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

[RIP-Felix]

I have some buck regulators that I can adjust to 3.3v. Would that work?

Buck converters are noisy and the PS3 is power hungry. I think they would do more harm than good. I would stick to the PS3's PS or a good 80 plus gold or platinum computer supply of 400W or greater.

...I was suspecting the IC voltage regulator mostly.
And now I remember @squeept had a board with abnormal voltage on RSX too. But was a total heatgun catastrophe with toothpaste flux.
My board has most likely been reflowed too, but actually doesn't look that bad to me. Maybe the Voltage regulator IC which is near the RSX could have been damaged by whomever used hot air there?

I would start there. It's trivial to remove the MOSFET and Driver, the IOR VRM is much harder. Start with the previous 2 and see if that helps.

 

[vyktormvmpay25]

Ok when I'm probing like this with the coils missing I'm reading 3 to 4 volts... This is already high right? And therefore means the high voltage problem is not caused by the side of the tokins and is back there as expected
The error is 09 3004 now as expected too

Ok is fine because without that coil (shunt resistance) and without cpu/gpu in parallel voltage is fine, what is wrong with cpu first. What is now resistance to cpu/gpu
Not sure if a mosfet problem or ior, I did not tested before. This should be done before on a working unit. If you want to wait until tomorrow I can tell by disassemble my cok002 with that modchip and test voltage without those coils.

 

[RCWD21]

Buck converters are noisy and the PS3 is power hungry. I think they would do more harm than good. I would stick to the PS3's PS or a good 80 plus gold or platinum computer supply of 400W or greater.

I would start there. It's trivial to remove the MOSFET and Driver, the IOR VRM is much harder. Start with the previous 2 and see if that helps.

Well I do have a spare power supply. Its one if the zssr models that I replaced in my a01.

Do you happen to know the pin out to get it to power on?

And would you recommend dual nand or one by one?

Sent from my SM-G965U using Tapatalk

 

[Pacorretaco]

Ok is fine because without that coil (shunt resistance) and without cpu/gpu in parallel voltage is fine, what is wrong with cpu first. What is now resistance to cpu/gpu
Not sure if a mosfet problem or ior, I did not tested before. This should be done before on a working unit. If you want to wait until tomorrow I can tell by disassemble my cok002 with that modchip and test voltage without those coils.

Resistance of gpu didn't really change. Still 1.8 ohms.
Cpu about 2.4 ohms

So you think the 3 to 4 volts there is normal? That un-narrows the problem.
I could remove the same on other board, wasn't too difficult

 

[RIP-Felix]

It's important to place LC filters correctly or they can interfere with the voltage sensing on the VRM. I'm guessing that's why the voltage increased after you removed them.

 

[vyktormvmpay25]

In electrical power we use triangle of power but yes YT link describes simplified, it's more complex than that with long formula, as long resistance did not change, not to much to be done is a fried cpu, it should be nearly 3.2 ohms.
Those 3 ohms is equivalent to those 3v will split voltage at 1.2v
I should explain simple if I found my manual.

 

[Pacorretaco]

In electrical power we use triangle of power but yes YT link describes simplified, it's more complex than that with long formula, as long resistance did not change, not to much to be done is a fried cpu, it should be nearly 3.2 ohms.
Those 3 ohms is equivalent to those 3v will split voltage at 1.2v

Ahh no man that sounds like voodoo.
I have cpus with 1.8 and machine is working well.
2.4 is acceptable and 3.2 is normal yes
Can't really say cpu is dead yet I think

 

[vyktormvmpay25]

Yes I had that situation with work at 1.8 ohms but not sure how those mosfet /ior are minimal/maximum values that should work. Now thinking to put a resistance in parallel with cpu to trick that 3 ohms, never test it.
@RIP-Felix what do you think will it work if he gets a 3 ohms on parallel with cpu?
No won't work.

 

[vyktormvmpay25]

Looking by my scrap board, do you get value for resistance of 3,5 ohms here?

 

[marciolsf]

http://s.go.ro/ax49drsu
I have add few schematics. Should be there.
Photos with tristate point.

This is fascinating! I swear the wiki says that you can't tristate cok001 boards. This is a very cool discovery (for me anyway)

 

[vyktormvmpay25]

I may be wrong but without those coils on cpu and add a 3.2 ohms resistance to those points and gnd you will see 1.2 v on those points of distribution, just in order to be sure cpu is powered fine. I will test this. This board is show kind of clock generator but won't give to much attention, already changed all clock ic beside sata.
It should get 1.2v here if they are common line after those 3 coils are removed and a resistance of 3.2 ohms is in parallel to gng. Then you can be sure power line is good.

 

[vyktormvmpay25]

This is fascinating! I swear the wiki says that you can't tristate cok001 boards. This is a very cool discovery (for me anyway)

Nobody tested under 3.21 otheos?
Yes there are different situations where you can, its not working at 3.55 not sure 3.30 if I've tested.

 

[marciolsf]

Nobody tested under 3.21 otheos?

I don't know about that. Early on as part of my research, I wanted to try setting the tristate on a cok-001 just to see what information I could get. But after a bunch of research, it seemed to me that you could only enable tristate on versions newer slims... The wiki definitely doesn't have the points you have above (last time I checked, anyway).

 

[Pacorretaco]

Looking by my scrap board, do you get value for resistance of 3,5 ohms here?

Shows 3.6 ohm there

I may be wrong but without those coils on cpu and add a 3.2 ohms resistance to those points and gnd you will see 1.2 v on those points of distribution, just in order to be sure cpu is powered fine. I will test this. This board is show kind of clock generator but won't give to much attention, already changed all clock ic beside sata.
It should get 1.2v here if they are common line after those 3 coils are removed and a resistance of 3.2 ohms is in parallel to gng. Then you can be sure power line is good.

Wait maybe there is a confusion here? This COK 001 board has 23 1002 error, and I noticed abnormal high voltage on RSX. I removed coils leading to RSX.
Maybe confused RSX with CPU?
My CPU is getting normal around 1v
I didn't really suspect problems with CPU side.

 

[RCWD21]

I don't know about that. Early on as part of my research, I wanted to try setting the tristate on a cok-001 just to see what information I could get. But after a bunch of research, it seemed to me that you could only enable tristate on versions newer slims... The wiki definitely doesn't have the points you have above (last time I checked, anyway).

Well mine was originally on 2.35 when all this started and assuming its not fully overwritten grounding tristate worked perfectly on my board. Now to just get everything hooked up to dump the nand [emoji28]

Sent from my SM-G965U using Tapatalk

 

[vyktormvmpay25]

Thanks to good developers that keep working hard trying to keep live those best info
https://www.psdevwiki.com/ps3/Teensy++_2.0
All his hard work.

 

[RCWD21]

Hey guys quick question! Is the chip on the top nand 0 and the nand on the bottom nand 1?

Sent from my SM-G965U using Tapatalk

 

[vyktormvmpay25]

Hey guys quick question! Is the chip on the top nand 0 and the nand on the bottom nand 1?

Sent from my SM-G965U using Tapatalk

Nand 0 is beside psata connection for bd drive.
It will be hard to read without clip ziff over nands. If you want to start first need to understand where all wires goes. Nands are harder then nor. A bad flash will ylod unit.

 

[RIP-Felix]

Yes I had that situation with work at 1.8 ohms but not sure how those mosfet /ior are minimal/maximum values that should work. Now thinking to put a resistance in parallel with cpu to trick that 3 ohms, never test it.
@RIP-Felix what do you think will it work if he gets a 3 ohms on parallel with cpu?
No won't work.

Anything above 1 Ohm should work. 1-2 ohms is precarious though. Since he said the board looked reflowed, we should consider the possibility it was a bad job. I noticed that flux residues can lower this resistance, so be sure the area is thoroughly clean. I use Electronics Cleaner (some harsh solvents) to get under the BGA. Then I finish off with 99% IPA. The soldering can affect these resistance too. My resistance went way up after reflowing these joints (and thorough cleaning). 1.8ohms seems unhealthy, but not quite dead. Only way to be sure is to pull the chip and ohm test it. Or you could try a reflow to see if it improves.

You could try a parasitic TaPol to see how it affects the behavior. No harm if you wire it up in parallel. Even if you have to scrape away some solder mask to get a good pad to solder to, it's not the end of the world.

 

[Pacorretaco]

Anything above 1 Ohm should work. 1-2 ohms is precarious though. Since he said the board looked reflowed, we should consider the possibility it was a bad job. I noticed that flux residues can lower this resistance, so be sure the area is thoroughly clean. I use Electronics Cleaner (some harsh solvents) to get under the BGA. Then I finish off with 99% IPA. The soldering can affect these resistance too. My resistance went way up after reflowing these joints (and thorough cleaning). 1.8ohms seems unhealthy, but not quite dead. Only way to be sure is to pull the chip and ohm test it. Or you could try a reflow to see if it improves.

You could try a parasitic TaPol to see how it affects the behavior. No harm if you wire it up in parallel. Even if you have to scrape away some solder mask to get a good pad to solder to, it's not the end of the world.

Ok, but are we talking about the CPU or the RSX now?
The RSX right? I'm a bit confused.

So ok, you mean the test removing the coils and reading 4 volts on there is normal?
I can answer tomorrow but that's the small question I have now.
Because if it is normal then sure, the tokins and the rsx itself are not ruled out. Time to become a caveman and to a RSX tokin job.

But if it's already not normal then it means I should instead begin replacing the IOR regulators and such.