has anyone done any case mods for airflow?

kaitlin4599 said:
@sandungas can you clarify what you mean when you said this (and/or if the CELL is going to die prematurelly just because his temperatures are a bit out of control) sorry im easily confused

keep in mind since i have a 3rd gen PHAT i also have the more cooler plastic shroud PSU i guess my main thought process was with the hole at the bottom the CPU & RSX would get a greater amount of airflow allowing the heat they generate to be moved away from them quicker i would of course add a dust filter to the whole and such
Click to expand...
The damaged thermal paste under CELL IHS is acting like a barrier, it doesnt allows the heat to be transfered normally to the heatsink

Im going to use some rought numbers for an example...
Lets say... in optimal circunstances the heat needs 1 second to be transfered in this direction:
CELL DIE ---> thermal paste ---> CELL IHS ---> thermal paste ---> heatsink
But in your PS3 it needs 4 seconds

Webman (and all the other custom apps that displays temperature) takes a temperature sample every 3 seconds. This means there are intervals of 3 seconds where the temperature value you see in the screen is not real (because is not monitored on real time with a precission of miliseconds)

Imagine webman takes a temperature sample right now (in this exact second where you are reading this sentence)
In the next second CELL increases his temperature 4ºC
In the next second CELL reduces his temeprature 4ºC
Then webman takes another sample

The result is the temperature displayed in the screen is the same... but the fact is CELL has been heated up 4ºC over the value displayed in the screen

--------------------
Under normal conditions this innacuracies in between what you see in the screen and the real temperatures doesnt matters much (because we know the real temperature should not be very far away from what we see in the screen)
Anyway... we dont really know how violently the temperatures changes, because is not posible to do it with software
But in a PS3 with cooling problems like yours this effect could be higher

In plain words, the temperaures of your CELL could be moving up and down very fast (lot more faster than in a heatly PS3) and is hard to know
That variations of temperature are bad for the hardware because the materials expands and contracts internally (specially the BGA solders)
 
Last edited:
sandungas said:
The damaged thermal paste under CELL IHS is acting like a barrier, it doesnt allows the heat to be transfered normally to the heatsink

Im going to use some rought numbers for an example...
Lets say... in optimal circunstances the heat needs 1 second to be transfered in this direction:
CELL DIE ---> thermal paste ---> CELL IHS ---> thermal paste ---> heatsink
But in your PS3 it needs 4 seconds

Webman (and all the other custom apps that displays temperature) takes a temperature sample every 3 seconds. This means there are intervals of 3 seconds where the temperature value you see in the screen is not real (because is not monitored on real time with a precission of miliseconds)

Imagine webman takes a temperature sample right now (in this exact second where you are reading this sentence)
In the next second CELL increases his temperature 4ºC
In the next second CELL reduces his temeprature 4ºC
Then webman takes another sample

The result is the temperature displayed in the screen is the same... but the fact is CELL has been heated up 4ºC over the value displayed in the screen

--------------------
Under normal conditions this innacuracies in between what you see in the screen and the real temperatures doesnt matters much (because we know the real temperature should not be very far away from what we see in the screen)
Anyway... we dont really know how violently the temperatures changes, because is not posible to do it with software
But in a PS3 with cooling problems like yours this effect could be higher

In plain words, the temperaures of your CELL could be moving up and down very fast (lot more faster than in a heatly PS3) and is hard to know
That variations of temperature are bad for the hardware because the materials expands and contracts internally (specially the BGA solders)
Click to expand...

@sandungas thanks for the reply i understand now sadly i dont have much choice none of the repair places in town can fix my ps3 ihs issue so its either cut a hole in the case for better more increased airflow or add more mounting pressure to the cpu clamps or cpu ihs so the cpu and rsx make better contact with the heatsinks

heres a video in case you wonder what im talking about let me know what you think

also i ordered new thermal pads for my ps3 think that will help with my heat issue a bit?

also aside from cardboard what else can i use that wont be conductive and or wont break the mobo
 
Last edited:
kaitlin4599 said:
@sandungas thanks for the reply i understand now sadly i dont have much choice none of the repair places in town can fix my ps3 ihs issue so its either cut a hole in the case for better more increased airflow or add more mounting pressure to the cpu clamps or cpu ihs so the cpu and rsx make better contact with the heatsinks

heres a video in case you wonder what im talking about let me know what you think

also i ordered new thermal pads for my ps3 think that will help with my heat issue a bit?

also aside from cardboard what else can i use that wont be conductive and or wont break the mobo
Click to expand...
I would never use vegetals to repair electronic equipment (the cardboard is made of vegetal fibers), the plastic is a lot better material
Also, the superglue emits some "fumes" when is heated that are corrosive, and after lot of days/weeks/months of heating cyles becomes "crackle" like founded sugar

Other that that... yeah, the mod of the video is a good idea, but keep in mind we are dealing with very small sizes
Anyway, the first thing you need to understand is how the force is spreaded in between the different parts
The real responsible of the pressure is the bending of the clamp... and it applyes all the pressure in a single point in his center, ok ?
That point touches with the squared part that can be seen in the video... so the forces are spreaded into a squared geometry, ok ?
But... the contact in between that square and the motherboard is made by 4 small "bumps" made of white plastic, located at the corners of the square

And that not so good technically... the ideal thing would be to maximize the surface in between the square and the motherboard.. thats what the mod of the video tryes to achieve ;)
But for it to work... the thickness of the material added needs to be bigger than the height of the 4 "bumps" at the corners
And is better to do it with some flexible material (rubber) because it acts like an spring, you know, the material is compressed when the console is assembled, but is constantly "pushing" trying to recover his original dimmensions
You could do it too with thermal pads instead or rubber
Also, keep in mind the CELL have a "hole" in the center... this is why some people "fills" the hole with a block of rubber

Anyway, the bending of the clamps is even more important than this (and is very easy to do, just dont bend them too much), that bending works also like an spring

----------------------------
The thermal pads of other components (different than CELL and RSX) are not going to help with your problem, is good to replace them for a manteinance though

Personally, what i would do if i was you... is to repeat the "idle test" we was discussing in the other thread
The point is we know the PS3 have a problem, but we dont know how bad is it
Is better to repeat that "idle tests" to have a better understanding of the problem, that tests are safe
 
sandungas said:
I would never use vegetals to repair electronic equipment (the cardboard is made of vegetal fibers), the plastic is a lot better material
Also, the superglue emits some "fumes" when is heated that are corrosive, and after lot of days/weeks/months of heating cyles becomes "crackle" like founded sugar

Other that that... yeah, the mod of the video is a good idea, but keep in mind we are dealing with very small sizes
Anyway, the first thing you need to understand is how the force is spreaded in between the different parts
The real responsible of the pressure is the bending of the clamp... and it applyes all the pressure in a single point in his center, ok ?
That point touches with the squared part that can be seen in the video... so the forces are spreaded into a squared geometry, ok ?
But... the contact in between that square and the motherboard is made by 4 small "bumps" made of white plastic, located at the corners of the square

And that not so good technically... the ideal thing would be to maximize the surface in between the square and the motherboard.. thats what the mod of the video tryes to achieve ;)
But for it to work... the thickness of the material added needs to be bigger than the height of the 4 "bumps" at the corners
And is better to do it with some flexible material (rubber) because it acts like an spring, you know, the material is compressed when the console is assembled, but is constantly "pushing" trying to recover his original dimmensions
You could do it too with thermal pads instead or rubber
Also, keep in mind the CELL have a "hole" in the center... this is why some people "fills" the hole with a block of rubber

Anyway, the bending of the clamps is even more important than this (and is very easy to do, just dont bend them too much), that bending works also like an spring

----------------------------
The thermal pads of other components (different than CELL and RSX) are not going to help with your problem, is good to replace them for a manteinance though

Personally, what i would do if i was you... is to repeat the "idle test" we was discussing in the other thread
The point is we know the PS3 have a problem, but we dont know how bad is it
Is better to repeat that "idle tests" to have a better understanding of the problem, that tests are safe
Click to expand...

@sandungas well i did the idle test using the syscon option in webman so that the PS3 had control of the fanagain and after 15 mins temps jump by 2-4 degrees on the cpu all the way up to 71 before i shut her down they kept climbing and never leveled out so i got scared she would overheat and shut her down

so heres where im at please hear me out i already have a PS3 slim the only reason i bought this phat was because it was 50$ which was too good a deal to pass up so the way i look at it is if in the end she overheats and dies on me im only out 50$ minus the cost of the thermal paste and thermal pads

so i think im just gonna do the fan mod i talked about and call it a day because with the thermal paste replacement and the thermal pad replacement the FAN MOD should be able to allow the fan to pull in more air

now i know this isnt the course you wanted me to take but i have no way to get the IHS delid and i dont know how to increase the mounting pressure i have no plastic no 3d printer to make a custom pressure peace and bending the clips will only do so much so in the end i want to thank you for all your help if there is anything you feel you need or want to add or comment on please feel free to do so thank you for your time and help
 
@sandungas i think i found something interesting i set webman to a fan speed of 40c jumped into irisman game manager and let her idle for 35 mins temps so far 20 mins have passed and temps for the CPU seem to have leveled out at around 57-58c and for the RSX 52c granted i have my window ac unit on granted its only 70F outside whatcha think i will run this same test for the same amount of time later in the day when its much hotter outside and report back will also try to test out GTA for at least half an hour if i have the time

i know i should have turned webman off but if i had done that she would have hit 80c under idle in less than 10 mins
 
No case mods are ever needed for PS3. The exhaust system is very efficient as long as all the components work. Unfortunately 90% of the time, the paste under the IHS for cell and rsx is compromised and needs to be replaced. With webman and fresh paste, the system works perfect and removes heat just fineZ
 
raidriar said:
No case mods are ever needed for PS3. The exhaust system is very efficient as long as all the components work. Unfortunately 90% of the time, the paste under the IHS for cell and rsx is compromised and needs to be replaced. With webman and fresh paste, the system works perfect and removes heat just fineZ
Click to expand...

@raidriar sadly none of the repair shops in town can do the delid and i dont have the tools nor the skills for such a risky job a case mod is my only option
 
kaitlin4599 said:
@sandungas well i did the idle test using the syscon option in webman so that the PS3 had control of the fanagain and after 15 mins temps jump by 2-4 degrees on the cpu all the way up to 71 before i shut her down they kept climbing and never leveled out so i got scared she would overheat and shut her down
Click to expand...
The test is safe because you can stop it at any point if you see the temperature is excesive, thats what you did, is ok... but you are not telling us the difference of temperature in between CELL and RSX when you was doing it !
The only reference i have to know how bad is your PS3 is the sentence you wrote when you said you had 18ºC of difference... but this is the kind of test that needs to be repeated 2 or 3 times to be completly sure the 18ºC you got was an accurate meassurement

Lets say... im a doctor, your PS3 is the patient, and you are a familiar of the patient
You are asking me what to do with the patient, and im telling you that i cant give you an accurate answer because i dont have enought info from the patient

You want to put the patient into surgery without the proper tests, is ok, but i dont know the results, and eventually im not going to be able to give you an straightforwrd answer to your questions

kaitlin4599 said:
@sandungas i think i found something interesting i set webman to a fan speed of 40c jumped into irisman game manager and let her idle for 35 mins temps so far 20 mins have passed and temps for the CPU seem to have leveled out at around 57-58c and for the RSX 52c granted i have my window ac unit on granted its only 70F outside whatcha think i will run this same test for the same amount of time later in the day when its much hotter outside and report back will also try to test out GTA for at least half an hour if i have the time

i know i should have turned webman off but if i had done that she would have hit 80c under idle in less than 10 mins
Click to expand...
The "idle test" with custom speeds is not much representative, keep in mind when you are ingame that difference of temperatures of 18ºC you was having is going to increase exponentially
In the idle test you got 18ºC because inside a homebrew app there is a low workload, but inside a game, when RSX starts kicking the problem is going to look worst

By now i dont suggest you to make any temperature tests ingame btw, is a lot more risky than the "idle tests" i was mentioning
 
Just to be clear what i was asking you to do in the idle test

1) Power ON the PS3, and prepare everything you are going to use later, configure fan speeds to syscon mode
2) Power OFF the PS3 and wait several hours to allow all the PS3 internal parts to return to ambient (more than 3 hours)
3) Power ON the PS3, enter the homebrew you are using to display temperatures/speed and keep an eye at the values
4) Take note of your results of CELL/RSX/FAN in a .txt (or use pencil + paper)

You can use this layout btw, fill the info in it
FAN | CELL | RSX
20% | AMBI | AMBI
21% | 74ºC | ??
22% | 75ºC | ??
23% | 76ºC | ??
24% | 77ºC | ??
25% | 78ºC | ??
Click to expand...

You need to take note of the values in the exact moment where the fan speed changes
I know the fan speeds are going to be that ones exactly because it can be seen in the graph i made with the factory thermal config of your PS3 model
vIDCyqQ.png

The other reason why i prepared the layout for you to fill it with info that way, is because in your PS3 CELL is a lot hotter than RSX, so the responsible of the fan speed changes is always CELL (not RSX)
In other words... everytime your fan speed changes, is because CELL have reached the values named "TempU", it works this way:
If CELL = 74ºC, then FAN = 21%
If CELL = 75ºC, then FAN = 22%
If CELL = 76ºC, then FAN = 23%
Etc...

One of the problems of the factory thermal settings of your PS3 model is when you boot the PS3 it does a "warmup" keeping the FAN at 20% speed for too much time
In other words... the fan speed only increases when CELL reachs 74ºC

If you stop the idle test with CELL having a temperature lower than 74ºC we are not going to be able to see what happens in the other fan speeds (and you are not going to be able to fill the info in the layout i made completly), but thats not much important, im just suggesting to continue the idle test up to 78ºC because it should not be dangerous
Just think in it, there is maaaaany people in OFW using the PS3 with temperatures over 80ºC and the PS3 doesnt dies so easilly

Anyway, if you dont feel confident stop the test at 75ºC or so... this way we have at least 1 sample of the first time the fan speed changes

The most important thing for me is to check again that difference of 18ºC you mentioned
kaitlin4599 said:
...with webman set to syscon the cpu hit 82c after 12 mins on idle in irisman and the RSX hit about 64c
Click to expand...
 
Last edited:
sandungas said:
Just to be clear what i was asking you to do in the idle test

1) Power ON the PS3, and prepare everything you are going to use later, configure fan speeds to syscon mode
2) Power OFF the PS3 and wait several hours to allow all the PS3 internal parts to return to ambient (more than 3 hours)
3) Power ON the PS3, enter the homebrew you are using to display temperatures/speed and keep an eye at the values
4) Take note of your results of CELL/RSX/FAN in a .txt (or use pencil + paper)

You can use this layout btw, fill the info in it


You need to take note of the values in the exact moment where the fan speed changes
I know the fan speeds are going to be that ones exactly because it can be seen in the graph i made with the factory thermal config of your PS3 model
vIDCyqQ.png

The other reason why i prepared the layout for you to fill it with info that way, is because in your PS3 CELL is a lot hotter than RSX, so the responsible of the fan speed changes is always CELL (not RSX)
In other words... everytime your fan speed changes, is because CELL have reached the values named "TempU", it works this way:
If CELL = 74ºC, then FAN = 21%
If CELL = 75ºC, then FAN = 22%
If CELL = 76ºC, then FAN = 23%
Etc...

One of the problems of the factory thermal settings of your PS3 model is when you boot the PS3 it does a "warmup" keeping the FAN at 20% speed for too much time
In other words... the fan speed only increases when CELL reachs 74ºC

If you stop the idle test with CELL having a temperature lower than 74ºC we are not going to be able to see what happens in the other fan speeds (and you are not going to be able to fill the info in the layout i made completly), but thats not much important, im just suggesting to continue the idle test up to 78ºC because it should not be dangerous
Just think in it, there is maaaaany people in OFW using the PS3 with temperatures over 80ºC and the PS3 doesnt dies so easilly

Anyway, if you dont feel confident stop the test at 75ºC or so... this way we have at least 1 sample of the first time the fan speed changes

The most important thing for me is to check again that difference of 18ºC you mentioned
Click to expand...


@sandungas heres the results ran the test 3 times waiting about 3 hours between each test and ran each test for 35 mins

FAN | CELL | RSX
20% | AMBI | AMBI
21% | 74ºC | ?? 63c
22% | 75ºC | ?? 64c
23% | 76ºC | ?? 65c
24% | 77ºC | ?? 66c
25% | 78ºC | ?? 67c

after each test i would power down and let it sit to cool off then rinse and repeat

oddly enough i noticed something that may or may not help you at the end of each test the final recorded temp was 79c for the CPU and 67c for the RSX i saw this in all 3 tests so naturally thinking this was a glitch or a fluke i re ran the exact same 3 tests under a different different homebrew application just incase irisman was reporting the wrong temp and got the same results in multiman after each 35 min test temps were the same including the end result of 79c for the cpu and 67 for the rsx as i type this message im letting the system cool down im gonna run one final test where i let it idle in irisman for about an hour and i will keep an eye on temps i will not shut it down unless temps reach 88c i shall keep you posted thanks for your time

keep in mind i ran these test at 1am in the morning with my windows ac unit's temp set to 60 which is as low as it can go
 
Nice tests @kaitlin4599 :encouragement:
First thing that worths to be mentioned is that there is no need to repeat this idle tests again
What i was trying to see is how much "far away" is RSX from CELL, your post from the other day when you said 18ºC of difference scared me a bit, but now you are having 11ºC and thats not so bad
And we are completly sure that 11ºC was meassured accuratelly because you repeated the test several times, so is fine... end of this chapter. Sorry to drive you into technicisms :)

Now back to what you was asking... are you going to do all this mods at the same time ?
-Hole
-Increase bending of CELL clamp (or both, CELL and RSX)
-Rubber/eraser under the clamps

Im asking because for everyone of them you are supposed to increase the cooling performance a bit, the better way to do it would be by dissassembling and assembling the PS3 several times (and use the idle test for reference to compare your new values with the ones you posted in the previous post)
But... this implyes replacing the thermal paste several times and it will take you some time
 
By now you could start by increasing the bending of the clamps, is very easy and doesnt requires reapplying thermal paste
But you should do it with a bit of precission, i was talking about it in this post
https://www.psx-place.com/threads/t...out-removing-the-ihs.23066/page-5#post-235675
And i even made a drawing (the clamp is the black line)
Dc9wk4n.png


1) Use an screwdriver to remove the clamp
2) Place the clamp horizontally on top of a table, like in the drawing
3) Meassure the distance in between the table surface and the "tip" of the bending
4) Thats the "factory geometry"... is good if you use a permanent market to take note of it in the clamp
5) Grab the clamp with your hands and bend it a bit
6) Place the clamp again on top of the table, meassure it again and compare it with the "factory geometry"

The goal is to increase the bending only a bit (lets say 1 or 2 milimeters only)
And the reason why im suggesting to take note of the "factory geometry" with a permanent marker in the clamp is because you know... maybe it happens that today you decide to bend it +2mm .... and some days later you decide to try with +1mm .... or +2mm more ... but eventually you realize that it was not needed to bend it so much so you will want to "unbend" it -3mm... and so on
Incase that happens (and you dont remember the "factory geometry") you are going to be lost... without a reference to know if what you are doing is excesive

After doing this mod you could repeat the "idle test" (to compare with your previous results before this mod) to have an accurate result of the improvement
 
sandungas said:
Nice tests @kaitlin4599 :encouragement:
First thing that worths to be mentioned is that there is no need to repeat this idle tests again
What i was trying to see is how much "far away" is RSX from CELL, your post from the other day when you said 18ºC of difference scared me a bit, but now you are having 11ºC and thats not so bad
And we are completly sure that 11ºC was meassured accuratelly because you repeated the test several times, so is fine... end of this chapter. Sorry to drive you into technicisms :)

Now back to what you was asking... are you going to do all this mods at the same time ?
-Hole
-Increase bending of CELL clamp (or both, CELL and RSX)
-Rubber/eraser under the clamps

Im asking because for everyone of them you are supposed to increase the cooling performance a bit, the better way to do it would be by dissassembling and assembling the PS3 several times (and use the idle test for reference to compare your new values with the ones you posted in the previous post)
But... this implyes replacing the thermal paste several times and it will take you some time
Click to expand...


@sandungas i was only planning on replacing the thermal paste and cutting the hole i wasnt planning on doing the eraser mod since several forums mention that the eraser mod isnt a permanent solution fix so i await your reply but in the end i only plan to replace the paste and cut the hole nothing else

also idk how to Increase bending of CELL clamp (or both, CELL and RSX) so if you have a guide for that i will read up on it to determine if i feel comfy with that mod

one final thing you should know after letting the PS3 cool down i let her idle in irisman with webman still set to syscon after about 25-35 mins in cant remember exactly but after that time frame she started to increase from 79c-80c before i got nervous and shut her down

i had planned to do all the mods at once since i dont really want to have to keep taking her apart not to mention im low on thermal paste and cant afford to buy a second tube atm

i await your reply ty for your time
 
The mod with the eraser/rubber is discussed in the thread i mentioned
https://www.psx-place.com/threads/t...l-dissipation-without-removing-the-ihs.23066/
Is not a permanent solution because you are going to have the thermal paste under CELL IHS in bad conditions... but other than that is a nice idea

Remember the 4 "bumps" of the plastic square piece i mentioned in a previous posts ?, there is a "hole" in the motherboard in the center of CELL
In the factory design we are transfering the pressure this way:
clamp--->square with the 4 bumps--->motherboard---> BGA solders--->CELL--->thermal paste--->IHS--->thermal paste--->heatsink
And with the rubber this way:
clamp--->square with the 4 bumps--->rubber--->CELL--->thermal paste--->IHS--->thermal paste--->heatsink

As you can see we are "bypassing" the motherboard and the BGA solders...as a result the pressure is applyed more directly to CELL, compressing like a sandwich the thermal paste under IHS

Remember, the rubber acts like an spring... it needs to be a material that allows to be compressed, but is constantly trying to recover his original geometry... in some way is like we are adding an spring in that "hole"

---------------------------
The same concept can be made with RSX... for RSX is easyer because there is no "hole" under it
This mod requires to "cut" that block of rubber with lot of precission, and keeping in mind that the rubber material is going to be compressed, there are some reference sizes mentioned in that thread

Not sure if suggesting you to do this before the case hole... both things could give you good results so you decide
This mod with the ruber depends a lot of the kind of materials you could find around your house "for free", probably you have several things you could use
If you are lucky to find something easy to cut in small blocks with a razor blade, and the amount of material allows you to repeat it several times, then go for it... is just a matter of cutting that blocks several times, the first one probably is not going to fit well, but after 2 or 3 you will make a good one :)
 
Last edited:
kaitlin4599 said:
@sandungas assuming i decide to do the clamp mod do i bend it inward toward the table or outward?
Click to expand...
Outward, you need to increase that distance
Lets say... for the first test try "factory distance" + 2 milimeters
Then try the idle test a bit (there is no need to do it completly, just take a look at the values around 70ºC and stop there)

Incase the improvement is not good enought... bend it a bit more
Second test try "factory distance" + 4 milimeters
And try the idle test again, this time the improvement should be notable (but we dont know how much)


Incase the improvement is not much notable i dont suggest you to bend the clamp much more, an excesive bending could be dangerous (and honestly i dont know where is the limit)
It would be time to go to plan-b (the case hole or the rubber)
 
You must log in or register to reply here.

Similar threads

Back
Top