PlayStation 3 CECH-A01 / COK-001 Nightmare (Help!)

[grantbennett@me.com]

I recently purchased a lightly used PS3 CECH-A01 which was manufactured January 2007. For some reason, not only is the glue that is holding the Cell/B.E. in place UNGOLDY strong, the clearance in which you would have on any other unit is MUCH slimmer than anything I've previously seen to break the seal of the glue. I have tried everything. Pallete Knifes, (which I DO NOT want to use going forward. I'll get to that) an 8 guage E guitar string, fishing line, unwaxed dental floss, everything. I feel like I have exhausted every method and I'm losing my freakin sh*t. If anyone could shed some light on whether or not very early manufactured models OR lightly used consoles are more difficult, I would be forever in your debt.

Moving on, while using said pallete knifes and working as slowly as possible, a slip happened. Yes. I know. I want to scream from the top of the highest mountain forever. This said slip caused me to break off 6 capacitors at the bottom of one of the 64MB XDR DRAM's and also slighly bend the DRAM chip itself. The components from the schematic are as follows:

ic1202 - 64 MB XDR DRAM (Elpida 061 10w 144 x51 16ac-3c-e)

c1243 - Capacitor
c1258 - Capacitor
c1259 - Capacitor
c1251 - Capacitor
c1254 - Capacitor
c1257 - Capacitor

This is my worst nightmare. I finally get my hands on a beautiful unit and I see it crumble before my eyes. I have included pictures of the troubled areas and would appreciate any help in terms of what the replacement capacitors would be as well as whether or not there is any fix for the bent DRAM chip. I have been on a great streak with my repairs lately and this is the one where the universe finally said "NOPE" Plz help!

[grantbennett@me.com]

For the love of god I have no idea why the bottom picture keeps flipping. This is not my night. Send help.

[Traceless]

appreciate any help in terms of what the replacement capacitors would be as well as whether or not there is any fix for the bent DRAM chip.

The pads of the six missing components seem to be okay as far as I can tell from the picture. If you don't have schematics for the board the easiest way to determine their values would be to measure them. In case they broke you can remove the corresponding components next to the neighboring chip and measure those. For this you will need a component tester or an LCR meter. Judging by the color not all of those components were capacitors.

Forget about the broken DRAM chip - this is probably easily replacable. The actual problem are BGA pads on the mainboard, which are relatively delicate to deal with. So by cutting under the chip there is a good chance that a few pads have been ripped of the PCB. To inspect the damage the chip needs to be removed with a SMD hot air station.

Edit:

For some reason, not only is the glue that is holding the Cell/B.E. in place UNGOLDY strong, the clearance in which you would have on any other unit is MUCH slimmer than anything I've previously seen to break the seal of the glue. I have tried everything. Pallete Knifes, (which I DO NOT want to use going forward. I'll get to that) an 8 guage E guitar string, fishing line, unwaxed dental floss, everything.

I just realized when you were talking about the Cell processor and not a coin cell here. The Cell Boradband Engine is the consoles CPU and it is not held in place by glue but by solder balls. I hope you did not try to break of the CPU or parts of it with the tools you mentioned.

[samnmax]

I just realized when you were talking about the Cell processor and not a coin cell here. The Cell Boradband Engine is the consoles CPU and it is not held in place by glue but by solder balls. I hope you did not try to break of the CPU or parts of it with the tools you mentioned.

He was most likely trying to delid the Cell CPU, removing the heat spreader so that the heatsink contacts the silicon die directly. It seems the glue holding the heat spreader is very strong.

[Traceless]

He was most likely trying to delid the Cell CPU, removing the heat spreader so that the heatsink contacts the silicon die directly. It seems the glue holding the heat spreader is very strong.

*Nods*, that's what I was referring to when I said "remove the CPU or parts of it". Either way delidding a BGA chip while it is soldered to the board is a very bad idea just because you will exert excessive force on the pads. The risk of damaging the PCB irreparably outweighs any potential benefits of this procedure by far, especially when considering the console and CPU were designed to operate with the lid attached. If one really absolutely has to delid a CPU remove the CPU from the board, put it in a small vise (similar to this) and perform the procedure off PCB. The problem is that the equipment to do this properly (SMD rework station) will cost you four-digit figures and you can still heat damage the chip and/or PCB if you are unlucky.

[grantbennett@me.com]

Hey there, apologies for the late reply. And yes, my aim was to just de-lid the Cell in effort to reapply thermal paste given that the console's fans were jet engines upon use. While I'm full aware of the risks involved, however if done properly this procedure far extends the life of the console by bringing down the temperature. I'm going to compare the torn components with a donor board I have to see if they can be swapped. I'm also going to have access to a component checker most likely tomorrow so I'll get further information on the torn capacitors/whatever the hell those things were.

[grantbennett@me.com]

Ok awesome! So after hunting around I found the information regarding the capacitors in another schematic. It reads that I need a 0.1uF 10v capacitor. And it also is showing that the ones I scraped up are all that value capacitor. Any good reccs on which exactly to buy? I found the listings on Mouser. Any input? (I've included my findings below)

[Traceless]

Great, having a schematic around is always a big help. Mouser is a good source you'll likely get quality caps pretty fast. Of course you could also order at RS components or Farnell or others (I'd decide based on handling fees/shipping costs). If you have more time and want to save money you can get capacitor kits like this one on Aliexpress very cheap. Delivery time will be probably between two and four weeks though. The good thing about having those kits around is that you'll likely have the right size cap around when you need it next time.

Any good reccs on which exactly to buy? I found the listings on Mouser. Any input? (I've included my findings below)

Looking at your mouser screenshot you have 01005 and 0201 capacitors and one 1210 capacitor in the basket. The 1210 is pretty big, but the 0201 and 01005 are microscopic (as in you'll need a microscope to solder them). It is hard to tell from a photograph but the components on your photo look larger to me. I guess they are at least 0402 (imperial) probably 0603 (imperial). For reference here are some components on the tip of a match. Take a look at this chart and order the right sizes. While your schematic screenshot shows sizes for C1509, C1510 (size 1608) and C1511,  C1512 (size 2012) I don't see any sizes for the caps you need. Also note that your schematic uses metric sizes whereas mouser uses imperial so 1608 metric is 0603 imperial, 2012 metric is 0805 imperial. The translation is also in the chart I linked. If you're unsure and since those SMD components are cheap you can simply order the same components in different sizes since and then just match afterwards.

While I'm full aware of the risks involved, however if done properly this procedure far extends the life of the console by bringing down the temperature. I'm going to compare the torn components with a donor board I have to see if they can be swapped. I'm also going to have access to a component checker most likely tomorrow so I'll get further information on the torn capacitors/whatever the hell those things were.

Disclaimer: I still strongly recommend against the delidding a) for the reasons stated above b) the board seems to be valuable to you which does not make it a good playground for experiments/learning. Also I personally never delidded a PS3 CPU so note that the following information are solely based on information I found online about the process. If you proceed with the delidding you do so at your own risk!

Okay so based on your experience you already found out that just cutting of the lid is not gonna get you far. If CPU/board have survived the procedure so far chances are you'll destroy CPU and/or board if you proceed with the current method. I found a PS3 delidding video on youtube from an austrailian tech who apparently works on consoles for a living (link below). He uses an SMD hot air station to heat up the glue attaching the IHS to the cell die. Once the glue has the right temperature it seems the heat spreader can be simply lifted off without any noteworthy force. The problem with that video is - we don't see which temperature he is using. Since CPUs are heat sensitive using high temperatures could damage the CPU. The risk of damage increases the longer the CPU is exposed to those high temperatures. I guess that the glue will soften well below the melting point of solder. So what you could try is to get an SMD hot air station (cheap ones are available on for instance on amazon), set it to the lowest temperature setting which for mine is (100°C) and then try to soften up the glue. If that does not suffice you can go up in steps lets say 120°C, 140°C, 160°C etc. until the glue softens. If you take too long give the CPU time to cool down and then resume at the latest temperature step. Alternatively you could also google for the correct delidding temperature. Be careful with the prying tool and don't shove it to far under the heat spreader to prevent damage to the CPU (remember: The Jedi approach is not going to work. So use heat and patience instead of the force .)

Looking at the video taking of the heat spreader is the easy part. Afterwards you'll have to remove the glue residue. You could try to wipe it of with IPA but since the tech in the video uses a razor blade I assume IPA wont work. The problem with the razor blade is: If you chip off the edge of the CPU die it is game over. Good luck and let us know how your project turned out!

[Norbert67]

Yeah, I know Your pain.
The glue on the GPU is different - its softer and its easier to remove. But the one was on the CELL processor on the other side....
I don´t know if only mine was like this, but the glue was like ceramics. Even with surgical knife I could barely make a visible cut. I was patient, tried with hot air, and insulated the other parts of the pcb with capton foil, but nothing was as easy as on the videos. I took my time - almost for a month, every day tried something. Later I DIYed a tool for that, but in the end just before the lid came off, I scraped the upper layer a little bit and did not worked since. 
So yeah, I  know exactly what You are talking about - and maybe You will have a better luck, so be cautious because before the lid came off there was no sign of the glue softening or a feeling of less resistance, just suddenly popped off.

[exige24]

Hi guys,

Sorry to resurrect this ancient thread, but I ran into a similar problem trying this same stunt that reading the thread almost answered for me with one caveat. I broke off two capacitors trying this also:





And as you can see, one of the capacitor values in easy enough, but the other is rated XX. Any help on what that could signify?