PS3 Frankenstein PHAT PS3: CECHA with 40nm RSX

DeadEnd said:
@sandungas @RIP-Felix I have completed the Excel sheet for the 90nm CELL. It is entirely based on the service manual. Sadly my patience has run out for the 65nm, so maybe one of you could continue (In the SEM datasheet they did not group the pins alphabetically but by function, so have fun with that :D).

I've tried to mark important notes as well as I could. For the voltage pins I put in parenthesis the name of the contact that it connects to on board (which does include voltage value in its designator). But yeah, feel free to use this to build a better image of Cell's layouts or whatever.

The file is in this folder among all the other documents with regards to Cell's designs https://bit.ly/3Ir1Cha

Or a direct link https://docs.google.com/spreadsheet...ouid=112686022450500831035&rtpof=true&sd=true

EDIT. Well , actually it was already kind of done. https://www.psdevwiki.com/ps3/CXD2964AGB Did I just waste my time again ? :sfun bangdesk:
Click to expand...
Ops, i thought you was doing the "side by side" comparison only, 90nm versus 65nm in only 2 columns and 1500+ rows, thats the kind of thing that doesnt exists in wiki
But for the 90nm pinout alone yeah, we have that page... is actually this link, intended to be displayed in all the wiki pages of all the CELL models using this pinout
https://www.psdevwiki.com/ps3/Template:CELL_pad_layout_41x41

If you click in the "View history" of the page and then in the "prev" links you will realize i been editing it a couple of days ago, in the edit when i commented "Ugly but handy" i added a feature to the table that allows to "sort" the table rows by clicking in the black arrows at most top of the table header... try it in the table by sorting contents by "Name"... then scroll down the page to the most bottom to see how are grouped the "Y0", "Y1", "RX", "TX" pads together alphabetically (for the wide data buses)

Thats the last thing i was doing, while checking his names i found many typos so i guess there are some more innacuracies. Now i guess i need to crosscheck your list too, this is going to be crazy, 3 different sources with 1500 lines each :sfun loco:
 
Well I suppose what I started can be easily continued because I didn't look at devwiki at all. My source was the manual and I copied it row by row. The same can now be done with 65nm but it would take a bit more time. Every row would have to be cross-referenced if anyone were to make an interposer... That was the idea behind it all. But yeah, feel free to use it as you wish.
 
DeadEnd said:
Well I suppose what I started can be easily continued because I didn't look at devwiki at all. My source was the manual and I copied it row by row. The same can now be done with 65nm but it would take a bit more time. Every row would have to be cross-referenced if anyone were to make an interposer... That was the idea behind it all. But yeah, feel free to use it as you wish.
Click to expand...
The plan is to compare your list with wiki, and if i find some mistmatch i will either check in the manual... or if is something controversial i wil take note about the pad name in a private .txt to ask for your oppinion here

I dont want to make this explanation of how wiki works too long, but maybe there is some wiki editor reading me so here we go...
In wiki main page there are links to every single CELL model known, even if the only difference is they changed the suffix "B" by "C", this was a decission made many time ago and the rule applyes to many other wiki pages, the reason is because sometimes we dont know whats different in them, also because usually is nice to add a photo about everyone of them just as a proof that exists all that "variants", and also as a preventive meassure because we are making a dedicated page for everyone of them to make room just incase there is some "hack" or interesting info that applyes to only one of them specifically
The problem derivated from that webdesign decission is sometimes is needed to display the same info in many differnt pages (in this case the pinouts, there are 4 CELL pinouts in total, for 14 unique CELL models)
The solution to avoid duplicated info (as example a massive pinout table of 1500 rows) is to move the content to a "template"
That concept of a "template" is something related with the wiki software, anyway... the point is we can create a template and display it in several pages

Or... (here comes the crazy idea) we can create a new wiki page named "CELL Pinout Evolution" and display 2 templates (the pinouts of 90nm VS 65nm) side by side
I never tryed it and is a bit massive, not sure if it will work but for this to work with a good alignment all the templates needs to have the same number of rows (the hole at the center of CELL is a problem because the number of pads missing in it varyes)
Dunno, just ideas... im not sure if im going to be able to "merge" 2 wiki templates like that... there are many tools intended to help in this kind of wiki table editions, is the same "mediawiki" software used by wikipedia, so for sure there is something to deal with massive tables, is just im not used to that softwares, i will try in the DIY way :P

Anyway... by now the other CELL pinouts doesnt have a template, and the template of the link needs to be renamed, i need to start by that
 
Last edited:
So for the statistics sakes, I present to you a working COK-001 refurbished with 65nm by Sony. Sorry for crappy quality.

Interesting notes would be that they used a plastic piece to cover the hole under CPU. Also I desoldered one of the replaced resistors and measured it to be 47k after all. The voltage regulator for RSX was definitely replaced too (sorry, forgot to take a picture), but there are no signs of tampering with lower thresholds.

Error log contains some of the recent forced shutdowns I had to make because certain programs didn't work on evilnat's 4.89 and froze up.

Firmware Version: 4.89 (build 50740)
Platform ID: Cok14
Product Code: 00 83
Product Sub Code: 00 01
Hardware Config: 00000000FFFFFFFF
Syscon Fimware Version: 0F29.0001000500000002 (EEPROM: 0001000500000002)
Bringup Count: 557, Shutdown Count: 534
Runtime: 66 Days, 14 Hours, 15 Minutes, 25 Seconds
Error Log
01: A0801001 Thu Jul 14 21:28:17 2022
02: A0801701 Thu Jul 14 20:54:56 2022
03: A0801001 Mon May 9 00:26:57 2022
04: A0902203 Sat May 23 09:00:38 2020
05: A0801001 Sat May 23 09:00:38 2020
06: A0801001 Fri May 22 11:20:17 2020
07: A0801001 Wed May 20 14:50:35 2020
08: A0801001 Wed May 20 14:49:08 2020
09: A0901001 Mon May 18 15:36:48 2020
10: A0801701 Mon May 18 15:36:48 2020
11: A0801001 Sun May 17 04:58:30 2020
12: A0801001 Sat May 16 18:15:14 2020
13: A0801001 Sat May 16 11:56:36 2020
14: A0801001 Mon Apr 6 16:41:19 2020
15: A0801001 Sun Apr 5 15:23:08 2020
16: A0801001 Tue Aug 7 16:16:33 2018
17: A0801701 Fri Jul 13 19:52:52 2018
18: A0801701 Thu Jun 7 15:34:55 2018
19: A0801701 Thu Jun 7 15:34:49 2018
20: A0902203 Thu Jun 7 15:34:38 2018
21: A0801701 Thu Jun 7 15:34:38 2018
22: A0801001 Thu May 24 11:21:23 2018
23: A0801001 Wed May 23 05:15:38 2018
24: A0801001 Fri May 11 13:20:22 2018
25: A0801001 Mon Mar 4 03:45:41 2013
26: A0801004 Mon Mar 4 01:54:49 2013
27: A0801004 Mon Mar 4 01:52:28 2013
28: A0801001 Mon Mar 4 01:46:34 2013
29: A0801001 Mon Mar 4 01:39:09 2013
30: A0801001 Mon Mar 4 01:37:58 2013
31: A0801004 Fri Jul 8 11:28:04 2011
32: FFFFFFFF Fri Dec 31 23:59:59 1999


20220715_002836.jpg 20220715_003029.jpg 20220715_003041.jpg 20220715_003240.jpg 20220715_003337.jpg unknown (1).png unknown2.jpg

I repasted it and ran GoW3 quickly, temps are ok but nothing like 40nm.
65nm CECHA.JPG
 
Last edited:
well the fact that there is a plastic to cover the cell hole is something that sony did to later revisions also. its in there to apply more pressure tp ihs heatsing. thats why some people do the rubber trick or eraser mod.
 
truemaster said:
well the fact that there is a plastic to cover the cell hole is something that sony did to later revisions also. its in there to apply more pressure tp ihs heatsing. thats why some people do the rubber trick or eraser mod.
Click to expand...
Do Not Ever do that!!! I did that and my Cell Die was Cracked in Half!!!!!!
 
RIP-Felix said:
I actually think the 65nm CELL will be much easier to fill in. I like the way SONY organized them by voltage better. Makes thing much easier to read.
Click to expand...
Maybe just give Up the idéia of Connecting Smaller Transistors Revision Cell B.Es, The Best and Safest motode to solve PS3 Cpu heating and loud fan is just to aply Líquid Metal and we get blessed with 58 degrees celsius on Cell 90nm transistors size in cok-001 with original heatsink and only 28% fan speed and mostly important all that Running The Last of Us which is a game Who Fully utilize all 6 Spes and the 1 Ppe to process Graphics and also A.I and game world etc so very Power heating game on Cell aswell as its an 2013 game, 7 years after ps3 launch and Naughty Dog said they used all drops of Power in PS3
 
DeadEnd said:
So for the statistics sakes, I present to you a working COK-001 refurbished with 65nm by Sony. Sorry for crappy quality.

Interesting notes would be that they used a plastic piece to cover the hole under CPU. Also I desoldered one of the replaced resistors and measured it to be 47k after all. The voltage regulator for RSX was definitely replaced too (sorry, forgot to take a picture), but there are no signs of tampering with lower thresholds.

Error log contains some of the recent forced shutdowns I had to make because certain programs didn't work on evilnat's 4.89 and froze up.

Firmware Version: 4.89 (build 50740)
Platform ID: Cok14
Product Code: 00 83
Product Sub Code: 00 01
Hardware Config: 00000000FFFFFFFF
Syscon Fimware Version: 0F29.0001000500000002 (EEPROM: 0001000500000002)
Bringup Count: 557, Shutdown Count: 534
Runtime: 66 Days, 14 Hours, 15 Minutes, 25 Seconds
Error Log
01: A0801001 Thu Jul 14 21:28:17 2022
02: A0801701 Thu Jul 14 20:54:56 2022
03: A0801001 Mon May 9 00:26:57 2022
04: A0902203 Sat May 23 09:00:38 2020
05: A0801001 Sat May 23 09:00:38 2020
06: A0801001 Fri May 22 11:20:17 2020
07: A0801001 Wed May 20 14:50:35 2020
08: A0801001 Wed May 20 14:49:08 2020
09: A0901001 Mon May 18 15:36:48 2020
10: A0801701 Mon May 18 15:36:48 2020
11: A0801001 Sun May 17 04:58:30 2020
12: A0801001 Sat May 16 18:15:14 2020
13: A0801001 Sat May 16 11:56:36 2020
14: A0801001 Mon Apr 6 16:41:19 2020
15: A0801001 Sun Apr 5 15:23:08 2020
16: A0801001 Tue Aug 7 16:16:33 2018
17: A0801701 Fri Jul 13 19:52:52 2018
18: A0801701 Thu Jun 7 15:34:55 2018
19: A0801701 Thu Jun 7 15:34:49 2018
20: A0902203 Thu Jun 7 15:34:38 2018
21: A0801701 Thu Jun 7 15:34:38 2018
22: A0801001 Thu May 24 11:21:23 2018
23: A0801001 Wed May 23 05:15:38 2018
24: A0801001 Fri May 11 13:20:22 2018
25: A0801001 Mon Mar 4 03:45:41 2013
26: A0801004 Mon Mar 4 01:54:49 2013
27: A0801004 Mon Mar 4 01:52:28 2013
28: A0801001 Mon Mar 4 01:46:34 2013
29: A0801001 Mon Mar 4 01:39:09 2013
30: A0801001 Mon Mar 4 01:37:58 2013
31: A0801004 Fri Jul 8 11:28:04 2011
32: FFFFFFFF Fri Dec 31 23:59:59 1999

And for those interested in making the plastic cover, I took some quick measurements and it's approximately like this.

View attachment 37966
View attachment 37964
View attachment 37965

View attachment 37956 View attachment 37957 View attachment 37958 View attachment 37959 View attachment 37960 View attachment 37961 View attachment 37962

I repasted it and ran GoW3 quickly, temps are ok but nothing like 40nm.
View attachment 37963
Click to expand...
Could you maybe...pretty please...test the VDDR voltage? I know you probably don't want to open it again, seeing how you already reassembled. But I strongly suspect the 65nm cell (edit, I meant RSX. I had been looking at the CPU schematics too much.) needs a voltage reduction also.

Currently in my tutorial I say you only need to reduce it for the 40nm. But AFAIK no one has confirmed this assumption. I want to be sure my tutorial includes proper advice for 65nm as well. So I would greatly like to see what sony did to VDDR. They did replace IC6200 with BD3504, which they would only do if they did reduce the voltage! So I want to know what voltage it's reduced to.
 
Last edited:
RIP-Felix said:
Could you maybe...pretty please...test the VDDR voltage? I know you probably don't want to open it again, seeing how you already reassembled. But I strongly suspect the 65nm cell needs a voltage reduction also.

Currently in my tutorial I say you only need to reduce it for the 40nm. But AFAIK no one has confirmed this assumption. I want to be sure my tutorial includes proper advice for 65nm as well. So I would greatly like to see what sony did to VDDR. They did replace IC6200 with BD3504, which they would only do if they did reduce the voltage! So I want to know what voltage it's reduced to.
Click to expand...

@botakompong has mentioned before that it would be reduced down to 1v. But I don't want to open it again at the moment...
 
RIP-Felix said:
Could you maybe...pretty please...test the VDDR voltage? I know you probably don't want to open it again, seeing how you already reassembled. But I strongly suspect the 65nm cell needs a voltage reduction also.

Currently in my tutorial I say you only need to reduce it for the 40nm. But AFAIK no one has confirmed this assumption. I want to be sure my tutorial includes proper advice for 65nm as well. So I would greatly like to see what sony did to VDDR. They did replace IC6200 with BD3504, which they would only do if they did reduce the voltage! So I want to know what voltage it's reduced to.
Click to expand...

Ok, opening ps3s has become my morning routine... So to give you a piece of mind. It's 0.990v or 1.0v in a perfect world.

 
RIP-Felix said:
Could you maybe...pretty please...test the VDDR voltage? I know you probably don't want to open it again, seeing how you already reassembled. But I strongly suspect the 65nm cell needs a voltage reduction also.

Currently in my tutorial I say you only need to reduce it for the 40nm. But AFAIK no one has confirmed this assumption. I want to be sure my tutorial includes proper advice for 65nm as well. So I would greatly like to see what sony did to VDDR. They did replace IC6200 with BD3504, which they would only do if they did reduce the voltage! So I want to know what voltage it's reduced to.
Click to expand...

Felix he never wrotte which chip was refurb with an 65nm Revision, so he just made us believe without Intention that he was talking about the Cell because Everybody was talking about the Cell, but Actually by his photo he has an Cxd2964gb Broadband Engine, so its Indeed an 90nm Original Cell.
 
Victor Hugo Alvarez said:
Felix he never wrotte which chip was refurb with an 65nm Revision, so he just made us believe without Intention that he was talking about the Cell because Everybody was talking about the Cell, but Actually by his photo he has an Cxd2964gb Broadband Engine, so its Indeed an 90nm Original Cell.
Click to expand...

Felix meant RSX. We are all talking about RSX, he just made a typo... Cell 65nm swap has never been done by Sony or anyone else... yet.
 
Last edited:
Guys Warning to the PS3 Cell; dont Ever place any Pressure with any kind of medium Strong material in the back of Cell; as an result of this bullshit I had my Cech-2001 with Cell Silicon Cracked in Half and Ylod.

Never put any kind of Pressure in the back of Cell!!!
 
Haha... Yeah, didn't catch that. I had been looking at cell schematics a lot to see if it would be possable to do a cell swap. But yeas, I meant the 65nm RSX. I wanted to update the frankie tutorial with accurate information. I have since done so. So from now on I do reccomend VDDR voltage mod for all frankies.
 
Victor Hugo Alvarez said:
Guys Warning to the PS3 Cell; dont Ever place any Pressure with any kind of medium Strong material in the back of Cell; as an result of this bullshit I had my Cech-2001 with Cell Silicon Cracked in Half and Ylod.

Never put any kind of Pressure in the back of Cell!!!
Click to expand...

25a5875d17a6973997a50bc7c94d6d35.jpg

Here's what I found my A01 years later after sending it to Sony. A lightning strike fried it a week after I brought it home December 2006. I sent it in under warranty, got it back and enjoyed it for 6 years. Opened it years later to find this behind the CPU.
84f7b6cd367aa30d22bdf0659cb1928d.jpg

95f5e37048c4a5303d7f3799c73f39b7.jpg

345e045d557aec4a892d9eb3f1645a50.jpg



Sent from my iPhone using Tapatalk
 
cfreddykrueger said:
25a5875d17a6973997a50bc7c94d6d35.jpg

Here's what I found my A01 years later after sending it to Sony. A lightning strike fried it a week after I brought it home December 2006. I sent it in under warranty, got it back and enjoyed it for 6 years. Opened it years later to find this behind the CPU.
84f7b6cd367aa30d22bdf0659cb1928d.jpg

95f5e37048c4a5303d7f3799c73f39b7.jpg

345e045d557aec4a892d9eb3f1645a50.jpg



Sent from my iPhone using Tapatalk
Click to expand...

Interesting... I had seen the plug before, but not those plates. Were they meant to be installed between the clamps and the board? It almost looks like the "official implementation" of the eraser mod.
 
You must log in or register to reply here.

Similar threads

Back
Top