PS3 slim fan speed

[dmr]

and fan speed? You forgot again?

 

[Ali888]

and fan speed? You forgot again?

No
Just forgot to write that

Fan speed was between 27% ~ 29%

 

[sandungas]

Yes all with syscon control. Someone in this thread said that the RSX chip is problematic one and that cell can take temperatures better? The 2504b model that I had before this had always hotter cpu then rsx, so I had situations in game like 72C CPU and 69C RSX is that better situation than opposite RSX at 73 and CPU 69C?

I think the problem was in your other PS3, it had an small problem in CELL, but i guess it was not much important
Most of the times you are not going to know if the problem is in the thermal paste over IHS or under IHS

If the problem was the thermal paste under IHS... and knowing the difference of degrees was not much important... is the kind of thing that is better to dont touch it

In the PS3 you have now i think CELL and RSX are very well, maybe RSX is having a small problem but is not important, being a couple of degrees out of the optimal is normal (because the paste under IHS cant be changed easilly, and is not posible to check it)

Play 20 minutes with SYSCON enabled and result looks similar to yours...

- Starts with 65°C CPU, 68°C RSX
- 15 minute: 68°C CPU, 71°C RSX
- 20 minute: 69°C CPU, 71°C RSX

Fan speed was between 27% ~ 29%

This time when you entered in the game the PS3 was well warmed up (around 65ºC). This is what most games are going to have in the main screen of the game... i mean, even before start playing

Something interesting to try also... (for the record)
The places inside a game where there is a lower temperature uses to be the small rooms (the smallest the better), you know, places where you are surrounded by very few polygons and textures
Like in survival horror games (resident evils, silent hill, the evil within), where the savepoints are an small room, if you stay quiet there for some minutes the temperatures gets very close to the ones you have before starting the game in the main menu

Your temperatures and speeds seems to be pretty similar to the ones dmr had (just a couple of degrees more)... yours is a CECH25xx too ?


-------------
Last btw... if you both wants to make a good comparison you should do exactly the same in the same game
The stress test with GTAV is good because is easy to replicate

In the last of us... the (huge) thermal peaks happens in the areas of the game where is water, there are only a couple
One is located at middle of the game (a lake in the niddle of the street of a destroyed city)... and a underground water channel at the ending

 

[Ali888]

yours is a CECH25xx too?

Yeah, mine is CECH-2501B

 

[sandungas]

Yeah, mine is CECH-2501B

Your PS3 have the thermal pastes very well, and all surfaces are touching each others pretty well too
This kind of tests needs to be repeated few times thought to have a better overview because there is always small innacuracies with the meassures, or at the time you took note of them you did not fast enought, or because the PS3 is not completly warmed up, etc...

Copypasted here for fast comparison,both playing uncharted 2

- 20 minute: 69°C CPU, 71°C RSX
Fan speed was between 27% ~ 29%

1h05m CPU69 GPU73 Fan29%

@dmr has an small deviation of 2ºC in RSX over @Ali888

This initially seems to indicate that there is a small problem in dmr RSX... or ali888 needs to repeat the test by playing in other areas more stressing of the game (in other words, ali888 have a small problem in CELL)

But anyway... you both should not worry much about this, because the deviations are very small... we are talking about couple of degrees only

The fact that temperatures in between CELL and RSX are very close to the other is a good signal that overall both PS3 are working fine

 

[Ali888]

@dmr has an small deviation of 2ºC in RSX over @Ali888

I think i know the reason of that 2 degrees...

@dmr Playing non-stop for about 1 hour, but I played 30 minutes first with Dynamic fan Speed, after that; quit the game and returned to XMB to set Fan control to SYSCON this time. Maybe that's why dmr has warmer RSX.

ali888 have a small problem in CELL

Can it be related with where i live ?
It's 33°C right here...

both PS3 are working fine

Good to hear that
Thanks

 

[sandungas]

I think i know the reason of that 2 degrees...

@dmr Playing non-stop for about 1 hour, but I played 30 minutes first with Dynamic fan Speed, after that; quit the game and returned to XMB to set Fan control to SYSCON this time. Maybe that's why dmr has warmer RSX.

Hmm, it could be what caused some inestabilities in the temperatures

Inside games/apps usually the minimal temperature is very stable, i mean... there is a temperature (different for every game) that is the minimal limit and temperature is not going to decrease under this value (not even in pause screen, neither in the main menu of the game)

For the game you was talking about (uncharted 2), it seems that minimal is around 65ºC (for a CECH25xx PS3 model using SYSCON factory fan speed settings)
And the max you both are having is around 73ºC
So for uncharted 2... the range of temperatures "ingame" is 65-73 (only 8ºC)

So is very easy to have small inaccuracies, because the range is small
To make a better comparison you both need to do EXACTLY the same, with the same times, and in the same point of the game, lol

With the small tests you both made is enought to say both PS3 are ok
But if you want more accuracy (and want to know which one wins the "hottest" competition) i think both you need to find an agreement (of what to do exactly) and repeat the tests

Can it be related with where i live ?
It's 33°C right here...

The comparison we was doing was mostly to see the deviations

The ambient temperature count a lot, of course
If you make the tests in winter... then some months later in summer the results of your tests are not valid

All this im talking about is with the factory fan speed settings (in the "syscon" mode)
But in webman if you configure a maximun temperature it will be valid for all the year

 

[dmr]

Ok thanks for this. What will be best method to put paste on cell and rsx? I change paste couple days ago and use pea method, just put paste in the center of both chips and let coller do rest. Is that ok?

 

[sandungas]

Ok thanks for this. What will be best method to put paste on cell and rsx? I change paste couple days ago and use pea method, just put paste in the center of both chips and let coller do rest. Is that ok?

if you are decided to change thermal paste again keep an eye at the corners of the IHS surfaces
The pea method makes the paste to spread like a circle, but the IHS is an square, so sometimes the corners are not covered

To avoid that what i do is first to extend a veeeeery thin layer of paste with a transparent plastic sheet (like a credit card but transparent, this way i can see under it)
And on top of the thin layer, a pea in the center (llike you did)

The point is to dont have corners without paste... for every corner without paste you are losing an small percentage of thermal dissipation efficiency... lets say dunno 2% each corner or more

-----------
Additionally, if you want to make some additional picky tests with paste, there is a trick i imagined that works nice to see how the surfaces touches each other, the only contra is you are going to waste some thermal paste, but is worthy to do it imo, at least one time

1) Apply the paste, then assemble all other parts, with the bolts thick (specially the black big bolts that keeps the heatsink metal block in place)
2) Dissassemble it completly again
3) Check the thermal paste over both IHS

If the paste surface has like "mountain peaks"... is because was touching the other surface (good)
If the paste doesnt have "mountain peaks" (specially in the corners), is because was not touching (bad)

You can repeat this several times, spreading the paste in different ways, adding/cleaning paste from previous tests, etc... no worry about the paste. But remember... when you finish this experiments clean all the paste and apply new one !!!
In the last step (when you are doing the final assembly of the PS3) is needed to apply new paste to avoid having small "bubbles"

 

[Naked_Snake1995]

isn't harmful for Fan itself ?

Nah, but ill be loud, but you will get use to it, ive set my 60Gb to 40%, but ive the PS3 have been keeping tabs to near 66º CELL and 60º RSX, and i didnt delid it, and i am getting paranoid.

 

[dmr]

if you are decided to change thermal paste again keep an eye at the corners of the IHS surfaces
The pea method makes the paste to spread like a circle, but the IHS is an square, so sometimes the corners are not covered

To avoid that what i do is first to extend a veeeeery thin layer of paste with a transparent plastic sheet (like a credit card but transparent, this way i can see under it)
And on top of the thin layer, a pea in the center (llike you did)

The point is to dont have corners without paste... for every corner without paste you are losing an small percentage of thermal dissipation efficiency... lets say dunno 2% each corner or more

Nice I did it like this and now temperatures are much better. After half hour in uncharted 2 CPU 65C RSX 68 Fan 28%. When I exit to XMB both temperatures lowered at 59C Thanks for this tip!!!

After game I left PS3 in xmb and temperatures lowered at 59/59 but fan didn't lower speed from 28% for 40 minutes. How long should I wait for syscon to lower fan speed?

 

[Naked_Snake1995]

Nice I did it like this and now temperatures are much better. After half hour in uncharted 2 CPU 65C RSX 68 Fan 28%. When I exit to XMB both temperatures lowered at 59C Thanks for this tip!!!

After game I left PS3 in xmb and temperatures lowered at 59/59 but fan didn't lower speed from 28% for 40 minutes. How long should I wait for syscon to lower fan speed?

Syscon will never lower the fan speed, is PS3 uses a Step fan process, Sony learned from that and the PS4 has a steepless fan mode.

Sent from my F8331 using Tapatalk

 

[dmr]

Syscon will never lower the fan speed, is PS3 uses a Step fan process, Sony learned from that and the PS4 has a steepless fan mode.

Sent from my F8331 using Tapatalk

To decrease speeds (when PS3 is reducing temperatures) it works in a different way though, it uses a timer to delay the steps backs

This??? And the motherboard inside this 2504a is JSD-001 if it is important?

 

[pinky]

I wouldn't worry about temps too much. your temps (70s while gaming?) are fine. 80+ is the danger area. I remember @Joonie showing a picture of the temps of his system while gaming. they were in the 80s, and he said he never had a problem with the system. since he's a member of team rebug, you've gotta imagine that he's had his ps3 turned on for extended periods of time.

 

[dmr]

Yes, now is good, but rsx temp was little off so it bothered me a little But this method of putting thermal paste lowered temperatures to cell also. Before I had 69/73 fan 29% and now 65/68 fan 28%

 

[pinky]

that's something I don't like about the state of gaming -- having to worry about whether or not the system is cool enough. I've heard of people boiling eggs on a ps4 pro. that's just not right.

 

[sandungas]

Nice I did it like this and now temperatures are much better. After half hour in uncharted 2 CPU 65C RSX 68 Fan 28%. When I exit to XMB both temperatures lowered at 59C Thanks for this tip!!!

Cool (literally), additionally to the overal reduction of temperatures in both CELL and RSX, it seems the gap in between them has been reduced a bit
So it seems the small problem in the RSX i was mentioning before of around 2ºC has been fixed, this is a good signal because it means the paste under RSX IHS (only visible if you delid it) is ok
That paste under IHS from factory probably is low quality though, and is several years old. But the point is both pastes under CELL and RSX IHS are working equally and thats good... in other words... both pastes under both IHS has been degraded equally along this years

After game I left PS3 in xmb and temperatures lowered at 59/59 but fan didn't lower speed from 28% for 40 minutes. How long should I wait for syscon to lower fan speed?

This??? And the motherboard inside this 2504a is JSD-001 if it is important?

The delays are several minutes long, and is made on purpose as a preventive meassure to cover all the posible "heat peaks" that happens "ingame"
The motherboard model is important, but to be more accurate is better to think in the syscon model, this needs to be explained with an example while looking at the table in wiki
http://www.psdevwiki.com/ps3/Thermal#Fan_settings_per_PS3_model_based_on_syscon_data_reads

There are 5 syscon chips series: CXR713, CXR714, SW, SW2, and SW3
Additionally we need to consider that most probably it allows to configure the fan speeds with settings, so it can happen that 2 different syscon chips could use the same fan settings... or the same syscon chip using different settings for different motherboards
But all this is unkown, the table in wiki did not have much success, we only have a couple of examples from a couple of PS3 models... but we dont know the values used by the other PS3 models :/

For the delays i need to write an example, im going to try to use realistic values, but im telling it from memory (tested lot of time ago) so maybe is not going to be much accurate, but is just for you to get the concept
If someone is curious and make some tests to measure this delays please post it here

---------------
Lets say... you are inside a game at 70ºC (CECH25xx using the factory fan settings in "syscon" mode)
This puts you into "step 9" of the table with fan at 0x4A speed

Then you quit the game and meassure temperature in XMB
Most probably the temperature has decreased several degrees, but the fan is still spinning at 0x4A
This delay is very long (around 10 or 15 minutes ?) because you are returning from a high temperature (the 70ºC you had ingame)
The fan speed is kept at the same speed for so many time because syscon doesnt really knows if you are "ingame" or in the main XMB
Incase you are "ingame", most probably the temperature is going to increase soon to 70ºC again (because frequent "heat peaks" that happens normally ingame). This is the "preventive" effect i mentioned before, they keeps the fan spinning to "match" with the temperature that was generated by the latest "heat peak" (because probability... if there was a heat peak at 70ºC couple of minutes ago is very probable of another one to happen now)
Anyway... this delays when you are returning from a high temperature are the longests

After 15 minutes in XMB you will see how it returns to "step 8" in the table... with fan speed at 0x48

If you wait 10 minutes more (note im reducing the delays) you return to "step 7" of the table... with fan speed at 0x45

And so on... 5 minutes more and it returns to "step 6"... with fan at 0x43

--------------
In short... the delays has 2 goals:
-predict the "heat peaks"
-reduce noise levels, by reducing the number of fan speed changes

The noise reduction is hard to eplain, but by experimenting it you will realize soon, the humans are specially sensitive to the noise level changes, but if you keep the same noise level at some point your brain starts ignoring it
So what they wanted to achieve is to reduce the noise annoyances by reducing the number of speed changes that happens while you are playing

Lets return with the example before of uncharted 2 running in a range of temperature ingame in between 65-73... if you keep the same fan speed along all this range the player will not feel much annoyances
But if you change the fan speed up and down lot of times the player will realize

---------------
Now im writing this im not sure if the delay changes happens in the other way around (lol), i mean... shorter when hot, and longer when returning to ambient

One thing i remember well.... is i could not see it returning to the "step 1" with fan speed at 0x33 ... even after waiting for dunno... an hour or more, i get bored and turned off the PS3
My guess is it never returns to 0x33 because that speed is used as the "warm up" (or preheat) speed

 

[sandungas]

that's something I don't like about the state of gaming -- having to worry about whether or not the system is cool enough. I've heard of people boiling eggs on a ps4 pro. that's just not right.

In the next gen probably is going to be fixed finally, with APUs made in 7nm
One thing i dont like from sony is everytime they reduces heat and power (by updating the electronics) they reduces the size of the console or/and they reduces the fan and heatsink sizes, and this results in similar noise levels and heat than previous model, genius :'(

With the PS5 and 7nm APU they have a chance to make a very silent and fresh console... i just hope they forgets a bit about reducing sizes, they just needs to make room for a big heatsink and a big fan
Bigger heatsink = bigger dissipation area
Bigger fan = bigger amount of air moved at the same speed (so you can use lower speeds)

With a 7nm APU i guess is not going to be needed to make fan and heatsink much bigger, but anyway... if is needed PLEASE DO IT !
If i have to choose in between having an small horny and noisy console... or a cool and silent console that is double the bigger (but is going to last for lot of years), then i have no doubt, gimme the big one


*PS4 PRON is a stepback imo (PS4 slim ftw)

 

[dmr]

Very well man, I agree with everything you said. That is why I avoid webman fan regulation, because sudden changes in fan speed. Is there another application for fan control that has at least 4 fan speed states? For example one speed up to 63C, second up to 68C third up to 70 and 4th up to 75C. That would be better solution than what weban gave us.

And how you explain that fan speed stay same at 28% from game state 65/67, and after 40 minutes in XMB at state 59/59 fan speed didn't change?

 

[sandungas]

40 minutes ? thats a lot, the only reasons i can imagine why you are having that are:
-you have not counted well the time
-you made something that created an unestable variation of temperatures and syscon entered in some "special" fan mode
-webman has a bug and is not displaying well the speeds when are stepping back

If you repeat the test try to make it in the most straightforward way posible... from ambient ---> ingame (up to 65ºC or so), then quit to XMB and wait
Im suggesting in doing it this way because if you enter-quit games several times maybe syscon does different things than the ones i noticed

And believe in your ears XD ... if you hear that speed is decreasing (because noise level does a notable reduction suddenlly), but the speed value in screen is the same, then this is a bug
Im sure the speeds decreases when the temperatures reduces (with that long time delays), and i made the tests in a PS3 like yours

I made all this tests with irisman btw, it has an screen where you can configure 5 temperature ranges and 7 speeds

That screen is the one i was using to see the values that i posted later in the table in wiki
You jusst need to turn ON PS3 (from ambient), enter in that screen and wait patiently (while taking notes in PC of your results, or chilling browsing some webs in the meantime)

At that time there was no software able to display temperatures/speeds inside games or XMB... so i imagined a handy trick to do it
Lets use again the example of uncharted 2 that we know moves around a range of temperatures of 65-73 (in CECH25xx with factory fan control)

Ok, so the range is 8ºC... and inside irisman i can configure 5 speeds... so...
speed 1 = triggered at 66ºC with fan speed at 0x60
speed 2 = triggered at 68ºC with fan speed at 0x70
speed 3 = triggered at 70ºC with fan speed at 0x80
speed 4 = triggered at 72ºC with fan speed at 0x90
speed 5 = triggered at 74ºC with fan speed at 0xFF (this is a "def con" warning)

The "defcon" warning is a temperature where im not supposed to be ever... and the speed asociated with it is huge
If for some reason i reach that temperature im going to notice it instantly because the noise level is going to be brutal

Well... the other speeds follows this rule too... are very different from the previous/next so everytime it changes the speed you are going to notice it for sure

Then after preparing that... i can enter uncharted 2 (and lower the TV volume, or even mute it if is the first time you do this) and play for some time
Is easy to count if the fan is doing "one step up" or "one step down" because the noise
And you know (because i configured it this way), every change is 2 degrees (66ºC --> 68ºC --> 70ºC --> 72ºC --> 74ºC)

But this im explaining was just a trick because at that time there was no other way to "see" the speed and temp values ingame or in XMB
And this kind of configuration is not good to use it normally, is only for tests because im covering a very small range of temperatures, and im using very weird and annoying speeds

It was a bit annoying because every 2 or 3 minutes of playthrought the speeds was changing. In some way it was excesively sensitive, but it was made on purpose for the tests
I completed the last of us with a config like that though, this is why i realized the temperatures rises up like hell in a couple of points of the game (with water, it seems naughty dog game engine suffers with water, i bet the same happens in uncharted because is the same game engine)