging the clock battery. Any timestamped codes will be erased by new errors every time the console is tested and it generates a new code without one.

Sounds like previous repair attemps caused more than one issue. OG issue is OFC the GPU (most likely). Now you have a 12v failure on top of that. Check fuse, inspect for damage, troubleshoot the circuit...Question life descisions.

If you clear the 3001 (get a temperature controlled soldering iron), you'll still have to deal with the GPU.

You're faced with a descision. What is your motivation for fixing this console? The fun of the project? If so, be warned it is an expensive a difficult project that more than likely will kill a few motherboard in the process.

If you just want a frankie, you will save money sending it for the servive.

 

[RIP-Felix]

The fix for the 2022 would be replacement of the AV OUT IC.
As for the 3040 error, I'm wondering if this would be fixed by the TOKINs being replaced.

In this case I would ignore the 2022. It's common to see generated at the same time as other, more diagnostic codes.

The 603040 has to do with Flash (NAND). It could explain your GLOD. If they lose power it'll generate 3040. If it's corrupted it can generate 3040. Likely what happened is someone attempted to update in safe mode to resolve the GLOD, which was caused by a dead or dying GPU. It failed during the update, corrupting the NAND and causing the 3040.

The 1701/1601 is common in 90nm models when the GPU is dying. 1001 is like 2022 and 2120. They're often generated at the same time and aren't diagnostic. Just associated. Like ancillary errors that can occur due to the instability cause by the actual issue.

1002 however is diagnostic. I am not aware of an instance where it hasn't been caused by RSX core voltage instability. Tokins being chief suspect (but not the only one).

 

[RIP-Felix]

I still am unaware if this is even related to my 3001 (12v power failure) error. but might as well post an update so my "progress" can be tracked.

It probably is related. I don't have the schematic open in front of me ATM, but you can DL it and find out what that broken signal is and where that ground plane extends to. If it's related to the 12v line, which is likely given it's proximity to 12v section, then that probably does explain it.

Once repaired, you'll still have to deal with the RSX.

 

[RIP-Felix]

Hi i got one cechg04 already delidded with wire. İt has 3 types of ylod i get the errorlog and it showed bunch of 1002 and 1200 and İt gives the 1200 when the consoles fan turn at max speed for a sec and shuts and it gives the 1002 when the fan turns so slow for a sec and shuts , i searched it and people say 1002 can be tokins and just for a try i heat the tokins for a test and all the 1002 1200 errors are gone now it just gives 3032 everytime. Btw this is my second time trying the heating the tokin method, the first time 3032 error last for 3 days and it turned back to 1002 1200. ( Now its still 3032 it can turn back to 1xxx any time) İ assume if i change the tokins everything except 3032 is going to get fixed but what the hell is 3032 error means can you guys help please. (Btw i get 2203 south bridge error for 4 or 5 times but it never came back after that)

Ok, what's happeing is this. You have 2 issues.

1. Bad RSX Tokins
2. CPU overheat

When you heat the tokins you're not just heating them, you're also heating the area of the board near them. Whci includes the MLCC caps underneath the RSX and OFC the GPU itself. This changes the parasitic inductance and capacitance of the circuit (Core voltage).

1002 is an RSX core voltage instability. When the filtering degrades far enough to begin triggering the Lower voltage drop-out or upper voltage threshold (+100mV or -163mV) it the VRM falls out of regulation and issues a no power good signal and the SYSCON steps in to shut down. Issuing a 1002 (RSX GPU) or 1001 (CELL CPU).

When this just starts hapening, it is right on the edge of working and not working. And any small change in parasitic inductane or capacitance can affect the outcome. So by adding heat you change this ballance and appears to work, temporarily.

But you still have an overheating cpu. And I should warn you that BGA and Bump defects (among other issues, like VRAM failure) can do the same thing. But I don't believe that's happening in your case. The 3032 is a Thermal monitor error, because the MB temp was too high, you probably tested the console while it was still hot.

So my diagnosis is a CPU delid and RSX tokin replacment are needed.

 

[Fallen777]

Ok, what's happeing is this. You have 2 issues.

1. Bad RSX Tokins
2. CPU overheat

When you heat the tokins you're not just heating them, you're also heating the area of the board near them. Whci includes the MLCC caps underneath the RSX and OFC the GPU itself. This changes the parasitic inductance and capacitance of the circuit (Core voltage).

1002 is an RSX core voltage instability. When the filtering degrades far enough to begin triggering the Lower voltage drop-out or upper voltage threshold (+100mV or -163mV) it the VRM falls out of regulation and issues a no power good signal and the SYSCON steps in to shut down. Issuing a 1002 (RSX GPU) or 1001 (CELL CPU).

When this just starts hapening, it is right on the edge of working and not working. And any small change in parasitic inductane or capacitance can affect the outcome. So by adding heat you change this ballance and appears to work, temporarily.

But you still have an overheating cpu. And I should warn you that BGA and Bump defects (among other issues, like VRAM failure) can do the same thing. But I don't believe that's happening in your case. The 3032 is a Thermal monitor error, because the MB temp was too high, you probably tested the console while it was still hot.

So my diagnosis is a CPU delid and RSX tokin replacment are needed.

I would say a 1200 is not always gonna be cpu overheat if the syscon returns normal tmp in uart script then you are in chase for the thermal alert circuit afterall it involves a resistor for signal and a capacitor as filter and we all know caps just have a thing for shorting out. Also if you have a superslim board with cell removed send me a photo of bga pads showing their surrounding trace you know what I mean so I can complete the circuit I found the capacitor now I need to find the resistor

 

[No_One]

You're faced with a decision. What is your motivation for fixing this console? The fun of the project?

If you just want a frankie, you will save money sending it for the service.

Yeah, I'm at my limit. I'm a novice and having a potentially working backwards compatible PS3 is more valuable to me than the repair experience. It probably would be a fun project but I will likely kill a potentially fixable board. for now I'll log everything I can find that needs to/should be addressed so I can pass it along to the servicer so they can get a good picture of the console.

With that said do you know of people offering the service in the States preferably around the Chicago land area? though I'll take anyone in the US tbh.

 

[Sampsonay]

Yeah, I'm at my limit. I'm a novice and having a potentially working backwards compatible PS3 is more valuable to me than the repair experience. It probably would be a fun project but I will likely kill a potentially fixable board. for now I'll log everything I can find that needs to/should be addressed so I can pass it along to the servicer so they can get a good picture of the console.

With that said do you know of people offering the service in the States preferably around the Chicago land area? though I'll take anyone in the US tbh.

Feel free to check out my store page, I can fix any ripped traces at no additional charge if you're looking to get the GPU replacement service.
*** DO NOT ADVERSITE: LINK REMOVED BY MODERATOR ***

 

[RIP-Felix]

With that said do you know of people offering the service in the States preferably around the Chicago land area? though I'll take anyone in the US tbh.

Well, there are a few options that come to mind that are closer to you. There's @Sampsonay (*** DO NOT ADVERSITE: LINK REMOVED BY MODERATOR ***). He's been around here forever and a friend of mine. Then there's the console king (*** DO NOT ADVERSITE: LINK REMOVED BY MODERATOR ***). I helped him learn the process and he's crank'n franken ATM. Then there's computer booter. Those are all the US vendors I can think of off the top of my head, that I know and trust (that offer it as a service). There are those who sell them premade like @squeept, who does probably the most premium level of work...and demands a price comensuate to that quality.

 

[Polynian]

Ok, what's happeing is this. You have 2 issues.

1. Bad RSX Tokins
2. CPU overheat

When you heat the tokins you're not just heating them, you're also heating the area of the board near them. Whci includes the MLCC caps underneath the RSX and OFC the GPU itself. This changes the parasitic inductance and capacitance of the circuit (Core voltage).

1002 is an RSX core voltage instability. When the filtering degrades far enough to begin triggering the Lower voltage drop-out or upper voltage threshold (+100mV or -163mV) it the VRM falls out of regulation and issues a no power good signal and the SYSCON steps in to shut down. Issuing a 1002 (RSX GPU) or 1001 (CELL CPU).

When this just starts hapening, it is right on the edge of working and not working. And any small change in parasitic inductane or capacitance can affect the outcome. So by adding heat you change this ballance and appears to work, temporarily.

But you still have an overheating cpu. And I should warn you that BGA and Bump defects (among other issues, like VRAM failure) can do the same thing. But I don't believe that's happening in your case. The 3032 is a Thermal monitor error, because the MB temp was too high, you probably tested the console while it was still hot.

So my diagnosis is a CPU delid and RSX tokin replacment are needed.

3032 is going for at least 3 days im not starting the console or getting errorlog when its heated, i already delidded with wire and i repad and repasted everything nothing is overheating. İ thought maybe it can be the tempreture sensor causing the 1200 error but im not sure about that. When i heat the tokins console is giving 3032 for 3 days thats not heat or something, i assume this as temporary fix for tokins but what am i need to do for 3032. İf i was able to boot the console after fixing this 3032 im going to change the caps for permanent fix.

 

[No_One]

Feel free to check out my store page, I can fix any ripped traces at no additional charge if you're looking to get the GPU replacement service.
*** DO NOT ADVERSITE: LINK REMOVED BY MODERATOR ***

I can't confirm the Firmware of the nand. Are you still interested?

 

[RIP-Felix]

I had to jog my memory. Looked back over some notes and alot of people are having 3032 errors from bad delids or eraser mods gone wrong. Cracked CPU die and etc. You should inspect the area around the CPU carefully for any SMDs that are out of place and the die itself for any hairline cracks. You could try a pressure test to see if it's just a CPU BGA as well. You could also check that +1.5V_YC_RC_VDDA is present and making it into the CPU. If so it shouldn't be the lack of VDDA causing it.

 

[RIP-Felix]

I can't confirm the Firmware of the nand. Are you still interested?

He knows how to flash the nand if he has to. It's part of the frankie process and part of why it's necessarily expensive. The extra headache you might have to do every now and again factors into the price.

 

[TurboYoda]

The symptoms are consistant with a dead RSX. Seeemingly confirmed by the 1802 syscon errorcode and the other codes associated with RSX issues (1701/1601/14FF). I'd reccomend ohm testing RSX_FBVDDQ (VRAM voltage) to see if it's dead.

The Blue Screen of Death you can learn about here. Scroll down to BSOD. I'm sure others are more knowlegable than I am about NOR corruption errors, so I'll let them chime in. But what I suspect happend is you had a normal GLOD caused by a failed GPU and misdiagnosied it as a failed update. And in your attempts to update while the RSX was in that bad state, you caused a BSOD from corruption.

EDIT: Worth mentioning it could be due to a bad HDD too. But the error codes don't fit with that IMO.

Is it possible to perform ohm test on the board itself without desoldering the RSX?
Apologies if this is a silly question.

BTW thank you so much for your contribution.

 

[RIP-Felix]

 

[Oxihixo]

I have two CECHA-01 systems that both suffer from the YLOD. One was sent to a repair facility back in 2018. They deemed it to be not repairable and put it back together. I decided to take it apart to have a look for myself. What I found is that two RSX NEC/Tokins were missing, the RSX has been delidded and new thermal paste added. Maybe it's an illusion, but it looks like the RSX isn't sitting parallel and has a slightly bigger gap on one end. I suspect they tried to re-flow it and they lacked the experience. I've included the SYSCON logs for this machine. Can someone with experience take a look at the log and let me know if this machine can be salvaged?

errlog
ofst[  0]:err_code:0xa0093003, clock:0xffffffff
ofst[  4]:err_code:0xa0093003, clock:0xffffffff
ofst[  8]:err_code:0xa0093003, clock:0xffffffff
ofst[ 12]:err_code:0xa0093003, clock:0xffffffff
ofst[ 16]:err_code:0xa0093003, clock:0xffffffff
ofst[ 20]:err_code:0xa0093003, clock:0xffffffff
ofst[ 24]:err_code:0xa0093003, clock:0xffffffff
ofst[ 28]:err_code:0xa0093003, clock:0xffffffff
ofst[ 32]:err_code:0xa0093003, clock:0xffffffff
ofst[ 36]:err_code:0xa0093003, clock:0xffffffff
ofst[ 40]:err_code:0xa0202120, clock:0xffffffff
ofst[ 44]:err_code:0xa0202120, clock:0xffffffff
ofst[ 48]:err_code:0xa0202120, clock:0xffffffff
ofst[ 52]:err_code:0xa0202120, clock:0xffffffff
ofst[ 56]:err_code:0xa0202120, clock:0xffffffff
ofst[ 60]:err_code:0xa0202120, clock:0xffffffff
ofst[ 64]:err_code:0xa0202120, clock:0xffffffff
ofst[ 68]:err_code:0xa0202120, clock:0xffffffff
ofst[ 72]:err_code:0xa0202120, clock:0xffffffff
ofst[ 76]:err_code:0xa0202120, clock:0xffffffff
ofst[ 80]:err_code:0xa0203010, clock:0xffffffff
ofst[ 84]:err_code:0xa0202120, clock:0xffffffff
ofst[ 88]:err_code:0xa0202120, clock:0xffffffff
ofst[ 92]:err_code:0xa0202120, clock:0xffffffff
ofst[ 96]:err_code:0xa0202120, clock:0xffffffff
ofst[100]:err_code:0xa0202120, clock:0xffffffff
ofst[104]:err_code:0xa0202120, clock:0xffffffff
ofst[108]:err_code:0xa0202120, clock:0xffffffff
ofst[112]:err_code:0xa0202120, clock:0xffffffff
ofst[116]:err_code:0xa0202120, clock:0xffffffff
ofst[120]:err_code:0xa0202120, clock:0xffffffff
ofst[124]:err_code:0xa0203010, clock:0xffffffff


bringup
[SSM] state: 0000 -> 0101
Bringup Mode #0 (0xFF)
[SSM] ssmCb_OnStartingBePowOn() called.
[SSM] First Boot.
[SSM] Bringup mode : syspm_stat=00000000/00000000
[POWSEQ] PowerSeq_Setup called.
[SSM] state: 0101 -> 0201
[POWSEQ] AV Backend Setup
[SSM] state: 0201 -> 0102
[ERROR]: 0xa0202120
[ERROR]: 0xa0202120
[ERROR]: 0xa0202120
[ERROR]: 0xa0202120
[ERROR]: 0xa0202120
[ERROR]: 0xa0202120
Press Ctrl+C to exit
>$ shutdown
[ERROR]: 0xa0202120
[ERROR]: 0xa0202120
[ERROR]: 0xa0202120
[ERROR]: 0xa0202120
[POWERSEQ] Error : wait attention timeout.(SEQ1)
[SSM] state: 0102 -> 0302
[SSM] PowSeq Fail : Detected !
[SSM] state: 0302 -> 0700
[POWSEQ] AV Backend Letup
[SSM] Shutdown mode : syspm_stat=00000000/00000000
[ERROR]: 0xa0203010
[POWSEQ] PowerSeq_Letup called.
[SSM] state: 0700 -> 0600
(PowerOff State) (Fatal)
shutdown
[SSM] state: 0600 -> 0000
[SSM] Error state is cleared.
(PowerOff State)


becount
Bringup : 4008 times
Shutdown: 3968 times
Power-on: 270day 02hour 25min 05sec

Additional questions.
1. The other machine that I have has half the power-on time (130 days) and only one SYSCON error, 0403034. Might this be an easy fix?
2. Is there anyone in Canada that is repairing these consoles?
3. @RIP-Felix, would you have interest in repairing one of these?

Ideally, I'd like to have the Frankenstein mod performed on one of these. Would it make sense to do this on the one that's missing the NEC Tokins since it's already been tampered with? Or would you recommend doing this mod to the other machine since it hasn't been tampered with?

In ending, I'd like to thank all those that have provided information and spent countless hours reverse engineering all things PS3. While I cannot name everybody, you know who you are. Thank you for keeping the mighty PS3 alive and making it better.

 

[RIP-Felix]

The RSX likely needs replaced and tokins replaced before we can see what errors remain. 3010 isn't a good error. It's one we oftem associate with a dead CPU, but that's not necessarily how this one was generated. I would agree with that shops descision not to work on that console. It has clearly been molested by butcher. Most shops wont work on them because they are a rabbit hole waste of time.

That said it may be salvageable for someone wanting to challenge themselves, but you will not likely find someone willing to offer any sort of guarintee. And they will rightly charge an attempt fee. I would consider it a donor board. A good candidate for learning to rework a PS3. There's still a chance, but if not you don't have to feel bad since it probably wasn't going to work anyway.

EDIT: I was just looking back over my spreadsheet and found one console that had 3010 after RSX_FBVDDQ was short. Could you check if your's is? Referr to the motherboard Ohms test location in the picture a few posts back.

 

[RIP-Felix]

Additional questions.
1. The other machine that I have has half the power-on time (130 days) and only one SYSCON error, 0403034. Might this be an easy fix?
2. Is there anyone in Canada that is repairing these consoles?
3. @RIP-Felix, would you have interest in repairing one of these?

Ideally, I'd like to have the Frankenstein mod performed on one of these. Would it make sense to do this on the one that's missing the NEC Tokins since it's already been tampered with? Or would you recommend doing this mod to the other machine since it hasn't been tampered with?

In ending, I'd like to thank all those that have provided information and spent countless hours reverse engineering all things PS3. While I cannot name everybody, you know who you are. Thank you for keeping the mighty PS3 alive and making it better.

1. No. BGA rework is probably the highest level of rework difficulty besides micro trace rapair. Having said that, with the right equipment and experience it's definitely doable.
   
2. I can't think of anyone off the top of my head.
   
3. I don't do them as a service. I have too many as it is I need to get to.

If you're contemplating paying for one of the services, choose a console that has the best chances of working. An unmolested console is preferable. Remember these services charge an attempt fee. If it doesn't work you lose that fee. So there is a chance you may have to try again and again before one works. And by that point it may cost you more than simply buying one that was successfully frankied already. That's the reason the service is cheaper, they don't have to buy the console and take the risk. You do. But if you are willing to take that gambit, it's best to stack the odds in your favor instead of shooting yourself in the foot.