PS3 Frankenstein PHAT PS3: CECHA with 40nm RSX

Pacorretaco said:
Remember the alternative way of solving this error. Botakompongs way with a modchip. Maybe worth trying if run out of ideas? This is still uncharted territory.

Also I vaguely remember him talking about possible GLOD and some resistors? But you probably have this way more controlled than in my memory. Maybe another thing to double check in case you havent already
Click to expand...
Haha...um, i just went back and watched. I don't remember if we performed the resistor mods! Thanks paco!

It was late.

@Computer Booter did we do the resistor mods to CGreset (10k diagnal to GND and remove R2001)? Those are required for both methods. I don't think we did. That is the change that causes the 4002 error that you were expecting, and may explain why we didn't get it. I have a very strong suspicion we forgot to do that! @Pacorretaco is right, it will cause a GLOD if we don't do this.
 
Last edited:
RIP-Felix said:
Haha...um, i just went back and watched. I don't remember if we performed the resistor mods! Thanks paco!

It was late.

@David Rainer did we do the resistor mods to CGreset (10k diagnal to GND and remove R2001)? Those are required for both methods. I don't think we did. That is the change that causes the 4002 error that you were expecting, and may explain why we didn't get it. I have a very strong suspicion we forgot to do that! @Pacorretaco is right, it will cause a GLOD if we don't do this.
Click to expand...

Lolol no, we did not do the resistors!! Now I'm excited haha


Sent from my iPhone using Tapatalk
 
Felix I wasn't online all night, I've sleep beteewn phases,I just wake up on time of reball.Pff that is good hint, I was so confident that rsx must work,we got 2,2 ohm without vcore jumper so was more then enough.So that may work .I say all this take time so we must take slow each process as learning curve,even if isnt working ,it must take precautions until last test.
 
Pacorretaco said:
That's why 3034 alone is not deterministic information.
Click to expand...
lol!!! that was my point! I'm glad we're on the same page then.

Pacorretaco said:
40-3034 appearing alone without the associated 40-44xx error typical of the funny RSX issues may actually be an early board with an early syscon. Logging the same issue in a slightly different way.
Assuming some things etc etc of course.
You need to look at the bringup for more info.

Is it a CECHA or B with a 201GB syscon?

Edit: sorry perhaps this could start to clutter this thread. I think trying to identify RSX problems to know when to replace is relevant, and the 201GB is still not 100% confirmed or understood... But maybe not the highest priority now
Click to expand...
I remember you talking about the funny rsx issue, but to be honest, I didn't understand them back then.


its a cecha01, but i don't remember the syscon version. i do have my bring up logs, i'll look them up. As far as the age, Im not sure. I bought it shrink wrapped all the way back in '06, and then the ylod in '07. I sent i back to sony, and they sent it back, but i dont know if it was the same unit. I sort think so, but i don't remember for sure.
 
Take another look at my post with the frankie patches, i been updatiing it with a new style, is more clear, and there is an important change, now im mentioning the RSX model names "by series" and i removed all mentions to the concept of "RSX 40nm v1" and "RSX 40nm v2" completly because if the IDs are swapped is veeery misleading
You know... if we tell someone the "40nm RSX v1" was used in the first superslim motherboards, and the "40nm RSX v2" was used in the last slim motherboards the first thing the other person is going to tell is "wait, can you confirm that again? because it looks like you did a typo when you wrote it", lol
https://www.psx-place.com/threads/f...cecha-with-40nm-rsx.28069/page-77#post-324662

Anyway... im wondering if i have to update the patches to replace all the mentions of 21EC by 21EB. In the current version of the patches im not doing yet, waiting for a confirmation from #m4j0r
 
Last edited:
vyktormvmpay25 said:
David got it working after resistor mod!
Click to expand...

Well at least we know what happens if you DON'T perform the "resistor mod" for franken phats. LOL. "Special GLOD", more like specially dumb dumb David for forgetting that. Cleaned up the board tonight and delidded Cell, tomorrow going to reassemble and test. I am still flabbergasted that 40nm on BCPS3 is a thing, and that I am able to do it. I can't thank all the contributors enough as I have personally received a LOT of guidance from a ton of people and used methods from others. Special thanks to @vyktormvmpay25 @RIP-Felix as they have been instrumental in my learning process and support. Also thanks to the only other people that I know personally that have worked on this @M4j0r @sandungas @botakompong, anyone else please receive my sincere thanks for your efforts. If there is any way I can possibly give back I would love to know! :-). So far having a blast climbing the reball ladder and learning how to repair these beasts even better every day. <3 <3
 
David Rainer said:
Well at least we know what happens if you DON'T perform the "resistor mod" for franken phats. LOL. "Special GLOD", more like specially dumb dumb David for forgetting that. Cleaned up the board tonight and delidded Cell, tomorrow going to reassemble and test. I am still flabbergasted that 40nm on BCPS3 is a thing, and that I am able to do it. I can't thank all the contributors enough as I have personally received a LOT of guidance from a ton of people and used methods from others. Special thanks to @vyktormvmpay25 @RIP-Felix as they have been instrumental in my learning process and support. Also thanks to the only other people that I know personally that have worked on this @M4j0r @sandungas @botakompong, anyone else please receive my sincere thanks for your efforts. If there is any way I can possibly give back I would love to know! :). So far having a blast climbing the reball ladder and learning how to repair these beasts even better every day. <3 <3
Click to expand...
Well without Pacorretaco you wouldn't be celebrating so soon either! The last piece of the puzzle may not be he most important, but sure is the most satisfying hehehehe.

I am very glad it worked. Just to clarify, was this a COK 001 board with a 201GB SYSCON (the last one to confirm)?

And be sure to test thoroghly yes. All weird tests you can think of

Cheers
 
sandungas said:
I mean... if what you wrote doesnt have any typo... you are telling that the 21EC was used in retail production before 21EB... so they inverted the order of the IDs for some weird reason ? :eek:
Click to expand...
Yes, if you look at the list of FlexIO IDs that's not the first time (the SX SB has a lower ID than the PX SB).
So the RSX CXD5301 identifyer is 21EC ?, then i guess what i said here is wrong:

It seems to be the other way around ?
5300 and 5301 (with IHS) are using identifyer 21EC
And 5302 (without IHS) is 21EB

I did read your post some hours ago and i been thinking in this, but now after reading RIP-Felix post it seems it matches with what you said, so...
5300 and 5301 (with IHS) are using identifyer 21EC ?
And 5302 (without IHS) is 21EB ?

Anyway... im wondering if i have to update the patches to replace all the mentions of 21EC by 21EB. In the current version of the patches im not doing yet, waiting for a confirmation from #m4j0r
Click to expand...
I don't know which chip corresponds to which ID, but the SUR-001 came only with the 21 EC. All later boards support 21 EC or 21 EB, Syscon detects which chip is used and selects the right data.
 
Workz_777 said:
Looking forward to this, was watching your stream. I'm concerned about how all the 8 x NEC/Tokins were replaced on that PS3 you were working on. The original NEC/Tokins act as a filter bridge, filtering out any AC noise on the line. However, when people put that bridge wire across, much of the current won't go through the caps and instead will just go via the bridge wire. Like water taking the path of least resistance. Meaning there is no filtering out of AC noise before the current reaches the RSX / CELL.

View attachment 36384

See image below, incorrect way...

View attachment 36377

This is what it should be doing, see image below...

View attachment 36378

I wrote about this 1.5 years ago in the YLOD / NEC/TOKIN thread, i even made prototypes and posted them back then, this fixed my YLOD PS3 and i was playing on it for hours each day, for months, all perfect.

View attachment 36379

View attachment 36380

Here below with added MLCCs Array, 47uF, 22uF (best for filtering out AC), 10uF, 1uF, 0.1uF. All 8 x NEC/Tokins were replaced like this.

View attachment 36381

Since then i didn't return to that thread until a few days ago, i was going to make a PCB prototype, but then i noticed RIP-Felix had already made a PCB version of the same idea...

View attachment 36382 View attachment 36383

In short, i feel that PS3 you had in your live stream David Rainer still needs it's NEC/Tokins done correctly, else it probably will still be a problem regardless of all the other mods and changes made.
Click to expand...
The 2 methods are electrically identical. The only difference between them is that my PCB has an identical and lower impeedance path for the current to travel. I made it just to make installing easier and to add MLCC's for slightly better performance.
 
Computer Booter said:
Well at least we know what happens if you DON'T perform the "resistor mod" for franken phats. LOL. "Special GLOD", more like specially dumb dumb David for forgetting that. Cleaned up the board tonight and delidded Cell, tomorrow going to reassemble and test. I am still flabbergasted that 40nm on BCPS3 is a thing, and that I am able to do it. I can't thank all the contributors enough as I have personally received a LOT of guidance from a ton of people and used methods from others. Special thanks to @vyktormvmpay25 @RIP-Felix as they have been instrumental in my learning process and support. Also thanks to the only other people that I know personally that have worked on this @M4j0r @sandungas @botakompong, anyone else please receive my sincere thanks for your efforts. If there is any way I can possibly give back I would love to know! :). So far having a blast climbing the reball ladder and learning how to repair these beasts even better every day. <3 <3
Click to expand...
I had another engaugment, but pulled the stream up on my phone. I saw you get it working. Congrats!

I was pretty confident in the RSX and reflow. Ohm test after was good. Heat didn't hurt it and it should have a solid bond.

Should be as reliable as a slim now...fingers crossed!
 
Last edited:
well this was an unexpected development! just read through the last 80 pages, and now the dream of upgrading my cecha12 with a 40nm gpu seems to have none from a big cope, to very likely to happen in the near future.

does anyone believe that they will be able to make the syscon compatible with replacement (or ideally 65/40nm) cpu's? iirc that was the main obstacle to resurrecting many ylod systems.

if my system can eventually be upgraded to become truly bulletproof i will be a very happy bunny.
 
Workz_777 said:
What 2 methods are electrically identical?

(A) View attachment 36386 (B) View attachment 36387

How are the above methods (A) and (B) electrically identical?

I'm not sure i understand what you mean.

Because a bridge wire allows the current to by-pass the filtering caps.

The photos below, with bridge wires, aren't filtering out the AC noise, because current, like water will just take the path of least resistance. In the photos below, the current just crosses over via the bridge wires. Why even have filtering capacitors in that case?

View attachment 36389 View attachment 36391
Click to expand...
A and B are the same. A has an unnecessary wire.

What you are not understanding is how bypass caps work. High frequency noise can travel through the capacitor to ground because of their less known property. Eveyone knows they act like a "battery" to sure up voltage during peak useage, but the second property is the frequence response curve. Noise above or below certain frequencies cant go through the cap. It looks like a resistor to those frequencies. But at certain frequencies the impeedance is smaller and they will go through the capacitor to ground because it provides a lower impeedance path to ground than the processor does. We don't want noise in the processor, where it interferes with signals
 
You must log in or register to reply here.

Similar threads

Back
Top