Discussion in 'Hardware Mods' started by kadorna2, Apr 15, 2019.
How is that NO ONE did that already and post it on youtube. YOU'RE GENIUS MAN.
Yep, posted about using it sometime last year on another thread. Works like a charm... Think I used 2lb or 3lb nylon line.
...on a slim console, you might want to read carefuly what other people post before jumping to conclusions, this is the kind of behavior that bricks consoles
When you used the fishing line, what was the technique? Did you use a sawing motion or was it more like straight motion (like using a wire cheese cutter)? While cutting the silicone, did you tension the fishing line so that it rides up toward the IHS to prevent it from possibly rubbing on the chip's PCB?
Did you notice any markings on the chip's PCB from the fishing line?
You need to move the fishline sawing left and right, start in a corner (and be patient btw, do it slowly)
And you should prepare a couple of "rings" to tie the fishline to them, to insert your fingers in the rings, otherway if you roll the fishline around your fingers and you start sawing with it repeatedly you know what is going to happen, right ? (your fingers very hurted)
You need to insert the index fingers in the rings... and your thumbs located at the opposite border of the IHS
This way the force you generate with the fishline in a direction is "neutralized" with the force of your thumbs in opposite direction
This prevents cracks in the sodler balls under it
What i dont get is why he mentioned an screwdriver, as far i undertand you are going to have both hands busy while doing this
*Is imposible to scratch the surface with a fishline, this method is very safe in that matter... the only risk is cracking the solder balls
Any force, risk cracking solder balls comen sense right? Screw driver method is for gfx chip not cpu, understand now? Sorry for my stupid explanation. But I respect the attempt to clear things up when people read your genuineness. p.s. I think my balls got cracked......lol
Ohh ok, the screwdriver was for RSX, got it
For the RSX i think the safest method is by using razor blades (several), you need to insert blades in the corners, in between the metal IHS and the memory chips
The memory chips can handle scratches, probably you are not going to scratch them, but incase you do it doesnt matters
What you need to do is inserting blades in the corners... in the first blade there is an small risk of scratches because is inserted in between the metal IHS and the memory chip... but the second blade (of that corner) is inserted in between the IHS and the previous blade (so in the second and next ones is imposible to scratch the memory chip)
You need 6 or 7 razor blades... 3 in a corner and another 3 in the next corner and the IHS is going to "pop" and unsticks entirelly
Incase it doesnt unsticks... you know... are needed moar blades to add more layers to the sadwich
Yes, I have used razor blades also, I think is safer because of what you say about cracked solder balls..hahaha he said balls. Anyway I agree with you on the razor blade method for rsx .great explanation!
TBH i found the wire and razors too much of a pain and u needed to hold the mother board on awkward angles at times for it to just work "ok", especially on the later models. I find making it simple and effective is best method but this will vary depending on user and their capabilities so either methods is not for everyone.
So for me a little heat on the IHS's with my wand just to soften the thermal adhesive or the silicon works wonders that allows u to put less force and stress on either BGAs.
For RSX a thin plastic card like about 0.5mm thick insert straight under the X on the RSX's IHS and with the slight heat u did previously, i use a very blunt rigid butter knife so theres no sharp corners like u do have on a screw driver etc... that can dent and pierce the BGA's substrate and just pry straight up with little to no force, remember the heat is key here.
For Cell i use a modified solder ball spreader that u can get with re-balling kits, its just a small version of a "spatula" pretty much. One end is bent on a 22.5* angle approx with approx 15-20mm long head. I used a dremal tool with a sanding disc to thin down its head from 1mm to 0.3 mm and rounded its sharp edges so u cant dig into the PCB at all. Then simply again with heat on IHS but also on the tip of the tool i use helps greatly to slice through the silicon with ease.
I DID THE LID!
Count me to the club of delidders. I just turned my PS3 slim on, and it works great (had a little heart attack because I forgot to put the hdd back in and everything stayed black).
I seems that I was able to reduce the temperature around 4-5°C on idle in XMB. I think that this is quite good, as I can now turn down the fan and make the PS3 a little quieter.
I probably shouldn't have done it as my temps weren't catastophical ([email protected]°C with 31% fan speed on idle in XMB). But I like to tamper with my stuff. Also, the wife and children were out of town, so...
What is still interesting: RSX is still about 5°C warmer than cell (just like before).
Anyways, happy gamin y'all
An improvement of 4ºC or 5ºC in standby in XMB could be good enought considering the thermal paste was not extermelly bad
Think in that differences as a "performance loss"... lets say you had a performance loss of 5ºC in XMB and you fixed it
But that performance loss increases exponentially with temperature... that performance loss of 5º you was able to meassure in XMB in standby probably was a performance loss of 10ºC while ingame (maybe more?, at this point is hard to know)
Btw, you can try to do a "surface lapping" to the IHS's, there are many tutorials and videos for PC... is the same stuff for PS3
The way i do it is to use a glass on top of a table (because the glasses are perfect plane surfaces), with a sandpaper sticked in it with double side tape (it needs to be a sandpaper with the most fine grain that exists)
Then drop some water on top of the sandpaper, grab the metal piece and start rotating it on top of the sandpaper slowly and SOFTLY
At the end you have a perfect plane metal surface like a mirror
Lol! You always have great ideas! I'll probably leave the PS3 as it is - for the time being. But now, you've put an idea into my head, and the optimization can go on! Nice!
I've dicussed that with some collegues, and I got the tip not to use sand paper but lapping paste. That got me thinking: probably the best would be to lapp the underside of the IHS, since the cpu/gpu dies are very flat to begin with. Otherwise you probably have to lapp the topside of the IHS and the big heatsink with the fan.
Also, how would you go about lapping the cpu heatsink from the underside, as it has a rim going around? Hmm I have to think about that a little more...
The sandpapers "gauge" are standarized with an scale... as far i remember the most fine are named "double cero" (or 00)... and also "water sandpaper" (because you need to drop some waterdrops on it when using it)
Thats the most fine that exist, you need to buy a couple of sheets of it
And additionally you should but 1 of the next gauge (not sure if is named "01")... and other of the next gauge (like "02")
You start with the 02.... then the 01... and finally the 00
And something i forgot to mention... after the sandpapers i continue polishing with a cloth + toothpaste, lol
The toothpaste is a nice/soft abrasive because it have fluor... and the cloth is nice to polish because the small fibers
The annoyance is the cloth is going to break several times, lol... (obviouslly, rubbing a metal piece repeatedly on top of a cloth destroys the cloth)
The way i do it is the same... i "grab" the cloth on top of the glass... then i rub the metal piece on the cloth repeatedlly (and this time with strenght) until the cloth breaks... then i replace the cloth and repeat until it breaks again... etc...
All this work requires time and insistence... most probably you are going to be moving your arm until it hurts
Lapping the PS3 heatsinks is an annoyance, because are double
I mean... you need to do the lapping on 2 different surfaces (are 2 blocks of alluminium), one for CELL and one for RSX
One of them is sticked with glue to the heatsink "body", and the other is not (is a bit loose)
The point is... you need to separate them to do the lapping individually... and this is tricky and hard to explain
Im goign to show you a photo of how is made the heatsink of a CECH-25xx
In this phtoso can be seent he 2 alluminum blocks, the block at botton (with a bump at his right for the copper pipe) is glued to the heatsink body (and is imposible to unglue it, btw)
When you look at it from this perspective you are going to notice the other alluminum block have several small bumps in between the fins.... are working as some kind of "retention bumps"
This block is not glued to the heatsink "body", the only thing that keeps it in place are that "retention bumps"
If you break that retentions you can take the alluminium block out of the heatsink
The "retention bumps" are what i marked in red in this photo, if you break them it allows to separate the heatsinks in 2 parts
After that you can do the lapping in the heatsink alluminium blocks individually
Im using the verb "breaking" because that "retention bumps" are assembled at factory by a hit/press... so is actually aluminium material that has been deformed (is imposible to restore his original geometry)
So the decission is either... break them... or dont break them... but there is no intermediate solution :P
I wonder if we could use this tool to do delid ihs on cell chip. its quite flexible
Separate names with a comma.