Seasonic S12 Fan Upgrade

PSUs: The source of DC power for all components in the PC & often a big noise source.

Moderators: NeilBlanchard, Ralf Hutter, sthayashi, Lawrence Lee, Devonavar

hmronin
Posts: 27
Joined: Sun Aug 28, 2005 11:38 pm

Seasonic S12 Fan Upgrade

Post by hmronin » Sun Oct 09, 2005 2:39 pm

I'm considering replacing the ADDA fan used in my S12-330 with something else. Has anyone tried doing this and found quieter results using another brand of fan? Also be interested in any other technical parameters you had to consider, since the S12 has its own fan controller, as well as any procedures or hints you discovered, while switching the fans. Thanks.

Badger
Posts: 228
Joined: Thu Nov 25, 2004 8:57 am
Location: West Michigan, USA

Post by Badger » Sun Oct 09, 2005 4:02 pm

I'm kind of wondering the same thing. Would it be alright to run the stock fan in the S12-330 at 5v all the time, or replace it with another 120mm to always run at 5v?

frankgehry
Posts: 1424
Joined: Sun Dec 12, 2004 5:00 am
Location: New York, NY

Post by frankgehry » Sun Oct 09, 2005 6:39 pm

Replacing the adda in the most recent s12 with the earlier version's yate loon D12SM-12 would be a little quieter. You can read about the difference between the previous and latest model in the reviews. Also, from looking at the photos you would simply unsolder the adda wires and solder replacement fan leads in. It doesn't seem to have a connector. Or you could cut the wires and splice the new fan in.

Since the fan controller doesn't ramp up I don't see why you would want to run it at a constant 5v, unless you notice a lot of fan speed changes. If the psu is loaded so that the controller puts out more than 5v, it probably needs more cooling. The other thing is that the yate loon has sleeve bearings and most fan manufacturers do not recommend mounting sleeve bearing fans horizontally or in a raised temperature environment, as some psu's have 3 yr warranties.

The adda tops out at 1800rpm, the yate loon D12SM-12 at 1650. I would use a JMC 1225L-12 with a top speed of 1600rpm and dual ball bearings. - FG

You could also use a papst 4412 F2/GL (1600rpm). I have the fans I've mentioned, but I don't have an adda so I don't know if a swap would really help much. But adda fans do not hava a great reputation for being quiet, although silenX claims they are.

sun.moon
Posts: 174
Joined: Wed Sep 21, 2005 5:57 am
Location: Germany

Post by sun.moon » Mon Oct 10, 2005 12:06 pm

I am also interested in considering changing the adda fan in my S12 430.

@ frankgehry: is there a reason you did not mention the Nexus 120 or the Papst 4412 f/2GLL? (I believe they are 1000 and 1200 rpm respectively.) They would both be quieter than the fans you mentioned. Do they not push enough air?

In my system even at 100% load, the S12 fan does not go above 800 rpms. With this setup maybe a lower rpm fan might be desirable... :?:

frankgehry
Posts: 1424
Joined: Sun Dec 12, 2004 5:00 am
Location: New York, NY

Post by frankgehry » Mon Oct 10, 2005 1:43 pm

The fan controller voltage at startup according to the s12-430 review is 4.6v. I'm not sure if the nexus or papst 4412 F/2GLL would start at that voltage. Maybe it doesn't matter if the fan starts right away. You could investigate adding a kickstart to the fan controller (www.cpemma.com). If the adda is at 800rpm most of the time, the yate loon D12SM-12 or papst 4412F/2GL would be at (1650/1800) * 800 ~ 730rpm which would be fairly quiet. Using a lower rpm fan would of course provide less cfm that may or may not adequately cool the psu or help move air in you system. - FG

sun.moon
Posts: 174
Joined: Wed Sep 21, 2005 5:57 am
Location: Germany

Post by sun.moon » Tue Oct 11, 2005 12:44 am

I suppose the key points to investigate are:

1. would the Nexus or Papst 4412 F/2GLL start at 4.6V, and
2. what is the cm/h or cf/h for these fans compared to the adda fan at the same low voltages.
3. anything else to consider?

If the airflow is comparable to the adda fans, the PSU would be adequately cooled in the same manner, with lower rpms = less noise, right? From what I read, the Papst fans have higher airflow than Nexus fans – maybe a good solution?

If anybody could shed some light on this, it would be appreciated.

Thanks

sun.moon
Posts: 174
Joined: Wed Sep 21, 2005 5:57 am
Location: Germany

Post by sun.moon » Tue Oct 11, 2005 11:33 pm

So, after some investigating on my own, I have found the following:

ADDA AD1212LB-A73GL (S12 fan)

RPM: 1800
CFM: 72
(dBA: 34.4)
2.88W


Papst 4412 F/2GL

RPM: 1600
CFM: 55
(dBA: 26)
1,25W


Papst 4412 F/2GLL

RPM: 1200
CFM: 41
(dBA: 18 )
0,72W


Nexus D12SL-12

RPM: 1000
CFM: 37
(dBA: 22.8 )
1,8W



With this data, I would conclude that the Papst 4412 F/2GLL and Nexus D12SL-12 would pose a higher risk for not providing adequate airflow to cool the S12 PSU. Seasonic must have chosen a fan with max 72 CFM for a reason. Out of this fan selection, the Papst 4412 F/2GL would be the better alternative. It is also according to manufacturer specification significantly quieter than the stock ADDA fan. (I know that one can not rely alone on these specs, but at least it’s an indication).

A question that rises for me: is there a linear relation to decreasing Voltage to reduced RPMs and reduced CFM and reduced dBA? How does one calculate that?

e.g. 12V = 1200 RPM, 41 CFM, 18 dBA and 5V = ?? RPM, ?? CFM, ?? dBA

This would be interesting since if the ADDA fan at low voltages is so quiet that the noise level difference compared to the Papst 4412 F/2GL at the same low voltages is not discernable. As mentioned earlier, even at 100% load in my system the ADDA fan doesn’t go above 800 RPMs. Then it would make little sense to change the fan in the S12 (and thereby forego the warranty).

Another question: is the temperature based fan controller in the S12 aligned to the specifications of the ADDA fan? I mean, on hand the temperature does the controller power the ADDA fan with a given voltage to achieve a given RPM? Or does the controller increase the voltage to reach a cooling level independent of the RPMs? (I hope this question is understandable).

Edit: the question also remains if the 4412 F/2GL starts at 4.6V. Has anyone tried this?

@ hmronin: Hope you don't mind me in your thread. I thought it wouldn't make sense to start another one on this this topic.

frankgehry
Posts: 1424
Joined: Sun Dec 12, 2004 5:00 am
Location: New York, NY

Post by frankgehry » Wed Oct 12, 2005 8:03 am

voltage/rpm is almost linear.
cfm/rpm is almost linear.
dba is not linear.

You should compare the test results from the yate loon version of the s12 to the adda version. The sould level results for the adda version are at the end of the s12 430 review. There is approx. a 2 dbA difference between the two versions so that is the best you can expect unless you use a lower rpm fan than either the yate loon D12SM-12 or 4412F/2GL. You could listen to the sound recordings to see if it is worth doing. - FG

DG
Posts: 424
Joined: Thu Feb 26, 2004 3:40 pm
Location: EU

Post by DG » Wed Oct 12, 2005 10:26 am


hmronin
Posts: 27
Joined: Sun Aug 28, 2005 11:38 pm

Post by hmronin » Thu Oct 13, 2005 8:44 am

sun.moon wrote: @ hmronin: Hope you don't mind me in your thread. I thought it wouldn't make sense to start another one on this this topic.
I don't mind at all. In fact, the discussion so far has been very helpful. I haven't decided yet which fan to use, however I'm considering one of the lower speed fans you listed, either the Papbst or the Yate Loon.

I just need to verify which will work better with the S12's fan controller.

My own thoughts on this upgrade are this: An S12 in a P180 case has much less demand requirements than in any standard ATX case, due to the fact that it is only exposed to its resident heat, and that of any hard drives in the lower chamber (My opinion is that this heat load plays a very small factor in overall heat load to the PSU).

This is not the case if the S12 is in a standard ATX, where it is exposed to the greatest heat generators in a PC: the CPU and GPU. As explained in an article here in SPCR, the ATX design spec plans for the PSU fan to participate in removing hot air from inside the PC. Therefore, PSU companies must design their products with the expectation that their PSU's will be exposed to a proportionally large amount of externally created heat. My guess then , is that heat management in any quality PSU is robust enough to handle excessive external heat loads, while operating under full capacity.

Since my S12 in the P180 will never see this kind of external heat load, I'm willing to reduce the airflow capability in the S12 by using a lower speed fan, confident that I can infringe upon the built-in safety factor that Seasonic has to put into this S12, assuming it will be used in a standard case.

Worst case scenario, I end up buying another PSU sooner rather than later... :D

I would encourage more feedback and insight in this discussion, its been helpful to me. :D
Last edited by hmronin on Thu Oct 13, 2005 1:27 pm, edited 1 time in total.

MikeC
Site Admin
Posts: 12285
Joined: Sun Aug 11, 2002 3:26 pm
Location: Vancouver, BC, Canada
Contact:

Post by MikeC » Thu Oct 13, 2005 9:17 am

I know that this topic is about modding an S12 with a quieter fan for use in a P180. I have a radical suggestion.

1. Remove the cover off any (halfway decent?) PSU.
2. Remove the fan.
3. Mount it coverless and fanless in the P180.
4. Block the extra vent holes below the PSU exhaust on the back panel of the case.
5. Mount the quietest 120mm fan you can get your hands on in the PSU chamber and run it at the lower voltage that it can start at.
6. Relax and enjoy the quiet w/o worrying about the PSU overheating.

Caveats:
1. POTENTIALLY DANGEROUS EXPOSURE TO HIGH VOLTAGE if you open the case when the system is powered up or plugged into AC.
2. It will work more "safely" (thermally speaking) with more efficient PSUs that have bigger heatsinks.
3. You might even want to bend (or cut away) the "intake" panel so that the airflow from the mid-way fan can go unimpeded straight into the PSU.
4. If you have a PSU with a smart fan controller (like the Seasonics), you have the option of powering the 120mm fan via the PSU fan controller.

I have been meaning to try this for some time. Maybe next week... I've always meant for the P180 to be a silent PC experimenter's dream case. ;)

hmronin
Posts: 27
Joined: Sun Aug 28, 2005 11:38 pm

Post by hmronin » Thu Oct 13, 2005 1:23 pm

1. Remove the cover off any (halfway decent?) PSU.
2. Remove the fan.
3. Mount it coverless and fanless in the P180.
4. Block the extra vent holes below the PSU exhaust on the back panel of the case.
5. Mount the quietest 120mm fan you can get your hands on in the PSU chamber and run it at the lower voltage that it can start at.
6. Relax and enjoy the quiet w/o worrying about the PSU overheating.
Neat idea Mike...I hadn't thought of that. I'll post results if I decide to go this route. :D
I've always meant for the P180 to be a silent PC experimenter's dream case.
The P180 case is very special because of its revolutionary design. I was not in the market for a new case, but when I saw the internals of the P180, I knew I had to get one. It just offered so many unique capabilities, and offered the chance to customize my PC in very substantial ways. Thanks Mike, for your contribution to its design.

sun.moon
Posts: 174
Joined: Wed Sep 21, 2005 5:57 am
Location: Germany

Post by sun.moon » Thu Oct 13, 2005 1:50 pm

MikeCs idea sounds pretty radical, but it wouldn’t be much use to me I suppose, since I have a SLK3000B, where the CPU also plays an exhaust role.

I figured this out earlier today - maybe it's helpful for others too:

Code: Select all

V        RPM         CFM        dBA* 

AD1212LB-A73GL (Seasonic S12 Fan) 
12,      1800,0      72,0       34,4 
9,5      1425,0      57,0    
7,0      1050,0      42,0    
5,0      750,0       30,0    

Papst 4412 F/2GL 
12,      1600,0      55,0       26,0 
9,5      1266,7      43,5    
7,0      933,3       32,1    
5,0      666,7       22,9    

Papst 4412 F/2GLL 
12,      1200,0      41,0       18,0 
9,5      950,0       32,5    
7,0      700,0       23,9    
5,0      500,0       17,1    

Nexus D12SL-12 
12,      1000,0      37,0        22,8 
9,5      791,7       29,3    
7,0      583,3       21,6    
5,0      416,7       15,4 
*manufacturers spec - just an inicator

After having just performed some work on my system this evening, I am questioning if I need to exchange my S12 fan at all. Status:

SLK3000B w/ 2 Papst 4412 F/2GLL each at 5V
Seasonic S12 430W
ASUS P4PE Mainboard
P4 2.4 GHz w/ SI-120 & Papst 4412 F/2GLL at 5V
Sapphire ATI 9600 Pro w/ passive AeroCool VM-101
1GB Infineon PC2700/333 DDR SDRAM
Seagate Barracuda 7200.7 120GB SATA HDD (suspended)
Sony CD R/W
Philips DVD

The changes I just made were replacing the stock Antec 3-Speed set at low as an intake fan with the Papst 4412 F/2GLL. Then I set all three Papst fans to from 7V to 5V via the fan cable supplied with the S12 (incl. one Y cable). And finally I opened the CPU duct on the side of the case and placed a dust screen on it.

So now, according to my calculations, I have an exhaust of about 47 CFM (Papst + S12 Adda fan – which even at 100% load doesn’t go over 780 RPM = appr. 5V), and intake of 17 CFM. Actually I imagine that the total exhaust must be less, since there must be quite a bit of resistance moving the air through the PSU, but I don’t know how much.

The CPU fan sucks cool air direct to the CPU which may play an neutralizing role to the negative pressure in the case. Even without the CPU duct in place, I am sure that this has a large cooling effect since the SI-120 plus fan sits so close to the case side. I feel air going in with my fingers.

With these changes – even with reducing all fans to 5V, I have a 2C temperature drop on the CPU. The case temps dropped 1-2C. Ambient temps are similar to before, appr. 22-23C.

And my system is remarkably quiet – which is why I am seriously reconsidering replacing the adda fan in the S12 and foregoing the warranty (I just had the S12 replaced due to a defective SATA connector!). I am sure that it would make a difference, but would I even notice it????? Believe it or not, the loudest noise which I perceive now is a slight buzz from my monitor!! So now an LCD monitor is in scope :lol: The Seagate HDD may also be nearing its end. I have my eye on a Samsung P80...

SebRad
Patron of SPCR
Posts: 1121
Joined: Sun Nov 09, 2003 7:18 am
Location: UK

Post by SebRad » Thu Oct 13, 2005 2:45 pm

Hi, I have swapped the fan in my PSU (for a Panaflo L1A 80mm). It's lower speed (1900 vs 2500rpm) than the original and significantly quieter. (Although the near sealed intake duct probably helps too). The theory I'm working on is that as it's still connected to the fan controller if the PSU temps get very high it should take the fan to high RPM (which never happens) The overall effect being that, for example, at 50% output you have the temperature and associated fan voltage of 75% output, so the PSU runs hotter that it otherwise would but still with in spec range. It seams to me like a better method than hardwired fan voltage as there is still some scope for increased fan speed. Anyone understand what I'm trying to explain and got any thoughts on it?
(I think this is relavent to this thread as we're talking about swapping quieter/slower fans in to an S12)
Thanks, Seb

patricksumner
Posts: 30
Joined: Fri Mar 04, 2005 11:16 pm

Post by patricksumner » Thu Oct 13, 2005 2:47 pm

sun.moon wrote:So now an LCD monitor is in scope

Careful! I never realized how loud the high pitched squeal coming from the transformer for my LCD monitor was until I made my system much quieter and ran it at 2:00AM. It's now the most annoying component in my system. It's not the loudest component, but it's the most intrusive.

MikeC
Site Admin
Posts: 12285
Joined: Sun Aug 11, 2002 3:26 pm
Location: Vancouver, BC, Canada
Contact:

Post by MikeC » Thu Oct 13, 2005 2:59 pm

sun.moon wrote:After having just performed some work on my system this evening, I am questioning if I need to exchange my S12 fan at all....

...And my system is remarkably quiet – which is why I am seriously reconsidering replacing the adda fan in the S12 and foregoing the warranty (I just had the S12 replaced due to a defective SATA connector!). I am sure that it would make a difference, but would I even notice it????? Believe it or not, the loudest noise which I perceive now is a slight buzz from my monitor!! So now an LCD monitor is in scope :lol: .
:lol: :lol: This was my first reaction when I saw your orig post. Glad you've sorted through it all. 8)

sun.moon
Posts: 174
Joined: Wed Sep 21, 2005 5:57 am
Location: Germany

Post by sun.moon » Fri Oct 14, 2005 1:13 am

Thanks MikeC. It is fun to figure out some things by myself!

So, now some concluding remarks from my side.

Thanks to dukla2000 in this thread,
The standard formula for this is
dBA2 = dBA1 - 50 * log (rpm1/rpm2)
so a 4412FGL is 7.0dBA @ 5V.

Yeah, right! The formula ignores among other things the bearing noise that may not decrease at all at lower rpm. Then again I cannot hear my 4412FGL @5V @ 1m: for quality fans the formula is arguably a fair indicator, for other fans I have had the misfortune to own (e.g. Evercool EC12025M12S) it is clear to my ears there are some assumptions behind the formula that will explain cases where it clearly is wrong!
I have been able to calculate the dBA levels for these fans at different voltages (with the friendly support of MS Excel). I stress this is only very indicative, since I am sure that there are numerous variables which from fan to fan need consideration. And I am sure that a dBA figure cannot be a negative number as the 5V of the 4412 2/F2GLL. Also, these calculations are based solely on the manufacturer’s specification, which one must treat with suspicion anyhow.

Nevertheless, the dBA findings are interesting to me, and I hope to others as well. Also confirms for me, I will not exchange my S12 fan, since the 15.4 dBA of the Adda fan at 5V is not perceivable for (most) humans. It’s probably the air resistance in the PSU at 5V that one hears at all, not the fan itself. And at 1m my system is now below the ambient noise level in my room.

Code: Select all

V	   RPM  	   CFM	   dBA

AD1212LB-A73GL
12,0	1800,0	  72,0	  34,4
9,5	 1425,0	  57,0	  29,3
7,0	 1050,0	  42,0	  22,7
5,0	 750,0	   30,0	  15,4

Papst 4412 F/2GL
12,0	1600,0	  55,0	  26,0
9,5	 1266,7	  43,5	  20,9
7,0	 933,3	   32,1	  14,3
5,0	 666,7	   22,9	  7,0

Papst 4412 F/2GLL
12,0	1200,0	  41,0	  18,0
9,5	 950,0	   32,5	  12,9
7,0	 700,0	   23,9	  6,3
5,0	 500,0	   17,1	  -1,0

Nexus D12SL-12
12,0	1000,0	  37,0	  22,8
9,5	 791,7	   29,3	  17,7
7,0	 583,3	   21,6	  11,1
5,0	 416,7	   15,4	  3,8
These findings also confirm to me in a theoretical sense the quality of the Papst 4412 F/2GLL fan which I can definitely sense in practice. Although I must admit, I have not seen/tested a Nexus fan myself; I cannot imagine a discernable difference between the two fans at 5V.

Even though my questions are answered, it would be nice to see this list expanded with other fans which are interesting for silencers, also including the actual dBA figures tested in SPCR labs, and even maybe posted as a Sticky. If anyone is interested in pursuing this, I would be happy to mail the excel table I started – you don’t need to start from scratch.

-sun.moon

p.s. if anyone finds an error in my calculations/assumptions, I would be happy for corrections.

hmronin
Posts: 27
Joined: Sun Aug 28, 2005 11:38 pm

Post by hmronin » Fri Oct 14, 2005 7:42 am

sun.moon wrote:
And my system is remarkably quiet – which is why I am seriously reconsidering replacing the adda fan in the S12 and foregoing the warranty (I just had the S12 replaced due to a defective SATA connector!). I am sure that it would make a difference, but would I even notice it????? Believe it or not, the loudest noise which I perceive now is a slight buzz from my monitor!! So now an LCD monitor is in scope :lol: The Seagate HDD may also be nearing its end. I have my eye on a Samsung P80...
After installing a T-Balancer to control the CPU, GPU, and case fans (CPU and GPU fans run at 40% now, case fan sporadically), I too noticed the coil whine from my LCD monitor. After finding the solution to that noise problem from this forum, the S12 PSU noise became the next apparent noise maker...or maybe not.

My initial thought was that the S12 fan was making the noise, simply because I could pinpoint it, and because after stopping the CPU, GPU, and case fan with the T-Balancer, the noise didn't go away. Its possible though that the S12 is spinning at full RPM. This is something I'll need to check, though I'm not sure how.

I'll list my setup in my signature. If any of you have ideas or suggestions on improvements, I would be grateful. :wink:

sun.moon
Posts: 174
Joined: Wed Sep 21, 2005 5:57 am
Location: Germany

Post by sun.moon » Fri Oct 14, 2005 9:38 am

Hey hmronin, the S12 has a fan rpm monitor cable - on mine its a two wires, black and blue. It has only the RPM monitoring wire and ground, no power. I just hooked up to my motherboard (ASUS P4PE) and read out the rpms over the Asus PC Probe software. I am not familiar with the T-Balancer, but maybe it would be possible to read out the RPMs of the S12 over that as well?

And btw, I really doubt that your S12 fan is running at full speed. Then it would be really really loud compared to the rest of your system. And on hand your signature, it doesn't appear at first glance that your system produces so much heat that your PSU must work so hard to expell it from your case.

hmronin
Posts: 27
Joined: Sun Aug 28, 2005 11:38 pm

Post by hmronin » Sun Oct 16, 2005 9:13 am

sun.moon wrote:Hey hmronin, the S12 has a fan rpm monitor cable - on mine its a two wires, black and blue. It has only the RPM monitoring wire and ground, no power. I just hooked up to my motherboard (ASUS P4PE) and read out the rpms over the Asus PC Probe software. I am not familiar with the T-Balancer, but maybe it would be possible to read out the RPMs of the S12 over that as well?
Thanks. I downloaded the Asus PC Probe software last night, and it appears the S12 PSU fan is running at a nominal 800 RPM. Assuming this is correct, either I have a bad/noisy fan, or I'm expecting too much from a fan cooled PSU... :wink: I'm going to continue monitoring for a couple of days, to see if I can confirm these readings independently. Then I may try replacing the ADDA fan with the spare Nexus 120mm I have. I'll post results if I decide to do this.

sun.moon
Posts: 174
Joined: Wed Sep 21, 2005 5:57 am
Location: Germany

Post by sun.moon » Sun Oct 16, 2005 10:03 am

Those RPMs sound about right. I'll be watching for your post if you do decide to exchange fans! :)

Badger
Posts: 228
Joined: Thu Nov 25, 2004 8:57 am
Location: West Michigan, USA

Post by Badger » Sun Oct 23, 2005 3:56 pm

To resurrect this post:

I've confirmed that the loudest part of my system is the S12's fan. It's not loud, but when I stop it for a second with a pencil the rest of the system is dead silent.

According to Speedfan, the S12's fan runs at idle at 796 rpm. During load it'll go up to the 900's. This makes me think I could get away with either the Yate Loon D12SL-12 or the Nexus version. I have a Nexus version already on my CPU at 5v and it's inaudible. I'd like to put another inside the S12, but I think it'd be best to run it off the S12's speed controller. Is the fan controller in the S12 going to ramp up the fan to get the CFM's it needs to keep cool? Is it going to be able to start the fan on a cold boot, or won't it matter since as it needs air it'll just push the volts up anyway?

According to the charts provided above, the Adda fan in the S12, when it's doing 750 rpm's, is doing about 30 CFM @ 5v. To get 30 CFM out of the Nexus, it's going to have to be running at 790 rpm at 9.5 v.

Is it going to do the trick? Any other suggestions for getting the S12 as quiet as a Nexus 120mm at 5v?

MikeC
Site Admin
Posts: 12285
Joined: Sun Aug 11, 2002 3:26 pm
Location: Vancouver, BC, Canada
Contact:

Post by MikeC » Sun Oct 23, 2005 4:50 pm

Badger -- Now you just have to try it! :wink:

Badger
Posts: 228
Joined: Thu Nov 25, 2004 8:57 am
Location: West Michigan, USA

Post by Badger » Sun Oct 23, 2005 5:32 pm

hahaha, yes yes I suppose I should. I wouldn't expect many others to have tried it.

The last time I modded a PSU was the one that comes with the 3700BQE - I took out the stock fan, rigged a Nexus 92mm to the outside of the PSU as exhaust, and ran it at either 5 or 7 v. It wasn't pretty, but it was quiet.

One night, about 15 minutes after putting the computer in "Standby" (I had rarely done this prior) my computer exploded! Well actually just the power supply, everything else was okay (whew). My guess is that because it was in Standby there wasn't any power being drawn from the hard drives and everything else, and the 7v trick I was using on my CPU fan was feeding the current (or whatever the electricians term is) back to the PSU. Or maybe my fan swap, if you want to call it that, finally caught up to the PSU and burned out.

So... Seasonic S12-330 here I come! Pretty good, but could be quieter, like Nexus 120mm 5v quiet, and I'll be happy. :D I guess the only way to achieve that would be to use a Nexus 120 at 5v, but ehhh, I'm wary of the little airflow.

frankgehry
Posts: 1424
Joined: Sun Dec 12, 2004 5:00 am
Location: New York, NY

Post by frankgehry » Sun Oct 23, 2005 6:16 pm

Done here - http://forums.silentpcreview.com/viewto ... 888#186888
with a yate loon D12SL-12

hmronin
Posts: 27
Joined: Sun Aug 28, 2005 11:38 pm

Post by hmronin » Sun Oct 23, 2005 8:18 pm

Here's an update to my last post. Sorry for the delay, but you'll see why it took as long as it did. I did go ahead and replace the ADDA fan in my S12 330 with the Nexus 120. The procedure is fairly simple. It does require you to break the warrantee seal on the PSU cover, so be aware of that. After removing the cover, I found that the power wires for the fan are connected to a circuit board via a quick disconnect. The only line I cut were the wires for sending fan rpm info to the motherboard. Since I chose to control speed on my PSU fan with the T Balancer, it was not necessary for me to use the fan controller on the PSU. After replacing the cover to the PSU and turning on the power, here is what I found:

1. The Nexus fan is noticebly quieter than the ADDA fan. Not tremendously quieter, but there's a detectable difference. There is also a difference in the character of the noise generated by the 2 fans. The ADDA tends to have a more "scratchy" or "brushy" sound to it, while the Nexus tends to have more "whoosh" noise. Pardon my technical descriptions, but its the best I can do. :D Overall, I prefer the Nexus sound characteristic and noise level over the ADDA.

2. After disconnecting the ADDA fan from the PSU, I ran a sanity check to see how quiet the computer is with a "fanless" PSU. To my surprise, I found that the noise signature had NOT changed substantially. After some additional troubleshooting, I found the culprit of the noise: vibration coming from one of my hard drives. Even in the lower chamber hard drive case of a P180, this particular hard drive (Western Digital Cavier 120) vibrated so much that it generated an audible hum, that pulsated at regular frequencies. After removing the hard drive, this noise went away. Bottom line, the S12 PSU was NOT the prime contributor to noise from my computer, the hard drive was, or so I thought....

3. After removing the hard drive, I still noticed an audible hum emanating from the area where the computer resided. At this point, I powered down the computer to see if it was something external to the computer...and found my culprit. To make a long story shorter, I found that the source of the hum was from the main electrical distribution box, mounted on the opposite side of the exterior wall where my computer sits in my office. The electricity meters, which measure electrical consumption for my house, vibrate and buzz (confirmed this with the screwdriver handle to the ear test). This wouldn't be such a problem, except for the fact that the meters make contact to the exterior wall at several points. These points are enough to transfer the vibration from the meter to the exterior wall, which then acts as a membrane (speaker cone) and generates this hum I'm hearing on the other side of the wall! :shock:

It sounds incredible, but its true. Kind of ironic too, since I put a alot of effort into making a quiet computer and achieved that goal, only to bring the noise of the computer below the ambient noise of the surroundings. As any of you silent computing folks can attest to, this ambient hum I now hear is driving me crazy, just because I know what's causing it. The only problem now is that I'll have to have an electrician relocate the meters, as they are live. I'll probably just live with the sound until I rewire the house anyways...this will motivate me to finish up on plans for remodeling the house now! :evil:

ozdoc
Posts: 138
Joined: Tue Aug 16, 2005 4:35 am
Location: Perth, Australia

Post by ozdoc » Sun Oct 23, 2005 8:51 pm

he only problem now is that I'll have to have an electrician relocate the meters, as they are live. I'll probably just live with the sound until I rewire the house anyways...
Live with the sound!!?? The horror of it all. No, you can't get away with it that easily, having come this far on the journey.
How about some sort of sorbethane or silicone gasket to go underneath the meter, a bit like what the Antec Noisekiller pack does for PSUs?

sun.moon
Posts: 174
Joined: Wed Sep 21, 2005 5:57 am
Location: Germany

Post by sun.moon » Sun Oct 23, 2005 8:52 pm

Hi hmronin. Glad to see that the swap worked for you. I have some questions for you if you don't mind.

First and foremost, did you take pictures?
hmronin wrote:After removing the cover, I found that the power wires for the fan are connected to a circuit board via a quick disconnect.
And what es a "quick disconnect" - is that simply a three pin molex connector? So you didn't have to soder any wires?

At how many RPMs are you running the PSU fan?

Have you noticed any temp differences in your case?

After having discovered what the real culprits that contributed to noise were, would you still have replaced the PSU fan in retrospect?

Badger
Posts: 228
Joined: Thu Nov 25, 2004 8:57 am
Location: West Michigan, USA

Post by Badger » Mon Oct 24, 2005 6:45 pm

There's a Nexus 120mm in the mail coming my way. I'll post back with results, probably around the weekend.

hmronin
Posts: 27
Joined: Sun Aug 28, 2005 11:38 pm

Post by hmronin » Mon Oct 24, 2005 8:47 pm

sun.moon wrote: First and foremost, did you take pictures?

And what es a "quick disconnect" - is that simply a three pin molex connector? So you didn't have to soder any wires?

At how many RPMs are you running the PSU fan?

Have you noticed any temp differences in your case?

After having discovered what the real culprits that contributed to noise were, would you still have replaced the PSU fan in retrospect?
Good questions. Here are my answers:

1. Sorry, I didn't take pictures. At the time, I was preoccupied with making the damn noise go away, so it didn't occur to me to take pictures, though I realize now that is would have been very helpful in this discussion.

2. You're correct, my reference to a quick disconnect is a molex connector. However, if I remember correctly (should have taken pictures :cry: ), the molex on the S12 is 2 wire and 2 pin, not 3. So your standard fan molex won't work, you'll still need to strip wire if you want to use the S12 fan controller. On a positive note, the S12 will operate, even if no fan is connected to the fan controller, in the event you use your own fan controller. Since I used my T-Balancer as the fan controller, I bypassed the S12 controls completely, no soldering necessary.

3. Currently, the Nexus fan is controlled by the T-balancer, which maintains the fan 500 RPM's during idle, and is programmed to increase in speed up to maximum when it senses a rise in temperature of the CPU and GPU. At 500 RPM's, I can't hear the PSU fan from my normal working position. The case fan also is programmed the same way. My thinking with this is that during intensive CPU/GPU activity (gaming / media editing / CAD work), the slight increase noise of Nexus fans at 12V won't be noticed.

4. No temperature differences in the case, and that's using an IR Fluke temp meter. This is what I expected, as the S12 has little heat to deal with, being in the lower chamber of a P180 case, with only one hard drive in the bay.

5. OK, this is the best question of all, and my answer is.....it depends.

:wink:

If you plan to use another fan controller to control the PSU fan, I think changing the ADDA out is justifiable. For one, you can get quieter fans, like the Nexus, that will provide adequate ventilation for normal use. Two, you gain the added functionality of having more control over the PSU fan. In my case with the T-Balancer, it allows me to program all my fans to work in concert with one another, and therefore support one another, based on changing conditions inside my computer.

However, if you plan to use the S12 fan controller, I wouldn't change the fan. First, the difference in noise, although noticeable, is not great. Depending on your setup and ambient noise level, you may not even hear a difference. Two, you'd have to determine how the S12 fan controller would interact with any other fan than the ADDA, otherwise, it would be unpredictable. Too much trouble for me, for so little benefit.

In conclusion, the S12 seems to have been designed quite well, and doesn't seem to benefit from tweeks. It remains to be one of the quieter components in my setup, the loudest now being my wall!!!:shock:

Hope this helps. Maybe when others chime in on their experiences on changing the fan on the S12, we'll get a better perspective.

Post Reply