The more I think about it, the more I'm convinced that @RIP-Felix is right about there being a CPU issue on the first one I tried with the chip. But then I got frustrated and just started swapping wires around since we didn't have perfect documentation then and I thought maybe some of those early pictures were wrong, so I think I fried both my mod chips by accidentally applying voltage to a data line. Then, the next two I tried worked perfect (each was 3034 then swapped and chipped still 3034) but my modchips are toast.
Either way, yeah, just gonna wait for some new modchips.
But then I got frustrated and just started swapping wires around since we didn't have perfect documentation then and I thought maybe some of those early pictures were wrong, so I think I fried both my mod chips by accidentally applying voltage to a data line.
I ordered a 40nm RSX in addition to the 90nm I already have on the way, I'm considering the Frankenstein mod. Might as well have both on hand to decide which path to take.
...
For the Indonesian "IC mod" chip, apparently the only way to buy this is from a third party though Fiverr, with questionable results? Has anyone already bought a bulk order and sitting on inventory? I suppose I can take this question to the Frankenstein thread.
Are there alternatives to the Indonesian mod on the horizon? I'd rather wait if something better is coming down the pike.
Honestly, if you have to take the RSX off anyway, you may as well put a 40nm on there. If you have the tools and experience to reball, then the necessary HW changes are trivial. Plus New old stock CXD5301's are still available for cheap on ebay. The ones I bought even came pre-balled (lead free, but that's fine for a 40nm). I have a sealed A01 ready to be diagnosed and sacrificed to the cause. I'm just procrastinating!
The biggest inconvenience is having to buy through forwarding services like Fiver. However, the process is not difficult. A couple of posts back @DeadEnd linked the person he's ordered from. I've asked @botakompong (the maker) to look into selling them directly on e-bay. I think he's just been busy with his local business and there is a language barrier. Last I heard he just hasn't gone through the hassle of figuring out how to make it work, or decided if it's worth it. Perhaps he will in the future, if there's enough demand. Correct me if I'm wrong @botakompong.
The mod chip is pretty much the ultimate solution. It's easier than the official SONY method, which we still haven't replicated yet. Currently no one has purchased a bulk order for MOD chips directly from @botakompong for regional distribution. I think you could easily ask $35 for it in North America, if it were an easy purchase. Maybe someone will add it to their mod shop or something. It just hasn't happened yet. It's still pretty new to western markets. The word hasn't really gotten out yet. Honestly, I think it's just us really.
I have been working on a PDN (Paint.net, basically a freeware version of photoshop) of Kiaw's pinout. I have put a lot of work into tracing out all the power and etc. @sandungas I have some photos you may want to upload to the wiki. This one I find useful for working on the board as you two seem to be doing.
I disagree with the schematic, though. It say's that all but one pad for VDDA are not connected (NC). Bottom left of the picture above, in green. Kiaw went with what the schematic say's, but if you probe the board they are connected. They're all on the same power plane and the solder balls are connected to those pads. At least on the board side they are connected. So the schematic lied! Perhaps only AP8 is connected on the RSX and the rest are NC on the RSX, @vyktormvmpay25 could you confirm?
I haven't forgotten about your tests required for cok002 board with modchip and, rsx AP8 points. I just been busy, I did not had time even to start repair Wi-Fi module for an super slim that I have boot loop of wifi short, I will do all with time.
Most of the time in summer I get working till late out of workshop.
So I've got 2 60gb and 1 20GB YLOD that I'd like if possible someone to do this mod to. I've got 3 mod chips and 3 40nm RSX chips. Has anyone started to offer this service yet?
So I've got 2 60gb and 1 20GB YLOD that I'd like if possible someone to do this mod to. I've got 3 mod chips and 3 40nm RSX chips. Has anyone started to offer this service yet?
You could order modchip, after receiving quite any local services for reball ps3/ps4 /Xbox. If they really do reball they should get this working as 40nm rsx is not sensible for heating as 90nm.
What I suggest is to tell to solder it without ihs (delid out of board is easy then on board).
Adding new thermal paste will be easy like that anytime.
Seeing so many people with dying ps3s and being victims of bad service, I have been trying to offer a bit of help. But not many care to ship consoles to EU and back.
I made this thread (sadly using a randomly generated name for the account, but ignore that)
Edit: Looks like @Sampsonay was doing the repair. After chatting with him, Sampsonay has done it all correctly. Seems like mail workers mishandled the package, or GPU was already in its last breath.
@RIP-Felix in post 934 you was right. Rsx don't use all VDDA pins aren't connected to it. Only on motherboard. This was tested with 40nm rsx. In rsx pinout that's why they are not all marked with green. Only AP 8, AD8, AB8, T8 are on rsx used. In have to test per case with 65 and 90nm. Not sure why they did not use all those points.
@RIP-Felix in post 934 you was right. Rsx don't use all VDDA pins aren't connected to it. Only on motherboard. This was tested with 40nm rsx. In rsx pinout that's why they are not all marked with green. Only AP 8, AD8, AB8, T8 are on rsx used. In have to test per case with 65 and 90nm. Not sure why they did not use all those points.
Got today a costumer willing to pay for this modchip if I can't recover his unit cok002 untouched board. On a quick look cpu 2.7 ohms /rsx 2.4 ohms. Not sure if worth trying but will do it anyway. Even if it's not working I still have board prepared for modchip.
PS3#8 - Continued
(Frankenstein Phat therml test)
I am continuing the thermal tests for PS3#8. I started this project in the YLOD/tokin thread, but now that it's become a frankenstein, it fall into this thread now. Here's a recap from earlier today.
Okay guys this will be the last post about this Frankenstein Phat console. At least on this thread. All future posts and thermal tests will be moved to the "Frankenstein PHAT PS3: CECHA with 40nm RSX" thread. Since this thread is about repair/research, particularly the NEC/TOKINs, but not strictly so, I figured it would be worthwhile to document my troubleshooting process on this thread to begin with. The SYSCON part was what I particularly wanted to point out to readers visiting this thread. I hoped it would help them see how important the SYSCON is as a first step of diagnosis! That's how I learned this was not a NEC/TOKIN issue, but an RSX issue (probably a BGA defect) that indicate it might be a good candidate for the Frankenstein Phat MOD. You have to diagnose the YLOD before you can fix it. That's what the SYSCON is for. Litterally, that's why SONY programmed in these error codes, so their technicians could diagnose and repair! We finally have access! And it only costs a few dollars for the USB adapter. Awesome!
Back to this post! This morning I fully assembled the console. Like I said before, I did not delid the RSX or CPU. I'm also reusing old dried out thermal pads. The only new thing about this console if the 40nm RSX and MX-4 thermal paste between the IHS and HS blocks. I routed the SYSCON wires through the vent holes under the power touch pad. That's a very convenient spot for them. I did tie DIAG to GND using aligator clips, but with the case closed I had to clip it to the RF shielding through the HDD port. That means I need to keep the console on it's side for this first termal test. That's important, because the PS3 can run a few degrees cooler when stood vertically. So basically, this should be the worst case scenario for temps! Old pads, no delid, console layed flat.
First stress test:
For this test I will use the SYSCON terminal to bringup the console. After syncing my controller and setting up HDMI to my TV I let the console idle on the XMB for 5-10 minutes. I figured this console has not run since the YLOD, so I have no idea if the CPU will overheat or not. I don't know the state of the TIC between the IHS/DIE. However, I decided to leave it to get a baseline. The worst case scenario for this console, given it's curent state of delapidation, is that the CPU will overheat in the XMB. It didn't, so I was free to proceed to stress testing.
In my experience, every single CPU has dried old paste making poor contact. They all need delidded IMO. I generally just go ahead and do it as preventative maintenance while cleaning. I made a special tool for it that safely cuts the silicone off every time without damaging the traces. It's super easy to do! It wasn't so easy to make, but not that bad either...
For the game, I chose NBA Live 2010. Why? Well, because you can set the computer to play itself under playoff conditions. That has a lot of people in the stadium crowd, which I've notice seem to tax the PS3. This game also had highly detailed player faces and models that pushed the hardware. Also, there's sweat physics, flashing lights from photographers, ribbons and smoke from pyrotechnoics, and lots of camera movements/panning. All of this while I have my hands free to type in syscon commands to monitor temps and fan speeds.
Remember, I don not have custom firmware installed yet. So I don't have WebMAN mod to monitor temps in game or change the default fan curve. There are 3 reasons why:
If there is a NEC/TOKIN or BGA issue waiting to YLOD my console, I don't want it to happen during the firmware flashing stage of a jailbreak and brick the console. I need to stress test to be sure it's stable first.
I don't plan on installing custom firmware on this console! I want this to be my new daily driver (replacing PS3#7, the reballed tokin fixed console). I want to be able to access PSN without worrying about getting banned.
I want baseline temperatures on a stock console to compare with later. So OG thermal pads, console on it's side, with the stock SYSCON fan curves, and a CPU that hasn't been delidded yet. The only thing new/different here is the 40nm RSX and TIC between the HS/IHS'.
Firstly, the Blue LED was illuminated, indicating there was already a disc in the BluRay drive. Yay a freebie! What is it...what is it! Oh...a DVD. Anybody ever seen "Dream Girls" (2006)? It's PG-13 so mayby I'll watch it later. This is the second freebie I've found stuch in a BluRay drive. The previous was "Transformers: Dark of the Moon" on BluRay. I can't say I'm much a fan of the series. It's alway's movies! I never get any games. Probably because when it eats their game they retrieve it. And since I only buy sealed consoles, the chances they left a desirable game in there instead of a boring movie is pretty low.
To the point then. After 15-20 minutes in game the the temperatures settled down and the Fan had them under control. Here are the Temps:
CPU = 72C
RSX = 53C
Fan = 30%
Ambient (room Temp) = 25C
Exhaust (out hottest part of back vents) = 46C
That last temp is important! I felt the air exhausting out the back of cther consoles I've delidded and have great thermal contact and they were hotter. Now you'd think that's bad, but it's actually good. The hotter the air coming out the back the better the thermal contact. It means the heat is efficiently being moved out of the console. This felt warm, but significantly cooler than I've felt before.
Now I have to manage that a bit. The RSX isn't contributing as much heat as it normally does. Usually in a test like this the RSX runs maybe 3-5C cooler than the CPU. So around 68C is generally is what I've seen for a well controlled BC console. 53C on stock paste and fan control is phenomenal! I was hoping I could keep it under 50C when all is said and done with this console. That's my arbitrary heat limit for BGA technology. Chips that run hotter than that flex the bonds too much for the tech to be reliable enough IMO. So If it's below that, I would consider it actually reliable.
So yeah, this RSX is only 3C away from being reliable now (with Lead-Free solder). And that's with my overly strict <50C for BGA rule. Pretty awesome!
I have plenty of ideas to improve these temps, so if you're interested in following this console further, you'll have to move over the the Frankenstein thread for more.
After this I disassembled it again. This time I delided both the CPU and RSX. As suspected the CPU needed it and the RSX did not. Regardless I wanted to delid, so I bit the bullet and just did it. It was scary! CPU was easy, but the RSX's thermal adhesive is strong on 40nm chips. It required an unsettling amount of force followed by a sickening crack as it finally gave way. Luckily, It went well and the BGA is holding strong. So far anyway.
I replaced all the TIC with MX-5 this time around. I also applied new thermal pads for everything except the BluRay drive, which doesn't need it. I used ThermalRight Extreem Odyssey Thermal Pads. They are supposed to have 12.8w/mk thermal conductivity, way better than the stock pads. They weren't cheap, but this console is worth it! Then I reassembled it for the last time.
Here are the temps with the same test procedure as before.
CPU = 68C
RSX = 54C
Fan = 28%
Ambient (room Temp) = 25C
Exhaust (out hottest part of back vents) = 48C
As expected, the exhaust temp went up. This means there was an improvement in thermal transfer. Perhaps counterintuitively the RSX temps are higher than before. So what gives? Well, I expected that. Because the CPU needed a delid, it was heating up more than it did this time. So it took a higher fan speed of 30% to keep it under control. Because the fan was higher and the RSX had good thermal paste, it was being cooled more. So this time a fan speed of only 28% was needed to keep the CPU cool, so the RSX would run a few degrees warmer. That's normal.
So what's next? Well, I'm happy with it as is, but I do want to install a custom fan curve. I'll need @Pacorretaco's help with that. He's been messing around with that more than I have. So how about it buddy, feel like walking me through that process? I don't want to mess with webMan mod, I want to edit the fan curves through the syscon.
Hello all. I set up a buy with Lusianaliu on Fiverr for qty (10) of the mod chip, but she said it is unfortunately out of stock. I used the link suggested in this thread: https://shopee.co.id/IC-MOD-RSX-i.4565788.3505669213
Does anyone know of another validated source of genuine mod chips I can refer her to?
If I may answer on his behalf, last time we talked, he said they were on lockdown because of government regulations. His shop is closed until july 20th. I think he also mentioned that his family is sick with covid.
But we nearly agreed that he will ship around 10 pieces to me after 20th and then I can sell it further. I'm based in EU, so slightly easier for me to ship them forward (don't worry, I won't add any unnecessary costs). Last I checked shipping unregistered envelopes to US would be around 5 euros, ~7 $, price goes up if you want tracking. So with tracking it would be like 20$-40$. Not sure if that's any cheaper than through Fiverr. Probably makes sense for buying in bulk. You can send me pm in case you are curious. Maybe I could get more than 10 pieces. I live near post offices so shipping is usually not a problem.
PS3#8 - Continued
(Frankenstein Phat therml test)
I am continuing the thermal tests for PS3#8. I started this project in the YLOD/tokin thread, but now that it's become a frankenstein, it fall into this thread now...
Do you think you would be able to offer this type of swapping service in the US ? Seems like the demand is there. Makes no sense for them to ship it across the ocean just to get this done.
