PS3 PS3 temperature vs ambient temperatures

[leandronb]

Thats perfect, just 1ºC or 2ºC of difference after 30 minutes inside an app, it means the heat is trasfered very well
If you repeat the test starting from ambient (around 30ºC) you are going to see the CELL is a bit hotter in the first lets say... 5 minutes... and after 30 minutes the temperatures are ballanced like in your results

Btw, for this test is not needed to use custom fan settings, you could disable webman (or configure to syscon to alllow it to use the factory settings), and the results will be a bit higer temperatures (maybe 4ºC or 5ºC bigger for both), but the difference in betweem them should still be very small


Ive seen people reporting 50ºC or 45ºC and my reactions are 1) wtf, 2) or is a lie, 3) or incorect meassurement

The "wtf" is because are temperatures smaller than the 55ºC used as a "warmup" in all PS3 models... i mean... is too much different than the factory settings, and personally i dont like to differ too much from what the engineers calculated
In other words... i would like to "copy" the temperature curve used by sony enginers and "customize" it a bit (of course boosting speeds at all ranges), but preserving the curvature i painted in a graphic in my previous post of this thread
Since some months ago when some syscon models was dumped (and the fantable mapped) we can do this adjustments a lot more accuratelly because we know exactly what sony engineers did
We dont have samples of the fantable used in PS3 slim models though, and this sucks because sony made an important change in the fan control for the slims (we could say what they was doing in PS3 FAT was bad, and they fixed it for slims, so the fantable of the PS3 slims is "the good one", even for using it in FATS)

The "lie" is because some people thinks this is a competition, if they sees the RSX is moving temperatures up and down 55ºC and 60ºC they reports 54ºC because it makes them feel better

The "incorrect meassurement" is because well... you know... what we need to check is the max temperature peaks (that usually happens in RSX), and to achieve this we really need to make a good "stress test"... and we should be able to replicate it as many times we need, becaue this allows us to compare the results before/after a thermal paste replacement, or some hardware modding, and to compare with other users, etc...
Some people just reports a value that is an intermediate temperature (not the max) so is pointless

------------
There are 2 kind of tests you could do, in the test i was talking about with @leandronb the goal is to have a very stable temperature, and the better way to achieve an stable temperature is either in main XMB, or inside a homebrew app (idle and waiting as many time is needed for the temperatures to stabilize)

The other is like an stress test, and is something different, here the goal is to generate the higher temperature posible to find the max

So do you think i should keep using AUTO mode in webman, or let syscon do the job? When i received the console after the delid and started playing call of juarez gunslinger in syscon mode the fans were not ramping up but temp was increasing fast and it worried me a bit, and i immediately changed to AUTO mode. In a previous post you said that 70c is a good limit for temps because it allows a bit of safe room if temps increase 5c really fast. Is this temperature for both CPU and RSX?

Running GTA V, that is one of the worst case scenarios, i noticed that CPU was really lower compared to RSX in auto mode (CPU 60, RSC 68), and with 47% fan speed is definitely audible. I did the same test again with MAX 68c and the result was almost the same, the fan sometimes increased to 50% but returned to 47%, and temps were the same as AUTO mode. If is safe, i would like a more silent fan speed, so can i run at MAX 70c to decrease fan speed or is better keep at AUTO?

Also does it make a difference running the PS3 in vertical or horizontal? I did some tests previously and vertical was a bit cooler, but it was before the delid, so the temps were so high that the difference could be just margin of error.

 

[pinky]

So do you think i should keep using AUTO mode in webman, or let syscon do the job? When i received the console after the delid and started playing call of juarez gunslinger in syscon mode the fans were not ramping up but temp was increasing fast and it worried me a bit, and i immediately changed to AUTO mode. In a previous post you said that 70c is a good limit for temps because it allows a bit of safe room if temps increase 5c really fast. Is this temperature for both CPU and RSX?

Running GTA V, that is one of the worst case scenarios, i noticed that CPU was really lower compared to RSX in auto mode (CPU 60, RSC 68), and with 47% fan speed is definitely audible. I did the same test again with MAX 68c and the result was almost the same, the fan sometimes increased to 50% but returned to 47%, and temps were the same as AUTO mode. If is safe, i would like a more silent fan speed, so can i run at MAX 70c to decrease fan speed or is better keep at AUTO?

Also does it make a difference running the PS3 in vertical or horizontal? I did some tests previously and vertical was a bit cooler, but it was before the delid, so the temps were so high that the difference could be just margin of error.

I'd use surround sound to drone out the noise. I've heard the ps4 pro can get very loud due to the heat. I was going to buy one for final fantasy vii remake, but I'd probably rather get a slim. I don't need something overheating on me.

 

[sandungas]

So do you think i should keep using AUTO mode in webman, or let syscon do the job? When i received the console after the delid and started playing call of juarez gunslinger in syscon mode the fans were not ramping up but temp was increasing fast and it worried me a bit, and i immediately changed to AUTO mode. In a previous post you said that 70c is a good limit for temps because it allows a bit of safe room if temps increase 5c really fast. Is this temperature for both CPU and RSX?

Running GTA V, that is one of the worst case scenarios, i noticed that CPU was really lower compared to RSX in auto mode (CPU 60, RSC 68), and with 47% fan speed is definitely audible. I did the same test again with MAX 68c and the result was almost the same, the fan sometimes increased to 50% but returned to 47%, and temps were the same as AUTO mode. If is safe, i would like a more silent fan speed, so can i run at MAX 70c to decrease fan speed or is better keep at AUTO?

Also does it make a difference running the PS3 in vertical or horizontal? I did some tests previously and vertical was a bit cooler, but it was before the delid, so the temps were so high that the difference could be just margin of error.

The temperatures you are having now are very good, dont get scared, try to play a demanding game in syscon mode... i dont think the temperatures are going to increase way over 75ºC... and thats still safe, a bit too close to the safe limit, and maybe not good to mantain them so high for long time, but a stress test to see some max peaks of 75ºC are not going to damage it

If you see something close to 80ºC yeah... thats the time to stop the stress test
Most probably you are not going to see 80ºC though (but you are going to be close, i guess something like 74ºC or 78ºC)

So is a personal preference, now that your PS3 is working nice (as good as the first day it was bought in the shop) you have a bit of freedom to try to reduce noise by lowering the speeds a bit

And btw, the difference of 8ºC of temperatures you had with a big workload is pretty much the same i had in my CECH-25xx when i was doing this tests
I mean... every PS3 results in different temperatures, but something in between 5ºC and 10ºC could be considered completly normal
If you only have 5ºC of difference then is awesome
If you have 10ºC of difference is ok, nothing to worry
The problem is if you see a difference bigger than 15ºC or 20ºC

*Note i use to round the values to multiplyers of 5 because i consider is a bit pointless to talk about exact values, i think we are dragging some error at all times, so better to think in them as small "ranges" of 5ºC each


----------
Edit:
About placing the console vertically... well, this was working better with the PS3 fats, but i think in the PS3 slims doesnt makes much difference, i never tryed to do a good comparison of temperatures in between vertical and horizontal because i prefer to have the consoles in horizontal

 

[leandronb]

The temperatures you are having now are very good, dont get scared, try to play a demanding game in syscon mode... i dont think the temperatures are going to increase way over 75ºC... and thats still safe, a bit too close to the safe limit, and maybe not good to mantain them so high for long time, but a stress test to see some max peaks of 75ºC are not going to damage it

If you see something close to 80ºC yeah... thats the time to stop the stress test
Most probably you are not going to see 80ºC though (but you are going to be close, i guess something like 74ºC or 78ºC)

So is a personal preference, now that your PS3 is working nice (as good as the first day it was bought in the shop) you have a bit of freedom to try to reduce noise by lowering the speeds a bit

And btw, the difference of 8ºC of temperatures you had with a big workload is pretty much the same i had in my CECH-25xx when i was doing this tests
I mean... every PS3 results in different temperatures, but something in between 5ºC and 10ºC could be considered completly normal
If you only have 5ºC of difference then is awesome
If you have 10ºC of difference is ok, nothing to worry
The problem is if you see a difference bigger than 15ºC or 20ºC

*Note i use to round the values to multiplyers of 5 because i consider is a bit pointless to talk about exact values, i think we are dragging some error at all times, so better to think in them as small "ranges" of 5ºC each


----------
Edit:
About placing the console vertically... well, this was working better with the PS3 fats, but i think in the PS3 slims doesnt makes much difference, i never tryed to do a good comparison of temperatures in between vertical and horizontal because i prefer to have the consoles in horizontal

Just tried using syscon with a game that i will play soon, Enslaved oddyssey to the west. And got a bit worried. CPU temps was just fine at 68c but RSX was really hot at 78c, and the fan just stayed at 28%. Waited a little bit more just to see if the fan would increase the speed, but still nothing, rsx increased to 79c and fan at 28%, and i quit the game. The game was only running for 16 minutes.
I always read that RSX is a bit more sensitive with temps than CELL that is why it worried me. Is this information correct or both CELL and RSX have the same thermal limits?

 

[sandungas]

Just tried using syscon with a game that i will play soon, Enslaved oddyssey to the west. And got a bit worried. CPU temps was just fine at 68c but RSX was really hot at 78c, and the fan just stayed at 28%. Waited a little bit more just to see if the fan would increase the speed, but still nothing, rsx increased to 79c and fan at 28%, and i quit the game. The game was only running for 16 minutes.
I always read that RSX is a bit more sensitive with temps than CELL that is why it worried me. Is this information correct or both CELL and RSX have the same thermal limits?

Ok, the test was mostly to see if your PS3 was able to stay under 80ºC with the factory settings, but it looks it cant. The results of this test depends a bit of the ambient temperature, maybe in winter it would be a few degress lower, anyway in your case is better if you use custom fan settings

I think the difference of 10ºC in between CELL and RSX is not something to worry, not sure if i mentioned it before, but that difference increases exponentially... so when you are in a high range of temperatures yeah... it could be notable, but your PS3 is fine
If you increase fan speeds to stay in a lower range that difference will be smaller (the 8ºC you mentioned you had before with webman or maybe 6ºC, is ok)

I dont really know about the thermal limits, i guess RSX is more prone to be damaged but is just a guess

 

[db260179]

Apart from the delidding, the fan bushes are probably on their way out.

I've had a few PS3's have the exact same issue, and it turned out the fan rotating speed had dropped, so wasnt pushing the air out fast enough.

The syscon default settings allow the temps to go to 85C, but i wouldnt recommend the RSX going to 79C as that will end its life sooner.

If you have the money spare, buy a new fan.

https://www.aliexpress.com/item/400...earchweb0_0,searchweb201602_,searchweb201603_

Just tried using syscon with a game that i will play soon, Enslaved oddyssey to the west. And got a bit worried. CPU temps was just fine at 68c but RSX was really hot at 78c, and the fan just stayed at 28%. Waited a little bit more just to see if the fan would increase the speed, but still nothing, rsx increased to 79c and fan at 28%, and i quit the game. The game was only running for 16 minutes.
I always read that RSX is a bit more sensitive with temps than CELL that is why it worried me. Is this information correct or both CELL and RSX have the same thermal limits?

 

[leandronb]

Yesterday I did a lot of tests with my PS3 to cover all the temperatures and fan speeds in most scenarios.
At ambient temperature around 30c, or a bit more because is really hot in here, and when using ar conditioner set to 22c.
I used XMB with sparks turned on and GTA V in the nearest corner to the left of Franklin's house.
I tested with fan speed set to manual at 40, 50, 60 and 70% and waited for the temperatures to stabilize. I also tested with AUTO #2 mode.
In the table it shows CPU/RSX temperature and on AUTO it shows the fan speed too.

 

[sandungas]

My oppinion about that results...
The difference in between 40% and 50% is notable, is an improvement of around 6ºC under all conditions
In between 50% and 60% the improvement is smaller
And in between 60% and 70% is even smaller

It makes me remember an story i did read about the bugatti beyron (not sure if completly true), it was designed to break an speed record, they started with a engine of 8 cilynders in V, after some revisions, improvements and tunning they was achieving a speed where it required a huge amount of horsepower to increment the speed in 1 unit
So they decided to use an engine of W16 (this is glueing 2 * v8 together), and they achieved more speed, not too much more
Is not a direct calculation because the most faster you try to move the biggest is the opposition of the air that is trying to stop you

So... there is a "hotspot" where the amount of energy required gives a reasonable result, but if you try to exceed that point you are wasting energy
In your PS3 it seems to be around 50% fan speed, just compare the results at 50% with 70%, the difference is not so big


Edit:
Or better said... there is a range of fan speeds that could be considered efficient, in the sense that the amount of energy invested results in a notable improvement
In this case the max speed that keeps you inside that range is 50%, but lower speeds are efficient too, of course

Spoiler

Is just we dont want to go too low