PS3 PS3 SLIM fan comparison data gathering

[Mitsu™]

Quite recently, my beloved Delta KFB1012HE died abruptly, which made me more interested in the different models available for the Slim (specifically for the 20XX/21XX/25XX).
There are some topics and videos about FAT's fans models comparison (efficiency, noise, etc...) but practically nothing for the slims. Scattered data, at best.
(Or did I miss something... ?)

That's why I'm opening this thread.
Feel free to contribute !

If possible, share your results like this :

- Fan model
- Slim model
- CPU/GPU temps in °C after 1 hour idle in Movian (yes, Movian. The XMB is fine and all, but is it a static wallpaper ? Waves ? Waves + sparkles ? etc...)
- Fan speed (would be better to set a manual % for this test, like 40%).
- Room temperature in °C

If you have ideas to improve the methodology, I'm listening !
I think this should be sufficient, but... ¯⁠\⁠_⁠(⁠ツ⁠)⁠_⁠/⁠¯

Don't forget to let it cool down between two tests.

Of course, the ideal tester would be someone with multiple fan models to try in one console, to get more significant results for comparison.
But even if you only have one model to test, it's still something. We can always try to cross-check the data later.

About noise, it's more complicated since there is a huge part of subjectivity. Different sensibilities for different frequencies, etc...
Still, feel free to share your experience.

There are 5 models for the 20XX/21XX/25XX :
• Nidec G10C12MS1AH-56J14
• Nidec G10C12MS2AH-56J14
• Delta KFB1012HE
• Delta KSB1012HE
• NMB-MAT BG1004-B045-P00

...and only one fan for the 30XX series, so nothing to compare :P
• Nidec G80E12NS1ZN-56J14

My personal results :

• Delta KFB1012HE
- Slim 2504A
- Temps after 1 hour idle in Movian : CPU: 46/GPU: 46
- Fan speed : 40% manual
- Room temperature : 21°C

• Nidec G10C12MS1AH-56J14
- Slim 2504A
- Temps after 1 hour idle in Movian : CPU: 49/GPU: 50
- Fan speed : 40% manual
- Room temperature : 21°C

• NMB-MAT BG1004-B045-P00
- Slim 2504A
- Temps after 1 hour idle in Movian : CPU: 49/GPU: 50
- Fan speed : 40% manual
- Room temperature : 21°C

The Delta KFB1012HE is significantly better for cooling. A shame that mine died...

As for the noise, I'll say that the Nidec is probably the quieter of these three. The Delta one is bearable, a bit more high-pitched though.

The NMB is the worst for me.
It's nice to create an 'original' design with more blades and all, but it would be even better if it actually improved something. Unfortunately, that's not the case at all.
The simpler Nidec G10C12MS1AH-56J14 is as effective and even quieter.

I found a Nidec G10C12MS2AH-56J14 that I should receive soon. Can't wait to see how it performs !

Also, interesting things I noticed:

The two Nidec have different amperage, okay. That's a noticeable difference. Maybe the G10C12MS2AH-56J14 will perform better ? I don't know. That would be logical.

But apparently, there isn't such difference between the two Delta. Same amperage, same blades count.
So, maybe they modified the shape of the blades ?
Or... the first model was defective ?
Whatever.

If someone has a Delta KSB1012HE (or even better, both Delta fans !), I'd love to know its effectiveness !
I'll try to find one but it seems to be rare where I live

 

[Skire]

In my opinion the only valid way is to test with equal noise levels. 40% fan speed for one fan could be a totally different RPM value and noise level for another fan. Also you would need quite a professional noise test setup to gather valid data.

 

[Mitsu™]

I see your point and you're right, but as you said, you'll need a professional noise test setup. Something that almost no one have at home, I guess.
We need an accessible methodology.

Another interesting way to test would be to let SYSCON manage the fan speed instead of setting a manual %.
But there are different SYSCON curves depending on models IIRC, so once again, it'd be better to have the "one user with multiple fans for one console" tester profile.

 

[NightLess]

Have you tried to install a 21xx heatsink on a 25xx? Are they compatible? The 21xx heatsink has better heat dissipation than the 25xx.

 

[Mitsu™]

Nope, I don't have one unfortunately.
Finding a 21XX at a reasonable price where I live is a challenge, and parts are rare too
From pictures, the 21XX heatsink doesn't seems to be drastically better than the one you'll find in a 25XX, idk.

The bests are the 20XX ones by far, it's really a shame we can't use them :\
Seriously, look at this beauty !

 

[NightLess]

Nope, I don't have one unfortunately.
Finding a 21XX at a reasonable price where I live is a challenge, and parts are rare too
From pictures, the 21XX heatsink doesn't seems to be drastically better than the one you'll find in a 25XX, idk.
View attachment 44228
The bests are the 20XX ones by far, it's really a shame we can't use them :\
Seriously, look at this beauty !
View attachment 44229

Yes, the heatsink of 20xx is the strongest one for slim, but unfortunately it is not compatible with later models.

Can you help me see what is in address 7f2 in syscon on a Slim 25xx? I messed up when setting up the dynamic fan in syscon, and the correct address to verify is w 7fe. I filled in w 7f2, and I have no idea what is in 7f2. Thank you.

 

[gytdoo]

This thread caught my attention due to the "jet engine" effect I'm experiencing on a 2501a (September 2010) slim I was given recently. It's been cleaned and repasted, I just think the fan itself is going out. The stock fan was a Delta KSB1012HE and I went ahead and ordered the Delta KFB1012HE you mentioned.

The stock fan would sound loud, whiny, and sliiiighy… wobbly? It certainly was working well regarding temps, just a little obnoxious to listen to.

• Delta KSB1012HE
- Slim 2501A
- Temps : CPU: 62/GPU: 64
- Fan speed : AUTO 2
- Room temperature : 21.6°C

The replacement delta seems to run slightly cooler and the sound is much more tolerable for me. Less high pitched, more of a blow than a whine.

• Delta KFB1012HE
- Slim 2501A
- Temps : CPU: 60/GPU: 63
- Fan speed : AUTO 2
- Room temperature : 21.6°C

In webman MOD I use the AUTO 2 setting, and I'm running a 650/800 oc. I tested in gt6 using replays, waiting about 30-45 minutes for temps to stabilize. The ps3 also had a chance to cool for a day between runs. I unfortunately haven't tested in movian, I'm pretty busy right now and am about to leave for vacation. Thought this was worth sharing anyway as I found the results interesting. I would compare the difference in sound to the steam deck lcd fans. While both worked well enough, one just sounds "better".

 

[Starship2]

Have you tried to install a 21xx heatsink on a 25xx? Are they compatible? The 21xx heatsink has better heat dissipation than the 25xx.

Good point. My personal result is that those delta fans in combination with the Heatsinks from 20xx/21xx models do perform the best. Balance between noise level and cooling performance is pretty decent.