PS3 FAT Overheating

snkplkn said:
Syscon fan control doesn't go over 30% ?
What does it mean by "No more fan steps ?"
Click to expand...
I guess you mean the info in the table for the PS3 fats CECHCxx models, right ?
Well... the person that made that test reported that values to me, and when i said that was weird he insisted that was right... i doubt it though, but this is the only test we have from PS3 fat
Also he stopped the testing at CELL=75ºC and RSX=68ºC this is why after that point is indicated with "no more fan steps ?" (with a question mark, because i guess there are more in PS3 fats)

The values in the table for the PS3 slim CECH-25xx are fine though, because i made it and verifyed it double and triple times in different days (im very picky with this kind of things to minimize errors and mistakes and im sure about it)
snkplkn said:
Also i noticed that Syscon responds to temperature when playing the game & raise the fan speed somewhat. But when in XMB, Syscon is not as reactive to rising temperature. Is it programmed that way? Looks like Sony made Syscon to work primarily in game (and that too poorly), but didn't programmed as such for idle/XMB.
Click to expand...
Syscon uses a timer when temperature is decreasing, the rule is something like this:
If the temperature lowers to the preceding "fan step" the speed doesnt decreases and the timer starts counting time. The speed is reduced only after an amount of time has passed (the value of the timer)

The purpose of timer is to "bypass" the temperature peaks that happens ingame making the fan control a bit "insensitive" to changes only when temperature is decreasing
This way the fan speed is not changing constantly up and down... but instead is keept "up" in the max speed you had some minutes before

Look at the values of the table for a CECH-25xx... imagine you are playing a game that keeps the console at 64º/64º (so we are at step 7)
Inside a game there are temperature peaks up and down that happens very fast, and could around 3ºC or 5ºC
So... in this example the fan should be changing the speeds from step 6, 7, and 8 and this would be very annoying, what the PS3 would do in this case is to keep the fan at step 8 at all times, and it would step back X minutes (the timer) after reaching the temperature values of step 7
Is hard to explain :D

-------
Anyway... i guess what happened to you is you was trying to see if the fan speed was changing after exiting a game... and thats confusing because this is the scenario i was explaining when is using the timer

In other words... when temperature is increasing the speed changes are very sensitive (syscon takes samples of the sensors with a precission of miliseconds) but when temperature is decreasing syscon is doing a "preventive" speed control
Also... syscon doesnt knows if you are in XMB or inside a game... or better said syscon doesnt cares about software, is going to do the same fan speed control even if the firmware is corrupted
 
Is Dynamic FCU option unsuitable for FAT PS3? Is it better to set fixed Fan speed (43% or 50%)?

In Dynamic FCU setting (Payload), i don't mind fan changing speed because i don't even hear the fan speed which think stays around 43% or lower. Or should i set to 43% fixed & let it be? I think i would be able to maintain temps (RSX below 70 & CPU below 70 at that speed). But i haven't tried some exotic game yet like TLOU or Gran Turismo. I have read somewhere that they just crash on FAT PS3 & specially when thermal paste not replaced under IHS.

Any more games that should be avoided on FAT models?

By the way, what are these so called Firmware updates? Why doesn't Sony just give a patch for Syscon specifically for old FAT models in latest releases of firmware?

If Syscon has problem in old FAT models, shouldn't it be a focus for Sony to patch stock Syscon to make it more versatile?

Also i am wondering what can be the true/ideal temps for CPU & RSX. I guess i have to replace thermal paste underneath IHS to find that out.
 
Last edited:
snkplkn said:
If Syscon has problem in old FAT models, shouldn't it be a focus for Sony to patch stock Syscon to make it more versatile?
Click to expand...

FYI PS3 is discontinued worldwide since May 2017 ;) $ony don't care if the consoles get fried, they only want to protect their IP adding security patches or maintaining their side businesses (movies, SEN, etc.).

IMO The ideal temperature is between 65-70 degrees, but there are other factors to consider like room temperature, ventilation and how you use your console (e.g 3-8 hours of continue play or just to play movies occasionally).
 
Inside all the PS3UPDAT.PUP files there is a PKG specific to be installed inside syscon chip, is the "syscon firmware"
https://www.psdevwiki.com/ps3/System_Controller_Firmware#Known_Retail_syscon_update_packages
Only exists a few applicable for every syscon model, and is unknown what was updated by installing them
You can see which version you have installed in your PS3 by doing a button combo and entering in this screen:
https://www.psdevwiki.com/ps3/More_System_Information

In DECR-1xxx (tool) PS3 models there is a daughterboard named "communication" processor that includes a secondary syscon with some unlocked features, and running an small embedded OS
Im sure when the engineers was making tests with prototypes of the PS3 models they was able to change the fan speeds in a easy way, in a DECR probably they was doing it by reprogramming onboard syscon with the communication processor
But this applyes to all PS3 models, think in it... a team of engineers designing the motherboard of PS3 superslims... if at some point they wanted to change the fan speed setting do you think they was using some kind of "flasher" to write in syscon ?... no way, most probably they was simply updating "syscon firmware" by installing a PS3UPDAT.PUP
Dunno, are just speculations

A reason why they never changed the fan speed settings after a PS3 model was launched is because that fan speed settings are a decission from the directives of the company, the goal is to have a console the most silent posible (with the payback of having high temperatures, but who gives a flying f*** about that ?)
Most consumers thinks the console is not overheating because the fan doesnt makes noise... genius

----------
Btw, some months ago there was published some datasheets of a different chip that should be considered "a brother" of the PS3 syscon, and it did look it have some circuitry inside for thermal and fan control
I always thought the PS3 syscon uses a table with the values for speeds and temperature hardcoded (this is why i decided to make the table in wiki because i guessed the way how is represented in that table should match partially the way how the settings are stored inside syscon).... but that datasheets made me doubt

Another btw... the whole "PS3 familly" uses several syscon models, that should be grouped as "syscon series"
The first syscon series seems to lack some features related with thermal control... it seems it doesnt have as many "fan steps" as the slims. This has not been completly verifyed, but by looking at the table in wiki it simply looks that (in plain words) they made it wrong
 
Consoles are weird!! I haven't owned a game console since PS2 as i don't have much interest. But i recently started to do a little research. The facts i found about consoles makes no sense.

I read somewhere that original XBox 360 heating situation is worse than Fat PS3. The fans are even less responsive. And they designed the exterior body of 1st Gen 360, even without considering the space for internal hardware! (its like designing car chassis without even considering which engine & spare parts have to installed in it) Then 360s started overheating and dropping dead with RRODs. 50% of consoles just failed!! And they had to spend Billion $$$ just for all the replacements & extended warranties. All because of bad designs and overheating.

Looks like "upper" management of these big companies are MORONS.
 
Some updates.

So i tried several games to check the performance.

I played TLOU and Tomb Raider. I can notice fan kick in hard and i suspect reach around 50% though i still don't hear any noise. But i can't be sure. FCU doesn't give in game Temps and fan speeds. The FCU only gives indication that Temps are running under 70 with LED indicators. As i quit the game and immediately go to FCU from XMB (takes about 5-10 seconds) i see Temps RSX at 55-60 Degrees and CPU at 65 Degrees. Fans at 00x56 - 00x60 (which are around 35%-37% i think). These are after 5-10 seconds immediately after quitting the game.

I Played around 4 hours of TLOU without any problems except FPS drop on some rare occasions. Everything seems to run smoothly.

I also observed that RSX is usually 5-10 degrees cooler that CPU at all times.

I read that Webman can give in game temps & fan speed. But its not a simple app like FCU as it comes with a lot of features which tie into many parts of cfw. Also its difficult to remove. I am presently running Multiman without any problems.

Is there any other simple and lite homebrew/utility which can be run in background to show in game Temps and fan speed?
 
That video is a "must watch", i just want to comment something

I totally agree in his analysis of the problem of the heatsink surface (lets say the first 6 minutes of the video), i also agree with the rude words he said, thats not aceptable, and i can tell you is true because i could see by myself lot of time ago

Also, if someone was not aware of the trick... the way how he checks the planarity of the surfaces with a razor blade + a lightpoint at the other side is used even by professionals... is very accurate
The point is... the razor blade should block all light completly

But what he does by sandpapering the surfaces cant be considered a "lapping" because he is doinig it by moving the sandpaper with his hand over the surface... and thats bad because is not precise enought (believe me, is imposible to create a perfectly plane surface in a block of metal with your hands and simple tools)

What you need to do is to take a glass (from a window, or whatever you have at home)... because the glass have a surface that uses to be perfectly plane... and i mean... at atomic level, the atoms of the glass are very well ordered following a "cristalization pattern"

So... you place the glass on top of a table... then you need to stick the sandpaper on the glass (with double side tape)... and then you move the metal surface (either the IHS or the heatsink) on top of the sandpaper, in circles, slow, and applying minimal pressure

Also, the sandpaper you should use is named "water sandpaper" of the most fine quality... with this sandpaper you need to add some waterdrops on top of it... and the material particles removed from the metal becomes a gray substance
From time to time you clean the metal with a cloth + more water to check how the surface is becoming plane.... and repeat... and repeat... and repeat
This way you are going to have a perfectly plane surface... actually it will be like a shiny mirror if well done

After that, if yo do the check with the razor blade you will see that it doesnt allows to pass the light, thats the proof your "lapping" was well made :encouragement:


In resume... when he says he is "lapping" the metal surfaces he is not doing a "lapping", what he is doing is simply sandpapering it, is ok because he is very picky at the details so his result is better than what he had from factory.... but if you do a real lapping the results are going to be much better :encouragement:
 
Last edited:
sandungas said:
Syscon uses a timer when temperature is decreasing, the rule is something like this:
If the temperature lowers to the preceding "fan step" the speed doesnt decreases and the timer starts counting time. The speed is reduced only after an amount of time has passed (the value of the timer)

The purpose of timer is to "bypass" the temperature peaks that happens ingame making the fan control a bit "insensitive" to changes only when temperature is decreasing
This way the fan speed is not changing constantly up and down... but instead is keept "up" in the max speed you had some minutes before

Look at the values of the table for a CECH-25xx... imagine you are playing a game that keeps the console at 64º/64º (so we are at step 7)
Inside a game there are temperature peaks up and down that happens very fast, and could around 3ºC or 5ºC
So... in this example the fan should be changing the speeds from step 6, 7, and 8 and this would be very annoying, what the PS3 would do in this case is to keep the fan at step 8 at all times, and it would step back X minutes (the timer) after reaching the temperature values of step 7
Is hard to explain :D

-------
Anyway... i guess what happened to you is you was trying to see if the fan speed was changing after exiting a game... and thats confusing because this is the scenario i was explaining when is using the timer

In other words... when temperature is increasing the speed changes are very sensitive (syscon takes samples of the sensors with a precission of miliseconds) but when temperature is decreasing syscon is doing a "preventive" speed control
Also... syscon doesnt knows if you are in XMB or inside a game... or better said syscon doesnt cares about software, is going to do the same fan speed control even if the firmware is corrupted
Click to expand...
Hmmm, this is something i wrote in april 2019 (before syscon was hacked), now im reading myself and i feel like i should add some corrections :)

What i said was a deduction based in a lot of tests and experiments i made with a CECH-25xx, at that point i thought the values stored inside syscon fantable for every "fan step" was composed by: TempU, Duty, and a Timer
But now we know are composed by: TempU, Duty, and TempD

In the practise it works a bit in the same way, the fan speed only decreases when you reach TempD... but is like a time delay because the fan is more "insensitive" to the decreasing temperatures

Sony did this to prevent the frequent speed changes... just to use a rought example:
In a game session of 4 hours (starting from ambient temperature) and without this system we would have a total of hundreds of fan speed changes
And with this system we have only 10 or 15 fan speed changes in the total gaming session of 4 hours... the fan speed "ramps up" but doesnt returns back unless very special conditions (if we quit the game, or if the game reduces his workload a lot for a lot of minutes)
 
Last edited:
You must log in or register to reply here.

Similar threads

Back
Top