"well ackchyually" rtv drys not from oxygen, but with humidity in the air. So if you live in dry climate, or doing this job during heating season, you might give it more time.
Ah...you are correct. Just looked it up again. In that case a moisture barrier is needed. That's easier to achieve, as water is a larger molecule than oxygen gas, but still not as easy as just vacuum packing. I should use mylar instead of the regular vacuum bags for longer storage, and add a silica gel dessicant pack. That'd do it. Thanks for pointing this out.
Can we take a moment to appreciate this dude's complete character arc? Went from complete electronic noob to basically an emminence in the coveted PS3 modding scene, and not only that he even went as far to document almost every step in the way, ensuring we don't fall for the same mistakes he might have. All for the sake of fixing the YLOD. Dog bless.
@@ripfelix3020 isn't possible to remove all the capacitors and other perishable components and store the motherboard in a proper place to have it for decades?
@@gabrielmanarte6809 we have the schematics. It’s not impossible to create a reproduction mainboard. Or better, I bet in 15 years we are gonna have BC PS3s built from 40nm CELLs or 65s with brutally reverse engineered syscons. Or better yet, “Syscon’t”s. Syscon in FPGA. Someone will def reverse engineer the COK-001 definitely. It’s a passion for a lot of people.
Last year I decided my childhood E01 was too loud and learned about delidding. Bought a heat gun and some painters knives and managed to get it off clean and got everything repasted and running so so much quieter. It ended up ylod a few weeks later because of a failed capacitor on the board. I assumed it was that I accidentally dumped too much heat into the gpu. I bought a few more BC consoles and delidded them great, learned to read syscon from your videos, and found out that the issue with my original E01 wasn't a dead gpu. Sent it off to Studio Halabi and he got it fixed for me. I love learning about the ps3 and hope that I can one day get into frankensteining the models I do have whenever they bite the dust. Thank you for these fantastic videos that do so much for the community!
I’ll never in my life do any of this stuff on your channel. That said I watch all your videos to the very end bc they yell passion and love of the craft. Keep it up brother.
In this case it became necessary to test the speculation and competing hypotheses. Too many people were being mislead by clues that could be interpreted to mean whatever they wanted to believe. A confirmation bias rich environment. So without a clear explanation with data to support the conclusion, people default to their desired outcome. I needed to know too. I wouldn't normally go to such effort, but this was causing alot of people to buy and fail to fix backwards compatible models in the hopes of fliping them for profit.
A perfect takedown of all the idiots drilling holes in enclosures and those who blindly swap power supplies because a higher model number means it's better ;-)
I once brought a PS3 where the back of the console was completely melted. I don't mean a bit melted, I mean all of the vents were drooping down. The top and bottom housings had fused together. My best guess is that they tried using a hairdryer to fix the YLOD without actually taking their console apart. Some people don't deserve to have consoles.
I wouldn't call people drilling holes in the console idiots, that is too harsh. They were misinformed by somebody else (who themselves likely had bad information) and made a poor decision that worsened their game console. They took a risk on repairing their failing device and it didn't pay off, but that doesn't mean they are idiots.
@@CallumFaulds1 While that is certainly a dumb idea to fix their console, I would suggest not saying they don't deserve to have consoles. They took a risk on repairing their defective-by-design console (if I'm understanding RIP Felix's videos correctly), which is already a bold move, and didn't realize the implication of the fix they were attempting to make. I'm sure they realized this, and they would not repeat that dangerous mistake. But saying they don't deserve to have consoles seems a little too rude.
Not intentionally defective, no. It was an unforseen, but not unforseeable defect that afflicted many in the industry. IBM knew better tho and avoided the issue, proving not everyone missed it. About drilling holes. I'm not condeming people that do it. It's harsh to call them idiots. I drilled holes too. It was a product of a time when we didn't know what exactly the issue was, but just knew that it had a temperature related component to the failure. So we just assumed overheating issues and circumvented the circulation pattern. It does dramatically cool the RSX and Cell, but the PSU get's hotter. People like NSC proved that and started adding fans to them. There's a popular mod in japan where they drill another hole in the top shell and mount a PC fan over it! But what no one checked was to see if there were other components cooled by the RF shield and passivly by the intended airflow pattern. I did that and shoed both the VRM and BD ICs get hotter as a result. It could be a tradeoff people are willing to make, but we don't know for sure that cooling the RSX more prevents YLOD anyway. And if it does forstall it, by how long. We onky know there is a thermal related defect that is part of the equation and we can lower temps to "hopefully" increase reliability and the time we have before the YLOD.
My biggest regret is not being able to see your video earlier today.. A few hours ago I delided cell ihs but goes wrong, YLOD and got ripped traces :/.. Though, nice video !
Ive never saved a video so fast in my life. Bought a launch model 60gb from Japan last year and plan to spruce that baby up right. Will 100% be consulting this video beforehand.
This video was suggested to me by youtube out of no where, and man, this is THE VIDEO, great video, I enjoyed every single second of it, thanks a lot for all those information
This is the 'Perfect Guide', you point out every small detail, and I think that's great! Keep up the great content! I reckon your saving a lot of peoples time and saving a lot of consoles from being ruined or being left to die.
Bro I LOVE how in depth you are on these videos. I’ve always been fascinated on not only how to repair electronic failures but the true reason for failure. Like yeah I know that the error is reporting a loss of communication, BUT WHY? Please keep doing videos like this! This really scratches an itch on my brain that most channels just can’t get right.
this is an absolutely amazing video. not just an intensely comprehensive resource on ps3 maintenance but honestly saved me from working this much on my ps3 myself because i 100% would have broken it this way. thank god i watched it in full when i did, out of a tepid curiosity
I think it would be a very good consideration to look into phase change material, such as PTM7950. I have this applied on my desktop (i7 12700K @5.1Ghz) and it works like a charm, and it is inherently reusable due to it hardening back into into a pad once cool, so you can scrap it off and put it back onto the die
you know it's serious when you find yourself in 20:52 vacuuming syringes with silicone. But shit gets really personal when you prepare yourself for a CellBE-apocalypse at 20:56 xD damn i love this dude
I just didn't want to waste the rest of the tube. You can get smaller quantities, but this size is the best value and I wanted to see if I could repackage it such that it wouldn't go to waste. So far it's been working well. Also, Thanks for the super! Very generous of you.🙏
the detail in this video alone, and potential videos to come is fantastic. I know that if i feel the need to delid in the future, i can rely on this video for the best methods and info. Good stuff.
FYI - When using liquid metal thermal paste, It's totally safe, and honestly a good idea to grab some cheap clear fingernail polish. Use the brush, and put a coat on anything that could be shorted out, if the liquid metal seeps off the die. Just don't put the stuff on the die itself. I did this when i used liquid metal on my 7950X CPU. There are a lot of contact pads, and surface mounted chips around the die. The nail polish will create a coating that stops the liquid metal from shorting anything out. Also by doing this, you can be slightly less scared of using tiny bit extra liquid metal, hopefully allowing you to not have to re-do the thermal paste as often. on the PS3 CPU, I'd honestly coat the entire cpu ( except for the die ), at least, after testing to make sure the silicone doesn't have a problem sticking to it.
I've tried that and it burns if you ever need to replace the RSX. Leaves behind a god aweful black residue that doesn't come off! I'd use UV curing solder mask instead if you intend to use LM on the GPU die. Not that it's necessary, mind you. Also, it can still run out the edges and get underneath the BGA if you apply too much. So LM is still a quite risky endeavor.
@@ripfelix3020 I'm curious but have you tried using PTM7950 and some thermal putty as I heard it improves cooling a lot? I've been interested lately in repasting laptops and I've come to learn about them recently.
I used 0.06mm molybdenum cutting wire to delid my Cell, and it didn't cause any damage to the substrate. It did take a long time, but it felt very safe, and the end result was perfect.
@@ripfelix3020 wouldn't it make sense to use that wire to slowly cut away the vram glue? probably with some kind of jig as to not cut into the dies. iirc the vram has quite a bit of plastic on top so cutting it on the corners should be "relatively safe". I only have my slim where i want to put on ptm but since it still works i wont dare touch it :P
Yes, the encapsulation around the corners can be cut into a small amount without reaching anything important. The razor blade method only works because of that, because as you start rocking it into the adhesive there, it will round the corner of the VRAM a little bit, but never causes an issue. I highly doubt cutting wire would be practical on the VRAM tho. There's only about 0.2mm of space that the thermal plaster fills. So the sawing motions will both sratch away at the IHS and upper surface of the VRAM (which might be fine), but it's still too hacky and tedius to be a legit strat.
@@ripfelix3020 yea, thats why i tought about a jig, cutting a bit into the ihs should be fine as long as the important components do not get fucked. If i had a cnc and some broken consoles i would have experimented with making a jig :D Another idea i just got in might be the safest method available: a small cnc that removes the area where the glue is. There are chinese cnc which sole purpose is to precisely grind away glued on ic on mobile phones, why not use that on ps3? :D I would rather grind away the heatspreader and make a replacement than to use the "butterknife" hack (which i did sucessfully years ago on a fat) on my slim. I think putting on that honeywell ptm stuff and sticking it back on afterwards should fix the issue for a long time, or?
Supposed to be good long lasting stuff. I bet it would work well.I got some, but it was from amazon and I think fake. I'm thinking of getting some legit from mouser and comparing.
I no longer fear the YLOD after realizing that I'm also going to die someday. I might as well use my PS3 and have it die before me than the other way around.
Nice vid Felix! I'm thinking it would be a good idea to do Honeywell PTM or the Thermal Grizzly Kyrosheet as I don't want to delid in the future if I don't have to
Update: I apparently misremembered, the backplates attached to my RF shield didn't actually have any such additional protrusions, just the small nylon feet. Also it turned out that I didn't put quite enough paste on the RSX and Cell dies the last time I had it open, so it's actually a very good thing I re-opened it to apply that gasket stuff you recommended to remake the cell processor supporting seal. This video might have indirectly saved my PS3 from myself. Original comment follows; I've actually delided both at this point, some time apart; the Cell a couple years ago and the RSX just yesterday. I may not have done it for the smartest reasons (You specifically call out deliding the RSX not being appropriate for preventative maintenance) but it seemed to go well. That said, I think I'm going to need to open it up again and apply your recommendation of remaking the silicone seal on the Cell. It has worked fine since my delid, but it's a backward-compatible Fat model and I'd like to give it the best chance of longevity I can. I considered ordering a brace for the backplates but, curiously, I think the integrated backplates on my model actually already have a protruding ring of metal that sits within the hole in the motherboard around the back of the cell and RSX. I don't know how common this is, I'm in a PAL region that only received the later backwards compatible PS3 models if that makes a difference. I'll double-check the next time I have it apart to re-apply the silicone seal.
I genuinely love you showing steps that are irrelevant to most viewers - packaging syringe after syringe of rtv for future use, which I'm sure *you* will get through, but I'm just picturing Joe delidding and relidding his PS3's heatsink after trying it a few times and maybe getting through a couple syringes.
Have you considered having the bracers machined? Not for helping thermals at all (Doubt you would want them machined from metal anyways), but more so for dimensional accuracy and rigidity at higher temps. Of course it would cost way more than a 5 cent printed part, my guess is at most 10 bucks for a one off. Anyway, Excellent video as always Felix
Great research, this will be immensely helpful when delidding mine. I went the route of buying a Slim but it is giving me issues, so will prob take the old one and delid. I will say that if the IHS is mostly stable, gluing seems unnecessary. Maybe it distributes force better but I don't think it will matter much. I agree it is important to keep it, as without a socket, there will be stress on the solder joints. I would also not reach the same conclusion on drilling holes. In semiconductor land, 35 vs 45C is irrelevant, especially for VRM chips. If that is the temperature, it will shorten the lifespan by 0. If it was 95 vs 85, that would be a huge difference. Could be some added stress for capacitors but would need to check if they run close to rated temps or not. Unfortunately the PS3 will always be a mess to repair. Thankfully RPCS3 exists and CPUs are getting faster. Thank you for all the work your doing to preserve the hardware. Getting some still working will always be important.
I read it's something like for every 10c increase the chance of death doubles. I'd have to look it up for a MOSFET, but 10c more is not good. Now one could make the argument that replacing 2x VRM modules instead of the RSX is a worthwhile tradeoff, but I would argue that ignores nuance. It reduces the complexity of what kills the RSX to a single factor. Heat! While that scratches a psychological itch for the world to be a simple place, that's not the reality we live in. PS3 is not a place where problems reduce to a singluar cause, with a simple solution. It's always a combination of unknowables and calulated probabilities. Pros vs cons or give and takes. Let's say the RSX would die in 5 years if you did nothing. So you drill holes and now it lasts 7 years, but the VRM die in 5 years instead of the RSX. Ok, you replace them, which is easier than the RSX! Then the RSX dies 2 years later and it needs replaced anyway. Was it wort it? Or would it have been better to only have 1 thing die and need replaced? OFC this is a hypothetical, we can't predict. Point is that wishfull thinking always makes the assumptuon that it'll be better, not worse. But it's just as likely to be worse. The fact that we (humans) assume our luck would fall on benificial side is a pshychological bias called gamblers ruin. Or Confidence bias. We overestimate our chances, content in the lie we tell outselflves that "it'll be fine. Felix is making a mountain out of a mole hill." Until disaster strikes. The astute observer of human behavior will see this pattern repeat like a poorly coded AI NPC. It makes us easy to predict and fool. All a deceiver has to do is figure out what we want to believe, and tickly our ear. But the convient lie is a jedi mind trick for the weak minded.
I love how much effort was made to make point to glue IHS back on and admit not doing that on own consoles.. as honestly there is no good point to do so from any point of view other than assumptions and guesses why it would be there other than just keeping it still. But give points showing how do gluing well. Though i want say when selecting silicone is best select one which wont be corrosive for electronics aka wont create acetic acid during curing process.
I killed a fully functional 90nm RSX when I attempted a delid (cracked interposer) with the popping off method. It sounds easy in theory but is harder in practice. Don’t use a screwdriver to lever on it, it puts to much pressure on interposer. In my opinion a butterknife is better due to it having more surface area and therefore it puts less pressure on one spot.
Yes, if you use a screw driver, a palate knife undeneath will protect the Interposer. Squeept does it that way and it works fine. You just need to have something thin and stiff enough that it can fit between, but still have enough rigidity to be levered without bending too much. And it needs to get far enough under to lever, not just dig in and scratch the traces.
First couple ones I've had were already delidded since I've bought them from someone in my city, and he's been working on his own delidding tool similar to the one you use, gave me a prototype of it along with a couple dead COK-001 boards (868-22 that had popcorned during prebaking), had very little issues testing on those. I do have a spare B01 that I can try it on after testing more on the other dead board. Never done anything like the case mods or eraser, and all the BCPS3s I've bought came with 226s already, which I suppose is convenient. It wouldn't make much sense for you to test but some C and G models from PAL regions had the APS-227 which is the same as the APS-226, but only takes 220-240V input. I'd doubt it'd have any real difference in efficiency/operating temperature from the 226, but could be something to test on my own later if I decide to get a C model. Mainly get A and B models since some of the PS2 games I have require the EE to be present as well (and because Japan only got those models and Japan is the cheapest place to get consoles from)
Couldnt you have uploaded this video two days earlier, before I listened to bad advice and cut some traces during the cpu delid? Well guess I am practicing microsoldering and trace repair next!
I love the content we have gotten from RIP Felix's dive into the blackhole of PS3 YLOD, and how to fix them. For me, I just bought a PS2 slim, and a PS3 slim, and have never looked back.
MX-4 should never be used when applied to a bare die. I have done this many times with laptop dies and after around 30 days of use, the paste has already been pumped out which leads to the same thermal issues as before the repaste. Instead for direct die applications a paste like MX-6 should be used which is less liquid and more solid. Arctic does not recommend using a tool to spread out MX-6; rather, you should use the pressure from the cooler alone to spread it. MX-4 is fine when used on a large IHS like a desktop CPU or the RSX/CELL.
pumped out means you used too much (some are expected, but not too much) -- don't forget: thermal paste is only intended to fill the gaps between interfaces (eg cpu and cooler) not to be another interface/medium; adding too much will reduce the thermal transfer
Hi Felix, I'm really glad this video has been released now. It can at least dispel all of the nonsensical misconseptions and BS advice thrown around in Reddit and such. I do have some things I'd like to know about though 1. I have actually delidded a CECH-E01 PS3 successfully, but opted not to re-glue the IHS since I frankly I was just lazy didn't feel like it (Buying the silicone sealent stuff is extra cost after all) but I actually still had temperature issues. Eventually, I was able to figure out by asking around that HY510 is pretty garbage paste and doesn't work very well on the PS3. But during my troubleshooting, I ended up removing the original sealent in an effort to see if it would make a difference. Oddly enough, I think my temps have actually gotten a bit worse from removing the original sealent. Is it because my CPU's IHS is now sitting on the die with uneven pressure or something? I may have to re-glue the IHS if this is the case, since the original sealent is now lost. The Vejle XCGPU on the Xbox 360 S/E is an example where the sealent is necessary, but it's also a MCM chip with two dies of different heights. 1.1. When re-gluing the IHS, what if you just do four corners instead of sealing the whole thing? I know it's what Octal450 personally suggests on the Xbox 360 S/E since it allows you to remove the IHS in the future. 2. Does heat actually help when cutting the CELL silicon? I see it recommended by people sometimes, but I'm skeptical if it actually helps compared to just using a good cutting tool. Is the CELL sealent even temperature sensitive? 3. If you're gluing the IHSes of the CPU/GPU back on, is it necessary to use clamps like in the video? What if you just assemble the console back to as it was normally? Would the pressure from the heatsink clamps be enough?
if you assemble the IHS will sit flat against the HS, not necessarily the die. You want it flat against the die while the glue cures. I changed to the clams due to this issue. Follow the same glue pattern as SONY used. I respect Josh, but when it comes to these things I default to the OG design when I'm not absolutely certain there's a good reason to depart from their design. I don't know if heat helps soften the silicone, since my tool doesnt require any help. I dont want to add heat anyway. I know people say it helps, but I think if you need heat, then your tool is wack.
IDK if anybody has mentioned this, but GD900 is a great paste for the price. I've compared it to my mx6 and couldn't tell a difference in temp, but iirc the cheaper paste did better than mx-6 when it comes to pump-out over time. Right now my 6700xt is running around 5 degrees hotter than when I first pasted it around 6 months ago with mx6. I'm sick of repasting every 6-12 months, so I'm going to try some ptm 7950. People say it's less likely to have issues with pump-out. Anyway, well done with the video.
39:40 i think that is important to consider the temperature of your region too, mine bc ps3 can pass 50% on hot days here in brazil to keep cpu below 65/66 Cº, but in cold days (which are rare) the speed can drop to under 40% without changing anything
Hi man! From seeing your ps3 videos, I understand you know your way into electronics. Even to the point of proving people wrong and finally putting an end on the wrong diagnosing and "fixing" of the ps3. You probably have better things to do than breaking your head with another console, but there's this other issue that seemingly no one knows how to fix on the original xbox, the tiple boot frag, almost every console plagued by this is deemed a loss. This is a rare-ish error where the xbox fails to pass the self checking test and doesn't give any type of indication to the actual failing part. No display no sound no error light or code just red and gren flashing (frag-ing [or Flashing Green And Red] means general system failure, xbox doesn't get to post). I'm losing my mind with my xboxes like you did with your ps3s some time ago, I hope more light gets shined on this. Thanks.
I have several friends deep into OGX and they are on the trail. I'll keep my ear to the ground and at some point I intend to attempt the 1.4GHz CPU upgrade, steller, HD+ mods. I'll dive deep, but I'm still captivated by PS3 and want to complete several ongoing projects with it before moving on to another console.
Man, the sheer increase in console hardware fragility/complexity between 6th and 7th generation is staggering. Even something as simple as a routine replacement of thermal paste requires specialized tools and surgeon-like precision to avoid destroying your machine. I've done lots of amateur work on 6th gen stuff and earlier to save myself a bit of money and learn new things, but after 2005 all the manufacturers really put their foot down. I'm almost glad that consoles seem to be dying, at least in 20 years when all the Xbox Series consoles start to fail from who-knows-what, nobody will have to scramble to save them to keep their nonexistent exclusives playable. I dread the day when I can no longer boot up a quick game of Tokyo Jungle.
For those wondering what kind of tool he was using to delid in the video to me I believe it was a metal nail file because I searched it up on Amazon since there was no link in the description for the tool he used or I’m just an idiot
Recently got the courage to repaste my Phat PS3 (non backwards compatible model). Own the original box and it was running hot even with web man. I recommend it, maybe not delidding unless completely necessary, but for making sure it makes less noise.
@@ripfelix3020 I think its dependant on the user, my PS3 even with the thermal paste replacement (65nm RSX) stays at 35 percent fan speed mostly. It then sometimes when I play intensive games shoots up to 50 percent fan speed; it slows down back to 35 after a few seconds though.(this usual happens every 10 minutes) I don't mind it but I can expect some people might not be too pleased with most consoles being relatively quiet.
Have you tried thermal pads instead? Curious to see a comparison. Also in your comparison clips here the RSX is always 5-10c lower than the CPU… my RSX is usually around the same temperature as my CPU.. it easily stays below 68c at around 30-35% fan speed but is my RSX too hot then?.. mines a Cech21xxx Slim.
There should be a der8auer style delid tool, there is one for the RSX chip, but not the CELL. Which needs it the most. It would be a great way to get away from Razors and Paint Knives which can render your PS3 useless after a slight mistake. And as for applying thermal paste I recommend manually spreading the paste with a spreader tool. From my test my temps were dropped by 10 degrees vs the peasize method.
Those wont work on board. And even if they did, it'd place too much force on the board/interposer for my commfort. The 3D printed one ebay is not good. It's the same as the pop method, except the plastic wedge will get hung up on the edge and not work. And the plastic springs off. It's trash.
@@ripfelix3020Yes I have not tried It. I think the RSX tool is pointless, RSX's are a lot easier to delid. But i can't trust the pressure it will put on the BGA solder underneath with the Cell CPU.
when repasting a bare die like on a Delid i like to make sure to spread the thermalpaste so the entire die is covered before putting the ihs or heatsink back on it that way i ensure that the entire chip is covered as simply putting on some paste and hoping it spreads on its own is not a 100% guarantee and you *may have a uncovered corner* , wich can lead to a dead chip certain components of the cpu like the memory controller love to sit in those corners and if that part of the chip cooks nothing will bring it back
@@ripfelix3020 that usually only happens on larger surfaces like the whole IHS and if you have a very uneven application and relatively low pressure all of wich you dont really have on a PS3 Direct die or GPU's
I don't do this as a service. I dumped a bunch of info gathered over the years into this video. It seems like alot, because I expanded it out with all the stuff I wished I knew before trying my 1st time. Things that need said and most don't. No need to be scared. Just onboard what you can and keep watching. Eventually you'll feel comfortable with the process and be ready to try it on a practice board.
Hi. 2nd time around here. I want to differ in something...... I have an E model,and i did almost all of this. Only i need delid Cell and new thermal pads. And i did some extras too. I used the last ps3 fan,the white label DC brushless from last models (make long cables for to attach) and used thermal grizzly. And my 226 psu flopped,so i have 231 instead,and i got 62 on Rsx,73 on cell at 40x speed at maximum. I don't think 231 is the worst one (ZZR with 19 blade fan are truly the worst option) And that's for me. Also i did remarry on BD drive,my 410 lens is great. Maybe one day i can get a frankie mod for this beast.
I am quite sure the APS-231 does run hotter in a BC model. Both the 226 and ZSSR run cooler in this scenario. Though the 231 does work, there's no telling for how long. About the fans from later models. I tried a G14T S1, it's the quietest fan. But the issue is that it's weak. In order to push the needed volume of air, it needs to run at like 40-60%, which is just as loud and doesn't have the headroom at 100% to push much more air. So if ever the higher percentages are needed, like when past wears out or a really hot day, it can't cool as well as those more powerful fans. Fun fact, the G14T and D14F have the exact same 15-blade blade. It slips on the rotor. The only difference is the motors power.
How reliable is the 65nm CPU/GPU? I have a PS3 slim(CECH-2004B) I bought in 2019 second-hand CIB with 2 OEM Controllers and to my surprise it was only used for 30 days (Total usage I got to know from webman settings which is readings given by syscon) to this day the total use it has had is 57 days (including the time I played on it). I read somewhere that this Slim model(CECH-2004B) is the "earlier" slim model which is one of the few able to launch full CFW and has 65nm chips that are evidently easier to delidd than later slim/Super-slim revisions. I don't play on it a lot to personal irl reasons but the temps on it are never good. Due to this I repasted it 4-5 times with MX-4 over the last 4-5 years of owning it(because I was always unsatisfied with the repaste application) and finally was satisfied with the repaste method I used last year. To give an example regarding bad temps I booted Last of us for the first time with Syscon fan settings and 5 minutes into the game it easily reached 81C GPU and 78C CPU. I have no idea what to do. I don't even how to soldier much less Delidding. Living in a third world country the best "expert" I could find for the repair of these consoles is an old sony technician that use to work on them and got my HDMI port fixed interesting thing is that he used a LIGHTER of some special kind to loosen the soldier joints at first I was afraid but later he used a small soldiering iron and I got it fixed. Anyway, Is it worth it for me? Or should I stop worrying about temps and just play the slim at 40% fan settings until it starts giving issues like a random reddit stranger suggested? EDIT: Wow, that's a wall of text. I really hope you'll read this and give your opinion as your opinion regarding this matters a LOT for me.
The 65nm RSX is a reliable chip, no need to worry about it dying from a heat related defect. I'm not sure what "normal" temps for a 20xx model slim are. I've never paid any attention it on mine. I just left it up to syscon to do it's thing. I don't remember there being any fan noise, it's always been pretty quiet. I should see what it get's up to. The only thing that model has that tends to die are the NEC/Tokin proadlizers on the RSX side only. But otherwise they are considered reliable. Most 25xx are jailbreakable, it's the late models that might not be. But it's actually pretty hard to find one. 30xx and later definately aren't.
you know what PTM7950 would probably be a really good solution and when it comes to ceramic thermal glue you freeze the heatsink, you do not heat it, that may work but you reduced the lifespan of the chip. In the field if a board fails ICT or FBT and a chip with a thermal glued heatsink is failing we use a freeze spray to freeze the heatsink and with a light twist or with light pressure from a pry tool it will release cleanly from the chip.
I tried a freze fracture using dry ice. The thermal adhesive becomes stronger and it guarintees the VRAM delaminates. You need heat to soften it if using the leverage method.
24:36 one question tough: why not mount IHS to metal shielding and apply paste into center of upside down IHS? it will still be present there, shielding also has to have right "curvature" (or rather of it) and we're saving ourselves hassle of deliding in the future when paste worns out. It will eventualy but at this time it will be as easy as catefully disassemble the console, wiping our solidified old paste and reapplying new, right?
2 issues. 1 the IHS will sit flat against the HS, not necessarily the die. I've seen that cause bad cooling. 2, most importantly the IHS is a stiffener. It provided structural support to the Package, likely increasing reliability which is the entire purpose of delidding in the first place. Why replace the paste to lower temps 3C, for an unknown longevity result, then not glue it back on reducing reliability by and unknown ammount? It's nonsensical until you realize that people are not making rational descisuons, they're making emotional ones. Choosing the bad mods they "want to believe" in, and disregarding good mods they "should." Why? Usually convenence and fear. They fear the extra 3C from a lower thermal conductivity paste, but that lasts longer. And instead choose Liguid metal, which puts the entire console at risk of being destroyed if you don't contain it properly. They don't want to have to put the IHS back on, so they choose not to believe it's necessary, instead coming up with all kinds of rationalizations about how SONY's thermal design teal of engineers were stupid and that dude on reddit knows better...and so on. It's a pattern of behavior that you'll see in humanity that's never going to change.
At 36:22 , are those 3d printed fan blades for the Nidec fan? I thought about this a couple years ago, and I am curious if you have found any benefit from different blade designs.
Zero. I tried many itterations and every one of them was worse is some way. We came close to matching the performance with one design, but not quite. Clearly Nidec simulated and optimized the blades they used. It was an intersting project tho. Gave me new respect for the engineers.
@@ripfelix3020 Interesting! I wanted to try this but didn't have a printer at the time to experiment. I guess it was time better spent elsewhere then. The Nidec guys clearly know their stuff.
I remember I was playing midnight club los Angeles and I thought to myself that the fans were a bit louder than usual i should have turned it off there
Just curious why use clamps to secure the ihs while the glue dries? Can't you use something else that will distribute the pressure so you don't need to use the 3d printed support plate. Is gravity not enough? Thoughts? I appreciate your videos!
When the paste was originally installed by the packaging facility the chips were not soldered to the motherboard. They could be clamped down with the pressure flat against the die and no worry about anything underneath the interposer. Like the BGA or the MLCCs on the backside of the motherboard. But we dont have that luxary! Well unless you want to desolder it first! So if you only use the backplate then now the pressure of the clamp pushing down on the die will press into the center. The interposer will flex under the die. The motherboard will bow under the die. That's because there nothing resisting bowing/sagging underneath the die. Only the 4 standoffs in the corners. When the glue sets the moment you release the clamp, that pent up strain will spring back and be locked in by the glue. And with thermocycling over thousands of cycles there's no telling how that would affect reliability. To remove that varible, I'm using these to keep the MB flat under the entire BGA region so that I know the paste and glue are perfectly flat while the glue sets. That way it'll be like it was from the packaging facility and I can trust the calculations/relibility simulations are still relevent to the design. I didnt change the a varible by locking in pent up strain.
I recently delided mine (just the rsx) and clean it properly. But i found an thermal pads (very thin brown) at the nec tokins under the cell and rsx. Is it normal ? From what i searched they mostly been found ouy on the other side... cechc03 model. The console was sealed and little bit dusty inside I assume it was never been opened before just good cared. 129 days in total, 30 errors in total mostly 1001 and only one 1200 from may2024. (bought the console in july.)
Someone added them after the fact. That's not OEM, no. The tokins don't dissipate heat throug their casing. It's just a resin cap, which doesn't even touch the metal part of the cap. It's an air gap. Lol. So putting a pad there does nothing. The pad under the CPU however is concerning. I've seen that cause BGA damage and cracked dies. People don't think it can. They get indignant about me pointing this fact out, but it can and has. So I do not reccomend it!
Since you're literally the PS3 god by this point, do you know any tool i could use to recover the whole HDD of my PS3? Something that can read all files off of that filesystem the PS3 uses. I have a Tool and the guy i bought it from only played PES without even installing it on the HDD. (His brother is a game developer and they got rid of the PS3 when the PS4 launched.) Presumably no data was re-written since it was first wiped. (That if the HDD hasn't been replaced with a new one.) I really want to recover the files and see what beta game i could find on there. Do you have any idea what kinda recovery tool i should use or if i should send the HDD to a data recovery service?
Unfortunately, the PS3 filesystem is encrypted on a per console basis. You need the EID Root Key (ERK). This can be dumped using CFM tools in recent versions of evilnat. With that you can decrypt a HDD using a tool (I'd have to look up it's name. Ask on psx place, they can direct you to it). Without the ERK, no. The encryption is solid and would take forever to brute force. The only option is to repair the console and get it to boot.
@@ripfelix3020 Oh, the console is not broken. It boots just fine. It's actually very clean inside too. I have it on its original firmware, some very early one. I forgot. 2.something... So i need to install a CFW, evilnat cobra that is. Hopefully it's available for DEX consoles too. Get the key, decript the HDD, then recover everything. Thanks. 😊
Hey man, I have finally decided to delid my PS3, after deliding my PS3 was having YLOD, so I decided to check syscon error logs, I'm getting error code : A0203010. Any help on how to fix this? I see no scratches on the CELL
You may have flexed the board and caused a CPU ball to break. That 3010 is one of the codes commonly encountered after a failed cpu delid. So if no traces were broken, no nearby SMDs are knocked or missing (check them carefully), then flexing could be the cause. That's why I use the MB brace to keep it from flexing. Sorry you had this. It's one of the risks of delidding the cell. It's why I dont reccomend it as preventative maintenence unless you actually need to. And it's why I reccomend a proper delidding tool. The painters knife isn't my favorite tool either. I feel the way it inserts places too much force on the board. But others like it and swear by it, so I've acquiesced to their experiance.
hey Felix, how can I clean my HDD properly ? because it is doing some errors sometimes .. I think this happened after I used the app that frees 8% of the HDD data. I use the PS3 in that way (with only 1% left for the HDD) for some time, and after that I restore the 8% of memory back however it is doing some errors from then onwards.
I printed various revisions and sand them as I need them. Since I don't do very many consoles I dont really think about efficiency. I just print what I need and sand as I need to. Preferrably seldomly.
im curious if you would ever sell your expertise as a service for people that would love to get Frankemodded PS3s? i am a nerd, but im too scared to wreck the rare E01 that i have, and most Frakenmodded ps3s are near 2x the price of a good A01, also, graphene pads for the IHS might be a good option as they dont expire
I'm not much interested in doing them for a service, no. For fun is one thing. Selling to recoup some of my investment costs too IG at some point. But I'm mainly in it for the science and challenge. To defend sellers, they are too much work to sell for less than they're going for. I don't think people realize the effort and skill that goes into it. The tools alone are expensive. The education through failure. The risk of offering a warranty. And OFC the ebay fees. There's actually good reason they cost as much as $800. A service for $550 is actually a great deal and less than it was a year or 2 back, thanks to healthy competition in the space. That was one of my goals by making these videos and it seems to be working. They're already cheaper than they were! But at some point if prices fall too much or people are not willing to spend what it takes to make it worth a modder's time, the frankies will not be available anymore. I happen to believe in free markets and the law of supply and demand. So it costs what it costs, becauae supply is low and demand is high. It's natural and good for those who can make them to profit while they can. If it's too rich for your blood, that's because it's not for those who are buying them. Until they stop buying them and modders have to lower prices to intice new buyers, modders will continue to meet demand at a price that keeps them busy.
My PS3 was delidded by the previous owner and still sits in the 40-50% fan range to keep CELL under 68C. I've repasted it twice, but the issue still remains. It's not glued on, but I've just accepted it's just gonna be loud at this point lol. At least CELL keeps the RSX in the 50 range anyways
I base my 40-50% reccomendation off an ambient temp of 20-25C.If your room is warmer, or the cabinet it's in gets warmer, then that reccomendation is relative. Also, it only applies to A-C models. The rest have different TDP, fans/Heatsinks, etc. So "normal" is relative to each model too.
I've used graphite thermal pads before. They perform slightly worse than MX4, but they never degrade and are fine. Best, they dissipate heat laterally away from hotspots better than most mateials. Graphene would be better, but to answer your question. No, I wasn't aware you could get graphene pads yet. Might be interesting to try. I have some PT., but I suspect it's fake. Was cheap on amazon, but I think the genuine stuff is only going to be had from digikey/mouser and be honeywell branded. Last I looked it was much more expensive than the one I found on amazon. I might get some legit stuff and compare them to see. Might make for a good vid. IDK.
I've bought a cech20 and opened it to find it had thermal paste beside the 2 IHS, how can one screw that up, idk. Put 2 carbonaut pads 38x38 on it, so i don't ever need to replace, and can just dust off. If I were to buy it again I would get kryosheet, same price and way better, but I didn't delid and don't think ever will, so I assumed it would bottleneck under the IHS, and be no difference.
Felix I have five CECH-A01 models. I disassembled cleaned and delided both CELL and RSX on all five. (Practiced on three dead boards first). NONE of them had a backing bracer plate on the back of the CELL. Should I be putting a bracer plate under the CELL? 😯 I did not glue/attach the IHS back on the CELL I marked the IHS where the old paste was. I let the area under the CELL open how it was when I opened the console. Question could I put a thermal pad that is not as thick as the CELL board channel? So the pad does NOT rise above the channel and does not touch the metal shell motherboard shell? Basically sits in the motherboard void of the CELL. Yes Felix I bought an ruler sequenced in MM just to measure the thermal pads of the originals Sony used I am that guy.🤷 I have never replaced parts as the consoles never quit working. Replacing parts in the instance of replacing the fan or the PSU what the console came with I reinstalled back in the console. I have a CECH-A01 with the ZSSR PSU still going. CECH-A01 are my favorite consoles ever. You are correct all these consoles will eventually quit working. I want mine to last as long as possible. Enjoy your channel my PS3 favorite Felix!
Ok, first respect the CPUs "void." It doesn't like forign objects inserted there. You don't have to do it the way I do. I'm pointing out why I do it my convoluted and overly cautious way. If you want to be extra careful like I do, I'm that guy, then you can follow suit. But no, you can have exelent results doing it other ways too. Is my way the most proper? Let's just say I've developed this procedure to address "minor issues and potential concerns", not necessarily "must dos or it'll die for sures."
I have the 20gb B00 model it’s been a tank so far on webman after being reposted an delidded it stays at 60 cpu and 55/58 gpu tad loud at 45/50 fan in games but it’s worth it too keep they 90nm cool
And here i thought i was being clever drilling the holes under the console. So to sumerize, having the fans run fast at all times wont solve the low fan speed hole problem as the components like psu and disc drive are just not getting air?
The syscon bases fan speed off CPU and GPU temps. The hole cools them more, the fans can run at a slower speed to maintain the same temps. The airflow through the console is circumvented and components cooled by that pattern get hotter. Regardless how fast you run the console, the holes will circumvent. So if say before you got 50CFM through the upper compartment at 25% and 300 CFM at 50%, the circumvention will lower whatever the CFM would have been at whatever speed you set it to. Dramaticly lower CPU/RSX temps are possable at the same %, but what most people are doing is actually setting webman at 68C and the holes only make the fan less noisy. So it's not like the temps are lower, it's just the fan is quieter. That makes them feel better and enjoy the noise reduction. But it comes at an unknown longevity consequence. And I have seen the VRM burn out and need replaced. It's not very easy and requires lots of heat...BUT, "If" it could prevent the 90nm from dying (I'm not convinced it can) one could argue it's a better outcome. I don't like it anyway.
I used a # 4 palette knife which is thin enough and the same width as a nail file. Everything else I did as you have done. Lol I used the same exact tool to remove the silicone 😅
trim removal tools, smartphone opening tools... I find they're actually unhelpul for those things because they're made of such soft plastic, but if they were made of tougher plastic they'd probably scuff up my stuff, so that's a tradeoff I'm happy with
Is this a reupload? I saw one of your videos in my subscription feed a few weeks ago but by the time I collected my pizza and sat down to watch it it was gone.
That's a myth. I've seen some that were glued on by idiots and subsequent attempts to delid caused the die to break off. The OP assumed it was soldered on. But everyone I've spoke with that has delided slim and SS models says the never solder the IHS on. I left that door open for a while, since I dont personally work on those models, but at this point I need proof before I'll believe it.
@ In that case, eventually even the slim models will need to be delidded, huh? Can’t say I’d be excited about going that route due to how easy it is to ruin a console by simply mucking up the delidding process, though it’s inevitable that delidding may be required in the future… which I personally don’t look forward to.
@ Let us know if a delidding tool for Cell ever does comes to fruition, as having a specialized tool that could make delidding easier (without killing the system) would be a real blessing for us that still hold onto PS3 systems, as we can finally better preserve these systems without needing to take unnecessary risks anymore. I look forward to hearing more about the delidding tool that could finally end PS3 casualties.
Given that I bought a CECHG01 *used* in 2013, played BF3 extensively with it cramped in a cupboard for years with only a single crash, and *it still works* (albeit it only sat in open air since then), I don't know what to think. I should at least soft mod it finally.
I haven't delid my console at all. It is CECHK and it's running time is 400+ days. I haven't changed termopaste at all. Is it okay to continue that way since it's working just fine, and with the most heavy games it's turning fans no more than 31% even after 3-4 hours of gaming of games like last of us, gta v and la noire.
So much trouble deliding it i'd rather pay someone else who knows what he's doing, this is not as near and easy as deliding a pc cpu processor, i love doing things myself and learning in the process but damage a console that was working feels bad
What are the odds of 3d printing a motherboard brace? Basically everything now days is designed for 256x256x256 but I wonder if there's some design that could make it a viable option; acquiring a brace for piece of mind just to do one console is kinda a stretch for me. If not AliExpress will for sure be the go-to.
They aren't easy to get anymore anyway. Last I checked they werent for sale. At least not the ones I got. I know some people have made a pattern and use a sheet metal service to cut them. Like OSHcut. So yeah, you can have them made that way and for a reasonable price.
@@ripfelix3020 Thanks for all you've done and have planned man; I'll look into this once I actually set into my CECHE01, with the help and confidence your guides provide of course.
i dont think pre 80c temps are really of an issue on the ps3 or any actively cooled system for that matter, its not the heat that kills it, its that heat cycling that slowly degrades the components
Partly. The higher the temperature delta, the greater the expansion forces per thermocycle. So it's both. But that's just normal thermocycling and assumes there's not a packaging flaw. Like poor adhesion promotor that the underfill is layed on top of, or poor curing of the underfill. Or inappropriate choice in underfill. Etc. That's what happened with the 90nm RSX in some phats. 65/40nm GPUs were fixed and are reliable. They can be reasonably expected to perform at normal 75-85c without reliability issues beyond expected.
@@ripfelix3020 thank you for explaining it, i have the 90nm gpu. did a repaste and played through all uncharted games on it, was loud, and it did freeze once but it made it through so far.
"well ackchyually" rtv drys not from oxygen, but with humidity in the air. So if you live in dry climate, or doing this job during heating season, you might give it more time.
Ah...you are correct. Just looked it up again. In that case a moisture barrier is needed. That's easier to achieve, as water is a larger molecule than oxygen gas, but still not as easy as just vacuum packing. I should use mylar instead of the regular vacuum bags for longer storage, and add a silica gel dessicant pack. That'd do it.
Thanks for pointing this out.
Can we take a moment to appreciate this dude's complete character arc? Went from complete electronic noob to basically an emminence in the coveted PS3 modding scene, and not only that he even went as far to document almost every step in the way, ensuring we don't fall for the same mistakes he might have.
All for the sake of fixing the YLOD. Dog bless.
Dog bless 😭
Dog?
@@TheM0nkeyBomb bless
Dog bless
I don't have a phat PS3 and I already know I'm still going to watch all 42 minutes of this
If you want to be 100% sure that you never get the YLOD then simply do not use your PS3. It's that easy!
To clearify. The PS3 can still die in the box, but if you never use it, you'll never know it has the YLOD. Ignorance is bliss.
Exactly, the best option is clearly a Schrödinger's PS3 scenario LOL
@@alexsmith5584 ffs beat me to it 😂
@@ripfelix3020 isn't possible to remove all the capacitors and other perishable components and store the motherboard in a proper place to have it for decades?
@@gabrielmanarte6809 we have the schematics. It’s not impossible to create a reproduction mainboard. Or better, I bet in 15 years we are gonna have BC PS3s built from 40nm CELLs or 65s with brutally reverse engineered syscons. Or better yet, “Syscon’t”s. Syscon in FPGA.
Someone will def reverse engineer the COK-001 definitely. It’s a passion for a lot of people.
Last year I decided my childhood E01 was too loud and learned about delidding. Bought a heat gun and some painters knives and managed to get it off clean and got everything repasted and running so so much quieter. It ended up ylod a few weeks later because of a failed capacitor on the board. I assumed it was that I accidentally dumped too much heat into the gpu. I bought a few more BC consoles and delidded them great, learned to read syscon from your videos, and found out that the issue with my original E01 wasn't a dead gpu. Sent it off to Studio Halabi and he got it fixed for me. I love learning about the ps3 and hope that I can one day get into frankensteining the models I do have whenever they bite the dust. Thank you for these fantastic videos that do so much for the community!
RIP Felix is back and I take note. Can’t wait
I’ll never in my life do any of this stuff on your channel.
That said I watch all your videos to the very end bc they yell passion and love of the craft. Keep it up brother.
Such respect for someone who actually tests, learns, examines their craft. Some scientific method in the retro gaming/modding scene is fantastic.
In this case it became necessary to test the speculation and competing hypotheses. Too many people were being mislead by clues that could be interpreted to mean whatever they wanted to believe. A confirmation bias rich environment.
So without a clear explanation with data to support the conclusion, people default to their desired outcome. I needed to know too.
I wouldn't normally go to such effort, but this was causing alot of people to buy and fail to fix backwards compatible models in the hopes of fliping them for profit.
A perfect takedown of all the idiots drilling holes in enclosures and those who blindly swap power supplies because a higher model number means it's better ;-)
I once brought a PS3 where the back of the console was completely melted. I don't mean a bit melted, I mean all of the vents were drooping down. The top and bottom housings had fused together. My best guess is that they tried using a hairdryer to fix the YLOD without actually taking their console apart. Some people don't deserve to have consoles.
I wouldn't call people drilling holes in the console idiots, that is too harsh. They were misinformed by somebody else (who themselves likely had bad information) and made a poor decision that worsened their game console. They took a risk on repairing their failing device and it didn't pay off, but that doesn't mean they are idiots.
Bro called out Ryan (Mystic) 😭
@@CallumFaulds1 While that is certainly a dumb idea to fix their console, I would suggest not saying they don't deserve to have consoles. They took a risk on repairing their defective-by-design console (if I'm understanding RIP Felix's videos correctly), which is already a bold move, and didn't realize the implication of the fix they were attempting to make. I'm sure they realized this, and they would not repeat that dangerous mistake. But saying they don't deserve to have consoles seems a little too rude.
Not intentionally defective, no. It was an unforseen, but not unforseeable defect that afflicted many in the industry. IBM knew better tho and avoided the issue, proving not everyone missed it.
About drilling holes. I'm not condeming people that do it. It's harsh to call them idiots. I drilled holes too. It was a product of a time when we didn't know what exactly the issue was, but just knew that it had a temperature related component to the failure. So we just assumed overheating issues and circumvented the circulation pattern. It does dramatically cool the RSX and Cell, but the PSU get's hotter. People like NSC proved that and started adding fans to them. There's a popular mod in japan where they drill another hole in the top shell and mount a PC fan over it!
But what no one checked was to see if there were other components cooled by the RF shield and passivly by the intended airflow pattern. I did that and shoed both the VRM and BD ICs get hotter as a result. It could be a tradeoff people are willing to make, but we don't know for sure that cooling the RSX more prevents YLOD anyway. And if it does forstall it, by how long. We onky know there is a thermal related defect that is part of the equation and we can lower temps to "hopefully" increase reliability and the time we have before the YLOD.
My biggest regret is not being able to see your video earlier today..
A few hours ago I delided cell ihs but goes wrong, YLOD and got ripped traces :/..
Though, nice video !
RIP
The perfect guide about temperature control on PS3s. I have been wondering what pad thickness should I use. Now I get all the answers from your video.
Ive never saved a video so fast in my life. Bought a launch model 60gb from Japan last year and plan to spruce that baby up right. Will 100% be consulting this video beforehand.
same with me, its sitting in its box in my closet ready for the day i run a full maintenance on it
This video was suggested to me by youtube out of no where, and man, this is THE VIDEO, great video, I enjoyed every single second of it, thanks a lot for all those information
This is the 'Perfect Guide', you point out every small detail, and I think that's great!
Keep up the great content! I reckon your saving a lot of peoples time and saving a lot of consoles from being ruined or being left to die.
Bro I LOVE how in depth you are on these videos. I’ve always been fascinated on not only how to repair electronic failures but the true reason for failure. Like yeah I know that the error is reporting a loss of communication, BUT WHY? Please keep doing videos like this! This really scratches an itch on my brain that most channels just can’t get right.
this is an absolutely amazing video. not just an intensely comprehensive resource on ps3 maintenance but honestly saved me from working this much on my ps3 myself because i 100% would have broken it this way. thank god i watched it in full when i did, out of a tepid curiosity
Another fantastic video from RIP Felix!
I think it would be a very good consideration to look into phase change material, such as PTM7950. I have this applied on my desktop (i7 12700K @5.1Ghz) and it works like a charm, and it is inherently reusable due to it hardening back into into a pad once cool, so you can scrap it off and put it back onto the die
I have it on my Frankenstein PS3 and it works perfectly. I trust it to be good for longer than MX-4.
you know it's serious when you find yourself in 20:52 vacuuming syringes with silicone.
But shit gets really personal when you prepare yourself for a CellBE-apocalypse at 20:56 xD
damn i love this dude
I just didn't want to waste the rest of the tube. You can get smaller quantities, but this size is the best value and I wanted to see if I could repackage it such that it wouldn't go to waste. So far it's been working well.
Also, Thanks for the super! Very generous of you.🙏
the detail in this video alone, and potential videos to come is fantastic. I know that if i feel the need to delid in the future, i can rely on this video for the best methods and info. Good stuff.
Remember to vacuum seal your silicone syringes for easy meal prep
Haha...I bought my first vacuum packer not for food, but for this! lol
Awesome to see another video, Felix.
41:27 i felt that line so much having tried to fix a minor PS2 issue once and I ended up breaking something important...
Fun stuff, thanks for the video Felix!! Some old ass footage of me delidding in there haha
FYI - When using liquid metal thermal paste, It's totally safe, and honestly a good idea to grab some cheap clear fingernail polish. Use the brush, and put a coat on anything that could be shorted out, if the liquid metal seeps off the die. Just don't put the stuff on the die itself. I did this when i used liquid metal on my 7950X CPU. There are a lot of contact pads, and surface mounted chips around the die. The nail polish will create a coating that stops the liquid metal from shorting anything out. Also by doing this, you can be slightly less scared of using tiny bit extra liquid metal, hopefully allowing you to not have to re-do the thermal paste as often. on the PS3 CPU, I'd honestly coat the entire cpu ( except for the die ), at least, after testing to make sure the silicone doesn't have a problem sticking to it.
I've tried that and it burns if you ever need to replace the RSX. Leaves behind a god aweful black residue that doesn't come off! I'd use UV curing solder mask instead if you intend to use LM on the GPU die. Not that it's necessary, mind you. Also, it can still run out the edges and get underneath the BGA if you apply too much. So LM is still a quite risky endeavor.
@@ripfelix3020 I'm curious but have you tried using PTM7950 and some thermal putty as I heard it improves cooling a lot?
I've been interested lately in repasting laptops and I've come to learn about them recently.
Great video. Sooo much detailes. awesome. thank YOU for that one :)
I used 0.06mm molybdenum cutting wire to delid my Cell, and it didn't cause any damage to the substrate. It did take a long time, but it felt very safe, and the end result was perfect.
I've seen it go both ways.
@@ripfelix3020 wouldn't it make sense to use that wire to slowly cut away the vram glue?
probably with some kind of jig as to not cut into the dies.
iirc the vram has quite a bit of plastic on top so cutting it on the corners should be "relatively safe".
I only have my slim where i want to put on ptm but since it still works i wont dare touch it :P
Yes, the encapsulation around the corners can be cut into a small amount without reaching anything important. The razor blade method only works because of that, because as you start rocking it into the adhesive there, it will round the corner of the VRAM a little bit, but never causes an issue.
I highly doubt cutting wire would be practical on the VRAM tho. There's only about 0.2mm of space that the thermal plaster fills. So the sawing motions will both sratch away at the IHS and upper surface of the VRAM (which might be fine), but it's still too hacky and tedius to be a legit strat.
@@ripfelix3020 yea, thats why i tought about a jig, cutting a bit into the ihs should be fine as long as the important components do not get fucked.
If i had a cnc and some broken consoles i would have experimented with making a jig :D
Another idea i just got in might be the safest method available: a small cnc that removes the area where the glue is.
There are chinese cnc which sole purpose is to precisely grind away glued on ic on mobile phones, why not use that on ps3? :D
I would rather grind away the heatspreader and make a replacement than to use the "butterknife" hack (which i did sucessfully years ago on a fat) on my slim.
I think putting on that honeywell ptm stuff and sticking it back on afterwards should fix the issue for a long time, or?
Supposed to be good long lasting stuff. I bet it would work well.I got some, but it was from amazon and I think fake. I'm thinking of getting some legit from mouser and comparing.
Your content is absolute gold
I no longer fear the YLOD after realizing that I'm also going to die someday. I might as well use my PS3 and have it die before me than the other way around.
Nice vid Felix! I'm thinking it would be a good idea to do Honeywell PTM or the Thermal Grizzly Kyrosheet as I don't want to delid in the future if I don't have to
Update: I apparently misremembered, the backplates attached to my RF shield didn't actually have any such additional protrusions, just the small nylon feet. Also it turned out that I didn't put quite enough paste on the RSX and Cell dies the last time I had it open, so it's actually a very good thing I re-opened it to apply that gasket stuff you recommended to remake the cell processor supporting seal. This video might have indirectly saved my PS3 from myself.
Original comment follows;
I've actually delided both at this point, some time apart; the Cell a couple years ago and the RSX just yesterday. I may not have done it for the smartest reasons (You specifically call out deliding the RSX not being appropriate for preventative maintenance) but it seemed to go well.
That said, I think I'm going to need to open it up again and apply your recommendation of remaking the silicone seal on the Cell. It has worked fine since my delid, but it's a backward-compatible Fat model and I'd like to give it the best chance of longevity I can.
I considered ordering a brace for the backplates but, curiously, I think the integrated backplates on my model actually already have a protruding ring of metal that sits within the hole in the motherboard around the back of the cell and RSX. I don't know how common this is, I'm in a PAL region that only received the later backwards compatible PS3 models if that makes a difference. I'll double-check the next time I have it apart to re-apply the silicone seal.
You're incredible Felix, keep up!
I genuinely love you showing steps that are irrelevant to most viewers - packaging syringe after syringe of rtv for future use, which I'm sure *you* will get through, but I'm just picturing Joe delidding and relidding his PS3's heatsink after trying it a few times and maybe getting through a couple syringes.
Excellent Video Like Always Felix
Thank you so much for your hard work. I really appreciate the themes of your video here: best practices and risk mitigation! ❤
Have you considered having the bracers machined? Not for helping thermals at all (Doubt you would want them machined from metal anyways), but more so for dimensional accuracy and rigidity at higher temps. Of course it would cost way more than a 5 cent printed part, my guess is at most 10 bucks for a one off. Anyway, Excellent video as always Felix
Eh, I just print them as I need to. It's tedius, but not too bad.
Great research, this will be immensely helpful when delidding mine. I went the route of buying a Slim but it is giving me issues, so will prob take the old one and delid.
I will say that if the IHS is mostly stable, gluing seems unnecessary. Maybe it distributes force better but I don't think it will matter much. I agree it is important to keep it, as without a socket, there will be stress on the solder joints.
I would also not reach the same conclusion on drilling holes. In semiconductor land, 35 vs 45C is irrelevant, especially for VRM chips. If that is the temperature, it will shorten the lifespan by 0.
If it was 95 vs 85, that would be a huge difference. Could be some added stress for capacitors but would need to check if they run close to rated temps or not.
Unfortunately the PS3 will always be a mess to repair. Thankfully RPCS3 exists and CPUs are getting faster. Thank you for all the work your doing to preserve the hardware. Getting some still working will always be important.
I read it's something like for every 10c increase the chance of death doubles. I'd have to look it up for a MOSFET, but 10c more is not good.
Now one could make the argument that replacing 2x VRM modules instead of the RSX is a worthwhile tradeoff, but I would argue that ignores nuance. It reduces the complexity of what kills the RSX to a single factor. Heat! While that scratches a psychological itch for the world to be a simple place, that's not the reality we live in. PS3 is not a place where problems reduce to a singluar cause, with a simple solution. It's always a combination of unknowables and calulated probabilities. Pros vs cons or give and takes.
Let's say the RSX would die in 5 years if you did nothing. So you drill holes and now it lasts 7 years, but the VRM die in 5 years instead of the RSX. Ok, you replace them, which is easier than the RSX! Then the RSX dies 2 years later and it needs replaced anyway. Was it wort it? Or would it have been better to only have 1 thing die and need replaced?
OFC this is a hypothetical, we can't predict. Point is that wishfull thinking always makes the assumptuon that it'll be better, not worse. But it's just as likely to be worse. The fact that we (humans) assume our luck would fall on benificial side is a pshychological bias called gamblers ruin. Or Confidence bias. We overestimate our chances, content in the lie we tell outselflves that "it'll be fine. Felix is making a mountain out of a mole hill."
Until disaster strikes. The astute observer of human behavior will see this pattern repeat like a poorly coded AI NPC. It makes us easy to predict and fool. All a deceiver has to do is figure out what we want to believe, and tickly our ear. But the convient lie is a jedi mind trick for the weak minded.
You are a guru in PS3 internals and this is another masterclass video. I love your content.
I love how much effort was made to make point to glue IHS back on and admit not doing that on own consoles.. as honestly there is no good point to do so from any point of view other than assumptions and guesses why it would be there other than just keeping it still. But give points showing how do gluing well. Though i want say when selecting silicone is best select one which wont be corrosive for electronics aka wont create acetic acid during curing process.
I killed a fully functional 90nm RSX when I attempted a delid (cracked interposer) with the popping off method. It sounds easy in theory but is harder in practice. Don’t use a screwdriver to lever on it, it puts to much pressure on interposer. In my opinion a butterknife is better due to it having more surface area and therefore it puts less pressure on one spot.
Yes, if you use a screw driver, a palate knife undeneath will protect the Interposer. Squeept does it that way and it works fine. You just need to have something thin and stiff enough that it can fit between, but still have enough rigidity to be levered without bending too much. And it needs to get far enough under to lever, not just dig in and scratch the traces.
First couple ones I've had were already delidded since I've bought them from someone in my city, and he's been working on his own delidding tool similar to the one you use, gave me a prototype of it along with a couple dead COK-001 boards (868-22 that had popcorned during prebaking), had very little issues testing on those. I do have a spare B01 that I can try it on after testing more on the other dead board. Never done anything like the case mods or eraser, and all the BCPS3s I've bought came with 226s already, which I suppose is convenient.
It wouldn't make much sense for you to test but some C and G models from PAL regions had the APS-227 which is the same as the APS-226, but only takes 220-240V input. I'd doubt it'd have any real difference in efficiency/operating temperature from the 226, but could be something to test on my own later if I decide to get a C model. Mainly get A and B models since some of the PS2 games I have require the EE to be present as well (and because Japan only got those models and Japan is the cheapest place to get consoles from)
Couldnt you have uploaded this video two days earlier, before I listened to bad advice and cut some traces during the cpu delid?
Well guess I am practicing microsoldering and trace repair next!
Ouch! Good luck with that! Last time I attempted it, they had me questioning life choices. Trace repair is next level microsoldering.
@@ripfelix3020 yeahhhhh its gonna be a shitshow. Luckily i can't make it any worse, so there is at least that.
I love the content we have gotten from RIP Felix's dive into the blackhole of PS3 YLOD, and how to fix them. For me, I just bought a PS2 slim, and a PS3 slim, and have never looked back.
🗣️ This is the future *banger comes in *
MX-4 should never be used when applied to a bare die. I have done this many times with laptop dies and after around 30 days of use, the paste has already been pumped out which leads to the same thermal issues as before the repaste. Instead for direct die applications a paste like MX-6 should be used which is less liquid and more solid. Arctic does not recommend using a tool to spread out MX-6; rather, you should use the pressure from the cooler alone to spread it. MX-4 is fine when used on a large IHS like a desktop CPU or the RSX/CELL.
Even better: use PTM7950 for the bare die.
MX-4 Is non conductive. I’ve used it on several different types of bare die chips for years and for long term. No problems.
@@henUSERNAMEwho said anything about conductivity
pumped out means you used too much (some are expected, but not too much) -- don't forget: thermal paste is only intended to fill the gaps between interfaces (eg cpu and cooler) not to be another interface/medium; adding too much will reduce the thermal transfer
I would just use Liquid Metal.
Hi Felix, I'm really glad this video has been released now. It can at least dispel all of the nonsensical misconseptions and BS advice thrown around in Reddit and such. I do have some things I'd like to know about though
1. I have actually delidded a CECH-E01 PS3 successfully, but opted not to re-glue the IHS since I frankly I was just lazy didn't feel like it (Buying the silicone sealent stuff is extra cost after all) but I actually still had temperature issues. Eventually, I was able to figure out by asking around that HY510 is pretty garbage paste and doesn't work very well on the PS3. But during my troubleshooting, I ended up removing the original sealent in an effort to see if it would make a difference. Oddly enough, I think my temps have actually gotten a bit worse from removing the original sealent. Is it because my CPU's IHS is now sitting on the die with uneven pressure or something? I may have to re-glue the IHS if this is the case, since the original sealent is now lost. The Vejle XCGPU on the Xbox 360 S/E is an example where the sealent is necessary, but it's also a MCM chip with two dies of different heights.
1.1. When re-gluing the IHS, what if you just do four corners instead of sealing the whole thing? I know it's what Octal450 personally suggests on the Xbox 360 S/E since it allows you to remove the IHS in the future.
2. Does heat actually help when cutting the CELL silicon? I see it recommended by people sometimes, but I'm skeptical if it actually helps compared to just using a good cutting tool. Is the CELL sealent even temperature sensitive?
3. If you're gluing the IHSes of the CPU/GPU back on, is it necessary to use clamps like in the video? What if you just assemble the console back to as it was normally? Would the pressure from the heatsink clamps be enough?
if you assemble the IHS will sit flat against the HS, not necessarily the die. You want it flat against the die while the glue cures. I changed to the clams due to this issue.
Follow the same glue pattern as SONY used. I respect Josh, but when it comes to these things I default to the OG design when I'm not absolutely certain there's a good reason to depart from their design.
I don't know if heat helps soften the silicone, since my tool doesnt require any help. I dont want to add heat anyway. I know people say it helps, but I think if you need heat, then your tool is wack.
This is coolest ps3 I have ever seen
I'll never be able to do any of this but fascinated nonetheless. thanks for sharing. Was curious able liquid metal so I'm glad you covered it
I see a rip Felix video. I Insta click and get some snacks.
IDK if anybody has mentioned this, but GD900 is a great paste for the price. I've compared it to my mx6 and couldn't tell a difference in temp, but iirc the cheaper paste did better than mx-6 when it comes to pump-out over time. Right now my 6700xt is running around 5 degrees hotter than when I first pasted it around 6 months ago with mx6. I'm sick of repasting every 6-12 months, so I'm going to try some ptm 7950. People say it's less likely to have issues with pump-out. Anyway, well done with the video.
39:40 i think that is important to consider the temperature of your region too, mine bc ps3 can pass 50% on hot days here in brazil to keep cpu below 65/66 Cº, but in cold days (which are rare) the speed can drop to under 40% without changing anything
I meant with an abient temp of 25C or less.
Yaaay new video. Can’t wait to get home and watch
(Also; first?)
ur not first
Hi man! From seeing your ps3 videos, I understand you know your way into electronics. Even to the point of proving people wrong and finally putting an end on the wrong diagnosing and "fixing" of the ps3. You probably have better things to do than breaking your head with another console, but there's this other issue that seemingly no one knows how to fix on the original xbox, the tiple boot frag, almost every console plagued by this is deemed a loss. This is a rare-ish error where the xbox fails to pass the self checking test and doesn't give any type of indication to the actual failing part. No display no sound no error light or code just red and gren flashing (frag-ing [or Flashing Green And Red] means general system failure, xbox doesn't get to post). I'm losing my mind with my xboxes like you did with your ps3s some time ago, I hope more light gets shined on this. Thanks.
I have several friends deep into OGX and they are on the trail. I'll keep my ear to the ground and at some point I intend to attempt the 1.4GHz CPU upgrade, steller, HD+ mods. I'll dive deep, but I'm still captivated by PS3 and want to complete several ongoing projects with it before moving on to another console.
@@ripfelix3020Exactly, keep going 👌
BABE WAKE UP HE POSTED
Man, the sheer increase in console hardware fragility/complexity between 6th and 7th generation is staggering.
Even something as simple as a routine replacement of thermal paste requires specialized tools and surgeon-like precision to avoid destroying your machine.
I've done lots of amateur work on 6th gen stuff and earlier to save myself a bit of money and learn new things, but after 2005 all the manufacturers really put their foot down.
I'm almost glad that consoles seem to be dying, at least in 20 years when all the Xbox Series consoles start to fail from who-knows-what, nobody will have to scramble to save them to keep their nonexistent exclusives playable.
I dread the day when I can no longer boot up a quick game of Tokyo Jungle.
Another certified banger
This is incredibly informative, even if I'm not brave enough to delid my silver phat PS3. I def need to send it to some one to do that, lol.
For those wondering what kind of tool he was using to delid in the video to me I believe it was a metal nail file because I searched it up on Amazon since there was no link in the description for the tool he used or I’m just an idiot
The IHS will also protect the IC from getting cracked. (Eg, the heatsink bearing down on the chip.)
Recently got the courage to repaste my Phat PS3 (non backwards compatible model). Own the original box and it was running hot even with web man. I recommend it, maybe not delidding unless completely necessary, but for making sure it makes less noise.
I figure if it needs more than 50% fan to maintain 68C underload (90nm phat models) then it's probably time.
@@ripfelix3020 I think its dependant on the user, my PS3 even with the thermal paste replacement (65nm RSX) stays at 35 percent fan speed mostly. It then sometimes when I play intensive games shoots up to 50 percent fan speed; it slows down back to 35 after a few seconds though.(this usual happens every 10 minutes) I don't mind it but I can expect some people might not be too pleased with most consoles being relatively quiet.
Have you tried thermal pads instead? Curious to see a comparison.
Also in your comparison clips here the RSX is always 5-10c lower than the CPU… my RSX is usually around the same temperature as my CPU.. it easily stays below 68c at around 30-35% fan speed but is my RSX too hot then?.. mines a Cech21xxx Slim.
instead of what?
@@ripfelix3020 sorry i meant instead of thermal paste on the cpu and gpu..
There should be a der8auer style delid tool, there is one for the RSX chip, but not the CELL. Which needs it the most. It would be a great way to get away from Razors and Paint Knives which can render your PS3 useless after a slight mistake. And as for applying thermal paste I recommend manually spreading the paste with a spreader tool. From my test my temps were dropped by 10 degrees vs the peasize method.
Those wont work on board. And even if they did, it'd place too much force on the board/interposer for my commfort. The 3D printed one ebay is not good. It's the same as the pop method, except the plastic wedge will get hung up on the edge and not work. And the plastic springs off. It's trash.
@@ripfelix3020Yes I have not tried It. I think the RSX tool is pointless, RSX's are a lot easier to delid. But i can't trust the pressure it will put on the BGA solder underneath with the Cell CPU.
when repasting a bare die like on a Delid i like to make sure to spread the thermalpaste so the entire die is covered before putting the ihs or heatsink back on it
that way i ensure that the entire chip is covered as simply putting on some paste and hoping it spreads on its own is not a 100% guarantee and you *may have a uncovered corner* , wich can lead to a dead chip certain components of the cpu like the memory controller love to sit in those corners and if that part of the chip cooks nothing will bring it back
You can trap a bubble doing that.
@@ripfelix3020 that usually only happens on larger surfaces like the whole IHS and if you have a very uneven application and relatively low pressure
all of wich you dont really have on a PS3 Direct die or GPU's
Thank you for the videos
Felix, can i just pay you to fix my console? This video scared me 😂.
I don't do this as a service. I dumped a bunch of info gathered over the years into this video. It seems like alot, because I expanded it out with all the stuff I wished I knew before trying my 1st time. Things that need said and most don't. No need to be scared. Just onboard what you can and keep watching. Eventually you'll feel comfortable with the process and be ready to try it on a practice board.
@@ripfelix3020 Thank you for that reply. I'm definitely going to keep watching this until I feel comfortable.
Hi.
2nd time around here.
I want to differ in something......
I have an E model,and i did almost all of this.
Only i need delid Cell and new thermal pads.
And i did some extras too.
I used the last ps3 fan,the white label DC brushless from last models (make long cables for to attach) and used thermal grizzly.
And my 226 psu flopped,so i have 231 instead,and i got 62 on Rsx,73 on cell at 40x speed at maximum.
I don't think 231 is the worst one (ZZR with 19 blade fan are truly the worst option)
And that's for me.
Also i did remarry on BD drive,my 410 lens is great.
Maybe one day i can get a frankie mod for this beast.
I am quite sure the APS-231 does run hotter in a BC model. Both the 226 and ZSSR run cooler in this scenario. Though the 231 does work, there's no telling for how long.
About the fans from later models. I tried a G14T S1, it's the quietest fan. But the issue is that it's weak. In order to push the needed volume of air, it needs to run at like 40-60%, which is just as loud and doesn't have the headroom at 100% to push much more air. So if ever the higher percentages are needed, like when past wears out or a really hot day, it can't cool as well as those more powerful fans.
Fun fact, the G14T and D14F have the exact same 15-blade blade. It slips on the rotor. The only difference is the motors power.
Great job,when is episode 2🙏🏻
What would you like to see in a series like this?
How reliable is the 65nm CPU/GPU? I have a PS3 slim(CECH-2004B) I bought in 2019 second-hand CIB with 2 OEM Controllers and to my surprise it was only used for 30 days (Total usage I got to know from webman settings which is readings given by syscon) to this day the total use it has had is 57 days (including the time I played on it). I read somewhere that this Slim model(CECH-2004B) is the "earlier" slim model which is one of the few able to launch full CFW and has 65nm chips that are evidently easier to delidd than later slim/Super-slim revisions. I don't play on it a lot to personal irl reasons but the temps on it are never good. Due to this I repasted it 4-5 times with MX-4 over the last 4-5 years of owning it(because I was always unsatisfied with the repaste application) and finally was satisfied with the repaste method I used last year. To give an example regarding bad temps I booted Last of us for the first time with Syscon fan settings and 5 minutes into the game it easily reached 81C GPU and 78C CPU. I have no idea what to do. I don't even how to soldier much less Delidding. Living in a third world country the best "expert" I could find for the repair of these consoles is an old sony technician that use to work on them and got my HDMI port fixed interesting thing is that he used a LIGHTER of some special kind to loosen the soldier joints at first I was afraid but later he used a small soldiering iron and I got it fixed. Anyway, Is it worth it for me? Or should I stop worrying about temps and just play the slim at 40% fan settings until it starts giving issues like a random reddit stranger suggested?
EDIT: Wow, that's a wall of text. I really hope you'll read this and give your opinion as your opinion regarding this matters a LOT for me.
The 65nm RSX is a reliable chip, no need to worry about it dying from a heat related defect.
I'm not sure what "normal" temps for a 20xx model slim are. I've never paid any attention it on mine. I just left it up to syscon to do it's thing. I don't remember there being any fan noise, it's always been pretty quiet. I should see what it get's up to.
The only thing that model has that tends to die are the NEC/Tokin proadlizers on the RSX side only. But otherwise they are considered reliable.
Most 25xx are jailbreakable, it's the late models that might not be. But it's actually pretty hard to find one. 30xx and later definately aren't.
you know what PTM7950 would probably be a really good solution and when it comes to ceramic thermal glue you freeze the heatsink, you do not heat it, that may work but you reduced the lifespan of the chip. In the field if a board fails ICT or FBT and a chip with a thermal glued heatsink is failing we use a freeze spray to freeze the heatsink and with a light twist or with light pressure from a pry tool it will release cleanly from the chip.
I tried a freze fracture using dry ice. The thermal adhesive becomes stronger and it guarintees the VRAM delaminates. You need heat to soften it if using the leverage method.
24:36 one question tough: why not mount IHS to metal shielding and apply paste into center of upside down IHS? it will still be present there, shielding also has to have right "curvature" (or rather of it) and we're saving ourselves hassle of deliding in the future when paste worns out. It will eventualy but at this time it will be as easy as catefully disassemble the console, wiping our solidified old paste and reapplying new, right?
2 issues. 1 the IHS will sit flat against the HS, not necessarily the die. I've seen that cause bad cooling. 2, most importantly the IHS is a stiffener. It provided structural support to the Package, likely increasing reliability which is the entire purpose of delidding in the first place.
Why replace the paste to lower temps 3C, for an unknown longevity result, then not glue it back on reducing reliability by and unknown ammount?
It's nonsensical until you realize that people are not making rational descisuons, they're making emotional ones. Choosing the bad mods they "want to believe" in, and disregarding good mods they "should." Why? Usually convenence and fear. They fear the extra 3C from a lower thermal conductivity paste, but that lasts longer. And instead choose Liguid metal, which puts the entire console at risk of being destroyed if you don't contain it properly. They don't want to have to put the IHS back on, so they choose not to believe it's necessary, instead coming up with all kinds of rationalizations about how SONY's thermal design teal of engineers were stupid and that dude on reddit knows better...and so on.
It's a pattern of behavior that you'll see in humanity that's never going to change.
At 36:22 , are those 3d printed fan blades for the Nidec fan? I thought about this a couple years ago, and I am curious if you have found any benefit from different blade designs.
Zero. I tried many itterations and every one of them was worse is some way. We came close to matching the performance with one design, but not quite. Clearly Nidec simulated and optimized the blades they used. It was an intersting project tho. Gave me new respect for the engineers.
@@ripfelix3020 Interesting! I wanted to try this but didn't have a printer at the time to experiment. I guess it was time better spent elsewhere then. The Nidec guys clearly know their stuff.
I remember I was playing midnight club los Angeles and I thought to myself that the fans were a bit louder than usual i should have turned it off there
Just curious why use clamps to secure the ihs while the glue dries? Can't you use something else that will distribute the pressure so you don't need to use the 3d printed support plate. Is gravity not enough? Thoughts? I appreciate your videos!
When the paste was originally installed by the packaging facility the chips were not soldered to the motherboard. They could be clamped down with the pressure flat against the die and no worry about anything underneath the interposer. Like the BGA or the MLCCs on the backside of the motherboard.
But we dont have that luxary! Well unless you want to desolder it first!
So if you only use the backplate then now the pressure of the clamp pushing down on the die will press into the center. The interposer will flex under the die. The motherboard will bow under the die. That's because there nothing resisting bowing/sagging underneath the die. Only the 4 standoffs in the corners. When the glue sets the moment you release the clamp, that pent up strain will spring back and be locked in by the glue. And with thermocycling over thousands of cycles there's no telling how that would affect reliability.
To remove that varible, I'm using these to keep the MB flat under the entire BGA region so that I know the paste and glue are perfectly flat while the glue sets. That way it'll be like it was from the packaging facility and I can trust the calculations/relibility simulations are still relevent to the design. I didnt change the a varible by locking in pent up strain.
Felix rocks!!!
I recently delided mine (just the rsx) and clean it properly. But i found an thermal pads (very thin brown) at the nec tokins under the cell and rsx. Is it normal ? From what i searched they mostly been found ouy on the other side... cechc03 model. The console was sealed and little bit dusty inside I assume it was never been opened before just good cared. 129 days in total, 30 errors in total mostly 1001 and only one 1200 from may2024. (bought the console in july.)
Someone added them after the fact. That's not OEM, no. The tokins don't dissipate heat throug their casing. It's just a resin cap, which doesn't even touch the metal part of the cap. It's an air gap. Lol. So putting a pad there does nothing.
The pad under the CPU however is concerning. I've seen that cause BGA damage and cracked dies. People don't think it can. They get indignant about me pointing this fact out, but it can and has. So I do not reccomend it!
lol @ the lil jon impression
Why thank you! I cracked myself up while making that. 😄
One of my favorite Chappelle Show skits
Mine too!
Since you're literally the PS3 god by this point, do you know any tool i could use to recover the whole HDD of my PS3? Something that can read all files off of that filesystem the PS3 uses. I have a Tool and the guy i bought it from only played PES without even installing it on the HDD. (His brother is a game developer and they got rid of the PS3 when the PS4 launched.) Presumably no data was re-written since it was first wiped. (That if the HDD hasn't been replaced with a new one.)
I really want to recover the files and see what beta game i could find on there.
Do you have any idea what kinda recovery tool i should use or if i should send the HDD to a data recovery service?
Unfortunately, the PS3 filesystem is encrypted on a per console basis. You need the EID Root Key (ERK). This can be dumped using CFM tools in recent versions of evilnat. With that you can decrypt a HDD using a tool (I'd have to look up it's name. Ask on psx place, they can direct you to it).
Without the ERK, no. The encryption is solid and would take forever to brute force. The only option is to repair the console and get it to boot.
@@ripfelix3020 Oh, the console is not broken. It boots just fine. It's actually very clean inside too.
I have it on its original firmware, some very early one. I forgot. 2.something...
So i need to install a CFW, evilnat cobra that is. Hopefully it's available for DEX consoles too. Get the key, decript the HDD, then recover everything.
Thanks. 😊
Hey man, I have finally decided to delid my PS3, after deliding my PS3 was having YLOD, so I decided to check syscon error logs, I'm getting error code : A0203010. Any help on how to fix this? I see no scratches on the CELL
You may have flexed the board and caused a CPU ball to break. That 3010 is one of the codes commonly encountered after a failed cpu delid. So if no traces were broken, no nearby SMDs are knocked or missing (check them carefully), then flexing could be the cause. That's why I use the MB brace to keep it from flexing.
Sorry you had this. It's one of the risks of delidding the cell. It's why I dont reccomend it as preventative maintenence unless you actually need to. And it's why I reccomend a proper delidding tool. The painters knife isn't my favorite tool either. I feel the way it inserts places too much force on the board. But others like it and swear by it, so I've acquiesced to their experiance.
hey Felix, how can I clean my HDD properly ? because it is doing some errors sometimes .. I think this happened after I used the app that frees 8% of the HDD data. I use the PS3 in that way (with only 1% left for the HDD) for some time, and after that I restore the 8% of memory back however it is doing some errors from then onwards.
Do you print the bracers every time you do a repair, or do you print a batch ahead of time, sand them, and reduce the overall process time?
I printed various revisions and sand them as I need them. Since I don't do very many consoles I dont really think about efficiency. I just print what I need and sand as I need to. Preferrably seldomly.
im curious if you would ever sell your expertise as a service for people that would love to get Frankemodded PS3s? i am a nerd, but im too scared to wreck the rare E01 that i have, and most Frakenmodded ps3s are near 2x the price of a good A01, also, graphene pads for the IHS might be a good option as they dont expire
I'm not much interested in doing them for a service, no. For fun is one thing. Selling to recoup some of my investment costs too IG at some point. But I'm mainly in it for the science and challenge.
To defend sellers, they are too much work to sell for less than they're going for. I don't think people realize the effort and skill that goes into it. The tools alone are expensive. The education through failure. The risk of offering a warranty. And OFC the ebay fees. There's actually good reason they cost as much as $800. A service for $550 is actually a great deal and less than it was a year or 2 back, thanks to healthy competition in the space. That was one of my goals by making these videos and it seems to be working. They're already cheaper than they were!
But at some point if prices fall too much or people are not willing to spend what it takes to make it worth a modder's time, the frankies will not be available anymore. I happen to believe in free markets and the law of supply and demand. So it costs what it costs, becauae supply is low and demand is high. It's natural and good for those who can make them to profit while they can. If it's too rich for your blood, that's because it's not for those who are buying them. Until they stop buying them and modders have to lower prices to intice new buyers, modders will continue to meet demand at a price that keeps them busy.
My PS3 was delidded by the previous owner and still sits in the 40-50% fan range to keep CELL under 68C.
I've repasted it twice, but the issue still remains. It's not glued on, but I've just accepted it's just gonna be loud at this point lol. At least CELL keeps the RSX in the 50 range anyways
I base my 40-50% reccomendation off an ambient temp of 20-25C.If your room is warmer, or the cabinet it's in gets warmer, then that reccomendation is relative. Also, it only applies to A-C models. The rest have different TDP, fans/Heatsinks, etc. So "normal" is relative to each model too.
I would give the RSX / GPU Delid Tool a try it worked great for me it was my 2nd time to delid and was way better than prying it up to delid
Song name: Max Brhon - This Is The Future
Do you do repairs and service Frankie?
I don't, currently.
"Kill Me' "Help" *cries in 90mn RSK*
Have you looked into thermal paste alternatives like ptm7950 or graphene pads?
I've used graphite thermal pads before. They perform slightly worse than MX4, but they never degrade and are fine. Best, they dissipate heat laterally away from hotspots better than most mateials. Graphene would be better, but to answer your question. No, I wasn't aware you could get graphene pads yet. Might be interesting to try.
I have some PT., but I suspect it's fake. Was cheap on amazon, but I think the genuine stuff is only going to be had from digikey/mouser and be honeywell branded. Last I looked it was much more expensive than the one I found on amazon. I might get some legit stuff and compare them to see. Might make for a good vid. IDK.
I've bought a cech20 and opened it to find it had thermal paste beside the 2 IHS, how can one screw that up, idk.
Put 2 carbonaut pads 38x38 on it, so i don't ever need to replace, and can just dust off.
If I were to buy it again I would get kryosheet, same price and way better, but I didn't delid and don't think ever will, so I assumed it would bottleneck under the IHS, and be no difference.
Felix I have five CECH-A01 models.
I disassembled cleaned and delided both CELL and RSX on all five. (Practiced on three dead boards first).
NONE of them had a backing bracer plate on the back of the CELL.
Should I be putting a bracer plate under the CELL? 😯
I did not glue/attach the IHS back on the CELL I marked the IHS where the old paste was.
I let the area under the CELL open how it was when I opened the console.
Question could I put a thermal pad that is not as thick as the CELL board channel?
So the pad does NOT rise above the channel and does not touch the metal shell motherboard shell?
Basically sits in the motherboard void of the CELL.
Yes Felix I bought an ruler sequenced in MM just to measure the thermal pads of the originals Sony used I am that guy.🤷
I have never replaced parts as the consoles never quit working.
Replacing parts in the instance of replacing the fan or the PSU what the console came with I reinstalled back in the console.
I have a CECH-A01 with the ZSSR PSU still going.
CECH-A01 are my favorite consoles ever.
You are correct all these consoles will eventually quit working.
I want mine to last as long as possible.
Enjoy your channel my PS3 favorite Felix!
Ok, first respect the CPUs "void." It doesn't like forign objects inserted there.
You don't have to do it the way I do. I'm pointing out why I do it my convoluted and overly cautious way. If you want to be extra careful like I do, I'm that guy, then you can follow suit. But no, you can have exelent results doing it other ways too.
Is my way the most proper? Let's just say I've developed this procedure to address "minor issues and potential concerns", not necessarily "must dos or it'll die for sures."
@@ripfelix3020 Thanks, Felix always eager to learn more about the PS3. 👍😃
I have the 20gb B00 model it’s been a tank so far on webman after being reposted an delidded it stays at 60 cpu and 55/58 gpu tad loud at 45/50 fan in games but it’s worth it too keep they 90nm cool
Nice. I like the B model's all black trim. It's peak.
And here i thought i was being clever drilling the holes under the console. So to sumerize, having the fans run fast at all times wont solve the low fan speed hole problem as the components like psu and disc drive are just not getting air?
The syscon bases fan speed off CPU and GPU temps. The hole cools them more, the fans can run at a slower speed to maintain the same temps. The airflow through the console is circumvented and components cooled by that pattern get hotter. Regardless how fast you run the console, the holes will circumvent. So if say before you got 50CFM through the upper compartment at 25% and 300 CFM at 50%, the circumvention will lower whatever the CFM would have been at whatever speed you set it to.
Dramaticly lower CPU/RSX temps are possable at the same %, but what most people are doing is actually setting webman at 68C and the holes only make the fan less noisy. So it's not like the temps are lower, it's just the fan is quieter. That makes them feel better and enjoy the noise reduction. But it comes at an unknown longevity consequence. And I have seen the VRM burn out and need replaced. It's not very easy and requires lots of heat...BUT, "If" it could prevent the 90nm from dying (I'm not convinced it can) one could argue it's a better outcome. I don't like it anyway.
I used a # 4 palette knife which is thin enough and the same width as a nail file. Everything else I did as you have done. Lol I used the same exact tool to remove the silicone 😅
I used to use my fingernail, but I like those plastic spanner things. Whatever they're called.
@ripfelix3020 I think it's called a spudger.
trim removal tools, smartphone opening tools... I find they're actually unhelpul for those things because they're made of such soft plastic, but if they were made of tougher plastic they'd probably scuff up my stuff, so that's a tradeoff I'm happy with
no lie, my superslim sounded like the intro before a repaste
Have you tried using Honeywell's PTM7950 for the IHSes? Then theoretically you won't have to replace it again
Yeah, I bought some but havent tried it yet. On the to do list.
@@ripfelix3020 Nice. Look forward to watching your next video! Love the content!
Is this a reupload? I saw one of your videos in my subscription feed a few weeks ago but by the time I collected my pizza and sat down to watch it it was gone.
This is the intended next video. The one last week was because I mistakely set a different video to public instead of private.
I wonder if any slims require a delid in the future as some are soldered down instead of glue that was previously used on all phat models.
That's a myth. I've seen some that were glued on by idiots and subsequent attempts to delid caused the die to break off. The OP assumed it was soldered on. But everyone I've spoke with that has delided slim and SS models says the never solder the IHS on. I left that door open for a while, since I dont personally work on those models, but at this point I need proof before I'll believe it.
@ In that case, eventually even the slim models will need to be delidded, huh? Can’t say I’d be excited about going that route due to how easy it is to ruin a console by simply mucking up the delidding process, though it’s inevitable that delidding may be required in the future… which I personally don’t look forward to.
A friend of mine is making a deliding tool that will hopefully release soon. Should make it much easier to do.
@ Let us know if a delidding tool for Cell ever does comes to fruition, as having a specialized tool that could make delidding easier (without killing the system) would be a real blessing for us that still hold onto PS3 systems, as we can finally better preserve these systems without needing to take unnecessary risks anymore. I look forward to hearing more about the delidding tool that could finally end PS3 casualties.
Given that I bought a CECHG01 *used* in 2013, played BF3 extensively with it cramped in a cupboard for years with only a single crash, and *it still works* (albeit it only sat in open air since then), I don't know what to think.
I should at least soft mod it finally.
I haven't delid my console at all. It is CECHK and it's running time is 400+ days. I haven't changed termopaste at all. Is it okay to continue that way since it's working just fine, and with the most heavy games it's turning fans no more than 31% even after 3-4 hours of gaming of games like last of us, gta v and la noire.
Then you're golden. No need to fear the potato.
So much trouble deliding it i'd rather pay someone else who knows what he's doing, this is not as near and easy as deliding a pc cpu processor, i love doing things myself and learning in the process but damage a console that was working feels bad
Yeah, delidding this consle has to be done carefully.
What are the odds of 3d printing a motherboard brace? Basically everything now days is designed for 256x256x256 but I wonder if there's some design that could make it a viable option; acquiring a brace for piece of mind just to do one console is kinda a stretch for me.
If not AliExpress will for sure be the go-to.
They aren't easy to get anymore anyway. Last I checked they werent for sale. At least not the ones I got. I know some people have made a pattern and use a sheet metal service to cut them. Like OSHcut. So yeah, you can have them made that way and for a reasonable price.
@@ripfelix3020 Thanks for all you've done and have planned man; I'll look into this once I actually set into my CECHE01, with the help and confidence your guides provide of course.
i dont think pre 80c temps are really of an issue on the ps3 or any actively cooled system for that matter, its not the heat that kills it, its that heat cycling that slowly degrades the components
Partly. The higher the temperature delta, the greater the expansion forces per thermocycle. So it's both. But that's just normal thermocycling and assumes there's not a packaging flaw. Like poor adhesion promotor that the underfill is layed on top of, or poor curing of the underfill. Or inappropriate choice in underfill. Etc.
That's what happened with the 90nm RSX in some phats. 65/40nm GPUs were fixed and are reliable. They can be reasonably expected to perform at normal 75-85c without reliability issues beyond expected.
@@ripfelix3020 thank you for explaining it, i have the 90nm gpu. did a repaste and played through all uncharted games on it, was loud, and it did freeze once but it made it through so far.