PS3 Frankenstein PHAT PS3: CECHA with 40nm RSX

Guys, if you want to focus on the cooling talk, perhaps you need to make a separate thread.

Gus21 said:
Hello everyone, has anyone tested a 40nm RSX without IHS on a cecha01, does the Cecha01's cooler come in contact with the die and the memories dissipating heat?
Click to expand...

If you're talking about the one found on SS models. I think Victor mentioned it's a bit lower than a regular 40nm. I might have it somewhere, I will try to measure the difference.
 
Just a question that Victor might be able to answer or anyone else. Any recommendations on good BGA rework stations or whether the more expensive machines do a substantially better job? I can't find anyone who offers it as a service in my country and I'm prepared to save up a couple of years if it'll get a better end result

I'm in electronics so it's a skill I've been interested in developing, not just for this
 
Have you tried cleaning and putting any ihs from previous models on top, looking at light on it and see space? If you don't have any space on die means is probably older version idk. We found that model with lower level probably.
 
DeadEnd said:
I'm not sure if it's been mentioned, but I have found one cause for errors 2120 and 3013. I'd like to thank whoever wrote the explanation in the devwiki https://www.psdevwiki.com/ps3/Syscon_Error_Codes#2120_.28HDMI.29

It my case it wasn't the fuse F6302 but a 5.6 ohm resistor R6345. I first noticed it had lost contact on one end. I touched it a little and it came off out of its place entirely. I have also tried to take the same one out of a spare board, but it nearly turned into dust as I touched it. It seems these little guys get a lot of stress with use.

View attachment 36530
View attachment 36531
Click to expand...

In addition to this example. I can also confirm that 10x 202120 with 1x 213013 can be caused by a blown fuse F6301. It's placed in the 12 line that is needed for forming 1.8v FBVDDQ down the line.
2120-3013 when FB6301 is blown.JPG

fuseJPG.JPG


And before I forget. It's also worth mentioning another cause that could create data errors 3034 with 40xx and Bittraining stuck on FlexID after newly 40nm is installed and the syscon code is modified. I recreated sony's method for swapping BD3520FMV to BD3504FMV voltage regulator to drop 1.2v to 0.95v, but later I had a little accident when I dropped the probe on the board, which actually blew the fuse above and then later gave me 3034... Anyways, Bittraining could not happen because 0.95v wasn't fed into RSX at all. I later speculated that it could have been also due to some of the newly installed resistors losing contact (I really suck at microsoldering), or perhaps the regulator itself just died... So syscon throws FlexID errors making you think RSX can't be recognized . I reality it can't start the training because there is no power for the Flex Core present... So yeah. Syscon is not very specific. Be careful when messing with voltages...

when no 0.95v present.JPG
 
Last edited:
Can anyone confirm if the exchange is correct?
RSX PWRGOOD R6205 = 20K by 27K and R6207 = 15K by 10K
CELL PWRGOOD R6111 = 20K by 27K and R6112 = 15K by 10K
 
O diag conectar onde nesta placa? alguém pode ajudar?
Assisti vídeos no you tube onde as pessoas só usam o rx tx e gnd e não usam o diag.
81DX9spnhBL._AC_SL1500_.jpg
 
Gus21 said:
Can anyone confirm if the exchange is correct?
RSX PWRGOOD R6205 = 20K by 27K and R6207 = 15K by 10K
CELL PWRGOOD R6111 = 20K by 27K and R6112 = 15K by 10K
Click to expand...

You got it right. But the other day I've talked with @RIP-Felix about lower thresholds and also a few other people. He convinced me that modifying pwrgood may not be such a good idea after all. It appears Sony also saw the capacitors being another potential cause for YLODs but instead of replacing them, they've decided to cut corners and lowered the thresholds to allow more ripple noise in (more unfiltered voltage) to enter into the RSX and Cell. This way, YLOD will not happen but at the same time it reduces processors' life over time. They need clean DC voltage, any high ripple (AC voltage) spikes will hurt it in the long run. Possible reason for not replacing Tokins was that they didn't want to risk BGA integrity and also Nec/Tokins were expensive at the time. It should be better to just add parasitic caps in parallel. Which will filter the voltage better without having to change lower threshold limits.

Here I have attempted to demonstrate it.

https://www.psx-place.com/threads/r...s-replacement-ylod.25260/page-250#post-335800

I will try to record a video where I talk about it in more detail.
 
Last edited:
DeadEnd said:
You got it right. But the other day I've talked with @RIP-Felix about lower thresholds and also a few other people. He convinced me that modifying pwrgood may not be such a good idea after all. It appears Sony also saw the capacitors causing YLODs but instead of replacing them, they've decided to cut corners and lowered the thresholds to allow more ripple noise in (more unfiltered voltage) to enter into the RSX and Cell. This way, YLOD will not happen but at the same time it reduces processors' life over time. They need clean DC voltage, any high ripple (AC voltage) spikes will hurt it in the long run. Possible reason for not replacing Tokins was that they didn't want to risk BGA integrity and also Nec/Tokins were expensive at the time. It should be better to just add parasitic caps in parallel. Which will filter the voltage better without having to change lower threshold limits.

Here I have attempted to demonstrate it.

https://www.psx-place.com/threads/r...s-replacement-ylod.25260/page-250#post-335800

I will try to record a video where I talk about it in more detail.
Click to expand...

Thanks
 
RIP-Felix said:
I believe it's a silicone oil. TechIngredients did a video not long back that describes how they made one.

Click to expand...


Hello, a RSX 90nm requires 4,800UF, the switch is made to a 40nm one, will I have to have 4,000UF?
We know that RSX 90nm is 4,800UF and 65nm is 4,000UF for 40nm is ??????UF ?
 
Gus21 said:
Hello, a RSX 90nm requires 4,800UF, the switch is made to a 40nm one, will I have to have 4,000UF?
We know that RSX 90nm is 4,800UF and 65nm is 4,000UF for 40nm is ??????UF ?
Click to expand...
No, the ripple/noise the filter is attenuating is generated by the Buck converters on the MB. So we size the filter to the same specification of the MB, not the chip. However, because the 65nm or 40nm RSX draws less current than a 90nm the Buck converters will not operate as heavily, generating less ripple/noise. This means the noise floor will actually be lower than it would for a 90nm. This means the 65nm or 40nm chips will last even longer than they would otherwise. Possibly even longer than they would in a slim.
 
RIP-Felix said:
No, the ripple/noise the filter is attenuating is generated by the Buck converters on the MB. So we size the filter to the same specification of the MB, not the chip. However, because the 65nm or 40nm RSX draws less current than a 90nm the Buck converters will not operate as heavily, generating less ripple/noise. This means the noise floor will actually be lower than it would for a 90nm. This means the 65nm or 40nm chips will last even longer than they would otherwise. Possibly even longer than they would in a slim.
Click to expand...

So I can use 4.800UF correct?
Thanks.
 
Gus21 said:
So I can use 4.800UF correct?
Thanks.
Click to expand...
4,800uF is the minumum. I use 12x 470uF caps, which is over, but that doesn't seem to be detrimental. The added capacitance does actually help. You can overdo it, but it's not a big deal going upto around 6000uF.
 
Hey everyone, so last month I purchased a PS3 CECHA01 from a buy and sell shop for $200 CAD. It had a warranty seal, which was one of the reasons I wanted it and later on in life, get someone to swap the RSX with 40/65nm. Fan wasn't loud but it was noticeable after few minutes. Like level 3. I opened it up, ruined the seal in the process. Found bunch of stickers on Blu-ray drive, indicating that it has been serviced, looked like Sony's work.

As soon as I went to the motherboard, it had tiny residue of flux left around the RSX and someone had written 40nm on the board. I was genuinely surprised, saw a few caps on the back side of the board to be diagonal and not straight.
It had the CXD5301DGB, which is a 40nm chip used in the earlier slim models.
I was quite happy to find that out and surprised that Sony was doing it? I posted it on a Facebook console repair group and someone there mentioned that Sony was doing swaps on some refurbished units.
I sent all of it to Mystic Ryan from YouTube as well, sent a DM on Twitter and he wanted me to share it here on this forum.

Since then, I've replaced thermal paste with Kryonaut and replaced each and every thermal pad. I haven't done any de-lid on it, I've killed one PS3 in the past and I don't wanna kill this super rare system now. I just wish I could find someone in Toronto to get Cell and RSX de-lid just so it has more life and runs a bit quieter, super used to of ps5 being super quiet now.

I haven't modded it or installed any CFW, I tend to keep it stock and play some games online. Also, replaced the 250GB HDD it came with, using a 500GB SSHD now.
kb423gv


kb423gv


kb423gv

https://imgur.com/gallery/kb423gv
kb423gv


kb423gv
 
Izhan Kayani said:
Hey everyone, so last month I purchased a PS3 CECHA01 from a buy and sell shop for $200 CAD. It had a warranty seal, which was one of the reasons I wanted it and later on in life, get someone to swap the RSX with 40/65nm. Fan wasn't loud but it was noticeable after few minutes. Like level 3. I opened it up, ruined the seal in the process. Found bunch of stickers on Blu-ray drive, indicating that it has been serviced, looked like Sony's work.

As soon as I went to the motherboard, it had tiny residue of flux left around the RSX and someone had written 40nm on the board. I was genuinely surprised, saw a few caps on the back side of the board to be diagonal and not straight.
It had the CXD5301DGB, which is a 40nm chip used in the earlier slim models.
I was quite happy to find that out and surprised that Sony was doing it? I posted it on a Facebook console repair group and someone there mentioned that Sony was doing swaps on some refurbished units.
I sent all of it to Mystic Ryan from YouTube as well, sent a DM on Twitter and he wanted me to share it here on this forum.

Since then, I've replaced thermal paste with Kryonaut and replaced each and every thermal pad. I haven't done any de-lid on it, I've killed one PS3 in the past and I don't wanna kill this super rare system now. I just wish I could find someone in Toronto to get Cell and RSX de-lid just so it has more life and runs a bit quieter, super used to of ps5 being super quiet now.

I haven't modded it or installed any CFW, I tend to keep it stock and play some games online. Also, replaced the 250GB HDD it came with, using a 500GB SSHD now.
kb423gv


kb423gv


kb423gv

https://imgur.com/gallery/kb423gv
kb423gv


kb423gv
Click to expand...
A very lucky find indeed! I think this makes about 5 reported on this thread now? Something like that.

Any way we could bother you to take more pics of the motherboard? Especially I would like to see the places in boxes below...
Frankenstein Changes COK-001 Side A.jpg
Frankenstein Changes COK-001 Side B.jpg


You mentioned some stickers. I'd like to see the service stickers too if you would.
 
RIP-Felix said:
A very lucky find indeed! I think this makes about 5 reported on this thread now? Something like that.

Any way we could bother you to take more pics of the motherboard? Especially I would like to see the places in boxes below...
View attachment 37786 View attachment 37787

You mentioned some stickers. I'd like to see the service stickers too if you would.
Click to expand...


I'll take it apart once I receive my APS-226 PSU which has been on backorder from last month. I expect it to be here by next week, I'll take pictures then.

The service stickers were on top of the Blu-ray drive, I think I posted those on Imgur link.
 
You must log in or register to reply here.

Similar threads

Back
Top