PS3 Frankenstein PHAT PS3: CECHA with 40nm RSX

If is same Xc2c32a it will work.White mask can not hide pinout from ic of green board so it is working if ic use same pinout, unless they change I/0 in out or bank adresess of that ic.
 
RIP-Felix said:
I'm not so sure. That looks like a different board than the one in the video I clipped the pics from. The pads could be different. I'm at work now, so I don't have the videos in front of me. Did one of those show these white boards? I'd double check before assuming I have it correct.
Click to expand...

He used the same white board in the video where you watched him install it. The pics you clipped at the end of the video appear to be a printed instruction paper, which came from an older revision of the modchip with a different color of the board and a larger package Xilinx IC. Since he was following steps from the paper, I'm willing to bet the modchip and its firmware are identical regardless of its revision/color or package size. The IC can be installed on the models originally containing either 90 or 65 nm RSX (All fats and early slims). The only differences are contact locations on the mainboards and resistor kinks. This would be my current guess.

To rephrase:

Based on what we know so far, the magic IC appears to be able to marry RSX series cxd2991-cxd5300 (models CHECHLxx through CECH-25xx) with any board from the models range CECHAxx - CECH-20xxB. Compatibility of the later RSX versions cxd5301-cxd5302 (CECH-30xx to CECH-43xx) probably needs to be confirmed (depends what's changed since cxd5300).
 
Last edited:
RIP-Felix said:
Are you supremely confident that the reball was successful? Where did that 4001 come from? You didn't remove any of the 47K's did you?
Click to expand...

I grabbed a 47k off of another board, I didn't even bother to measure the one I took off to throw away. So, it's still the same then I guess, lol.

I'm fairly confident. The board had only A0403034, and the 90nm RSX measured like 0.9 ohms. The 65nm RSX was known working with no prior issues ever. However, the board has been through 4 rework cycles with a lot of baking too, and well, anything goes at that point. Another reason I'm excited about this chip. Cutting out 2 rework cycles is a HUGE plus, both in time and reliability.
 
squeept said:
I grabbed a 47k off of another board, I didn't even bother to measure the one I took off to throw away. So, it's still the same then I guess, lol.

I'm fairly confident. The board had only A0403034, and the 90nm RSX measured like 0.9 ohms. The 65nm RSX was known working with no prior issues ever. However, the board has been through 4 rework cycles with a lot of baking too, and well, anything goes at that point. Another reason I'm excited about this chip. Cutting out 2 rework cycles is a HUGE plus, both in time and reliability.
Click to expand...
Would be helpful next desoldered rsx if would like to read rsx markers, may not be relevant but on 90nm there is a simple note to one corner "Rsx" and may be other codes marked. On other 40 nm rsx I have seen "RSX4". I have on work another 65 nm and will look as well for marker on this moodel. It may help to identify it somewhere in middle or inside syscon.
 
squeept said:
I grabbed a 47k off of another board, I didn't even bother to measure the one I took off to throw away. So, it's still the same then I guess, lol.

I'm fairly confident. The board had only A0403034, and the 90nm RSX measured like 0.9 ohms. The 65nm RSX was known working with no prior issues ever. However, the board has been through 4 rework cycles with a lot of baking too, and well, anything goes at that point. Another reason I'm excited about this chip. Cutting out 2 rework cycles is a HUGE plus, both in time and reliability.
Click to expand...
The reason I ask is because the test before and after that 47K are electrically identical, so that 4001 came out of nowhere! Unless the minimal heat or mounting pressures were different, you should have had identical results. In other words, you may have popped a BGA like I did on the bad tokin board you sent me. I was just testing/soldering, I dumped some heat into the board, but not so much I would be worried about the BGA. If the BGA is that sensitive, I'd be concerned you're having problems with your equipment or these boards had badly oxidized pads.

Maybe try the pressure test?
 
I was looking for some rsx IDs, may not be relevant but who knows All are on same RSX4. Rsx 65 does not have, well at least this dyn001 board from table.
bef583ce6e933d6a7b733f6f628f74ba.jpg
b0ea57a07b506468a13c801c54129aad.jpg
c9d0d7556fe87ac24bcf19fe70cac7e1.jpg
 
I got some new photos from the seller, they're absolutely atrocious pictures but one of them is at least worth sharing. Compare the first image with the other COK-001/2 installation diagram and you'll notice that this one is actually installed on the complete opposite side of the motherboard.
IMG-20210312-WA0017.jpg
higher res resistor guide
IMG-20210312-WA0016.jpg
 
Okay I got my multimeter out and traced where the soldering points were on the board based on what points matched up with the points on either side. They do indeed line up to exactly where @RIP-Felix said the points were in the first place.
The only thing I cant figure out is the resistor/capacitor setup on the back. Maybe it's because the board used in the video was a COK-002?
There's 2 blue SMD on my COK-001 that don't seem to be there on the board used in the YouTube tutorial. I've got a water damaged CECHE I'll open up later to check and take photos of but I figured I would post the COK-001 photos now for anyone looking to make a guide on how to do this.
COK-001 top side solder points
20210312_182835 - Copy.jpg
COK-001 bottom side solder points
20210312_183416 - Copy.jpg
COK-001 RSX top side SMD layout
20210312_202327 - Copy.jpg
20210312_202455 - Copy.jpg
edit: added new photos
 
Last edited:
DoublesAdvocate said:
...The only thing I cant figure out is the resistor/capacitor setup on the back. Maybe it's because the board used in the video was a COK-002?

There's 2 blue SMD on my COK-001 that don't seem to be there on the board used in the YouTube tutorial.
Click to expand...
If you're using a COK-001, then you don't have to mess with the 47K resistors. They are already populated from the factory. That step is only for COK-002, because they were removed.
 
DoublesAdvocate said:
...I've got a water damaged CECHE I'll open up later to check and take photos of...
Click to expand...
For the COK-002 use as reference the last photos posted by @DeadEnd and the BGA layout i mentioned in this post
The points you named 1,2,3 (in the top/components side of the motherboard) have a trace directly connected to a pad of the syscon border (well, 2 of them, the other is located in second row)
At this point that comparison should allow to identify which one is each (CLK, DO, DI)
 
I found all my answers, just going to report more when mod chip is here. We may not be able to get a copy of the Xc2c32a, I have tools to program Xc2c32, 64, 128, etc it has digital fuses and can be programmed, can not retrieve, only right file svf/xsvf can be written.
 
Last edited:
Yeah, I don't know that that guy is trustworthy... I remember asking around here, a while back, about his reball services. All I heard is that he's not that great to work with and to avoid it.
 
DoublesAdvocate said:
Yeah, that guy is a scammer who got kicked off of ebay. I wouldn't trust anything he says or does.
Click to expand...
Fair enough. I guess he is not going to help us figure it out so. :)

Hopefully this mod is open sourced for everyone soon, and without the chip someday too.

Funny the way he makes a big deal out of lead free solder being better and safer, afaik old solder is better really and only less safe to those applying it or when you dont dispose of it correctly.
 
Last edited:
@squeept I've got several dead PS3 motherboards and I'm wondering if you'd be willing to harvest the chips from them and then send the bare boards and individual components back to me so that I could document them and hopefully work on designing a PCB schematic that I could print myself. I've got a hot air station so the smaller chips and smd components are fairly easy but I know if I try to remove something like the CELL or RSX I'm just going to tear the traces to shreds. As far as I can tell every model of PS3 used a 6 layer PCB.
 
Half of the time when people message me on eBay asking about buying a CECHA01, they start the message with "I was recently scammed by ps3specialist.com"

@DoublesAdvocate There are already schematics of several boards, what are you looking to figure out here? If you've already got hot air and you're gonna dive in to this kind of stuff, grab that Puhui preheater that everyone uses and you're set.
 
You must log in or register to reply here.

Similar threads

Back
Top