How to slow down Coolermaster 380W fan

chazza · Post by **chazza** » Mon May 29, 2006 9:01 pm

The fan in my Coolermaster 380W is variable based on the PSU temp. It starts of quietly in the mornings and gets louder after 10-15 minutes. Working the CPU/GPU hard causes it to get quite loud. I already have a case exhaust fan.

Is it really necessary for a PSU fan to work so hard?

I want to mod it somehow to slow it down. I don't want to replace the entire PSU just because of the fan.

I'm thinking either:

1. Build a ramp to let it breathe in cooler air (dont think this will make a enough difference as the case temp is quite low).

2. Cut off the fan grills on the back. Will this make a big difference?

3. Open the PSU carefully and add a resistor (about 30 ohms?) to slow down the fan.

Any other suggestions?

P.S. Is there any way I can measure the PSU fan speed or temp?

thegoldenstrand · Post by **thegoldenstrand** » Tue May 30, 2006 4:37 pm

Is there anyway to tweak the temp settings? I would like to have larger fan running slower.. I am looking at the toughpower from Thermaltake because it has that 14 cm fan.. it is expected to rev up to 1900 rpm at load.. I would much rather have it slowed down slightly..

chazza · Post by **chazza** » Tue May 30, 2006 6:28 pm

Thanks for replying but there is no way to tweak anything.

TomZ · Post by **TomZ** » Tue May 30, 2006 6:44 pm

I would guess that, if you are reporting it gets loud after 10-15 minutes, then the internal case temperature has probably risen quite a bit and the PSU is now taking in rather warm air. Maybe the internal case temperature is higher than you think in that condition?

I would suggest the following experiment. Run your computer under the same conditions, with the case open, and put some cardboard or something similar in to let it draw in fresh air instead of the air from within your case. This is an easy way to tell whether creating a bypass duct will help out.

I wouldn't suggest putting in a resistor. The PSU designers probably pretty carefully considered temperature/load/CFM relationships, so I wouldn't mess with that. Another option might be to replace the fan with a quieter, similar airflow model.

Other option is to replace the PSU with something inherently quieter, e.g., Seasonic S12.

chazza · Post by **chazza** » Tue May 30, 2006 9:11 pm

TomZ wrote:I would guess that, if you are reporting it gets loud after 10-15 minutes, then the internal case temperature has probably risen quite a bit and the PSU is now taking in rather warm air. Maybe the internal case temperature is higher than you think in that condition?

I would suggest the following experiment. Run your computer under the same conditions, with the case open, and put some cardboard or something similar in to let it draw in fresh air instead of the air from within your case. This is an easy way to tell whether creating a bypass duct will help out.

That's a good idea Tom. I will try that out tonight.

Although I have already tried keeping the side of the case open and even laying the case on its side so as to prevent the hot air rising to the PSU and it did not make a noticeable difference.

Edit: What about cutting off the fan grill at the back? Will that help reduce some white noise or is this usually insignificant.

disphenoidal · Post by **disphenoidal** » Tue May 30, 2006 11:48 pm

Instead of a resistor, install a diode or two in series. Each one will drop the voltage the fan recieves by .7V.

chazza · Post by **chazza** » Wed May 31, 2006 5:00 am

Well I managed to open the PSU and inspect it (and survived to tell the story).

The fan is a brushless ADDA AD0812HS-A70GL (See Google: http://www.google.com/search?q=AD0812HS-A70GL)

Rated at 12V 0.25 A and I think 36 cfm.

More importantly I found the temp sensor stuck to the heatsink (I couldnt pry it off safely) and I've inspected the way the air flows through the PSU so maybe I can come up with a trick to keep the fan speed down (I'll try it tomorrow).

SebRad · Post by **SebRad** » Wed May 31, 2006 5:41 am

Hi, another option is to replace the fan with a quieter/slower one. You PSU appears to have 3000rpm (@12V) fan and mine had 2400RPM ADDA fan that I've swapped for a Panaflo 80L1A. I think the Panaflo is rated for 1900 rpm so it's about 80% of the original fan's speed. I'm not worried about overheating my PSU for 2 reasons. Firstly it's ducted (to blow hole as intake) so it's getting cool air and secondly the new fan is connected to the PSU's fan controller so if things get really hot the PSU can still ramp the fan up. I guess it runs 5-6V all the time in my system. see here.
What's the spec of your system? As long as it's not a power guzzling monster and your case temp are reasonable I would think you could safely use a slower fan, Yate Loon D80SL-12 (2000rpm) would be good candidate or a Panaflo like mine or any other equivilent fan that runs smoothly/quietly. I would advise against a Nexus 80mm as I think it's too slow running.
Good luck, Seb

Le_Gritche · Post by **Le_Gritche** » Wed May 31, 2006 6:34 am

chazza wrote:The fan in my Coolermaster 380W is variable based on the PSU temp. It starts of quietly in the mornings and gets louder after 10-15 minutes. Working the CPU/GPU hard causes it to get quite loud.

The fan gets louder because the temp sensor is subjected to a hotter temperature, then in turns raises the fan voltage.
You can either act on the fan or the sensor.
The sensor is heated by the heatsink and by the airflow :
- do a fresh-air duct for the PSU intake
- dump the heat from CPU+GPU outside by ducting them to the case fan
- put the sensor farther apart from the heatsink (you should try again, that's very efficient particularly if the exhausted air is just mildly warm, indicating room for reducing the fan speed safely)
- reduce the heat radiated by the heatsink (by buying a more efficient PSU or by undervolting your CPU and GPU if possible)

The fan :
- improve the airflow by cutting the fan grills on the back (as you said)
- lower the restrectiveness of the intake (cut the intake grills and have a case intake big enough to accomodate for both the PSU and case fans exhaust airflow)
- swap the fan with a quiter/slower one
- solder a resistor, diode or potentiometer on the fan power wire (the fan speed was set with safety margins in mind to accomodate for different ambient and case setup as well as safety, you can get rid of them if your own setup allows it.)

P.S. Is there any way I can measure the PSU fan speed or temp?

Following one of the link in the google search leads to 3010 rpm, 34.4 dB and 38.6 CFM. That's at 12V.
For the real speed you can either mesure the real voltage or try to guess the frequency of the noise with that website (it will work with the motor noise, not with the broadband airflow noise obviously)

In my personnal experience, insuring that the PSU get fresh air was efficient, but less than moving the sensor (my far less power hungry PC than yours allowed a big gap between sensor and heatsink). Cuting the fan grills helped a bit.
At this stage undervolting the CPU and underclocking the GPU helped very little with the noise, because with the sensor moved away from the heatsink, the fan speed was quite independant from the power consumption. Nevertheless it helped my piece of mind by lowering the exhaust air temperature.
I'm still not happy with the noise (mainly worn out bearing noise now) so short from a fan swap, I'm considering soldering a potentiometer to allow for further tweaking according to the ambient temp (summer/winter) and the power usage (web browsing/gaming with all overclocked).

Cuting grills, moving the sensor, doing an air duct, swapping the fan, reducing the power consumption are all usefull, so do as many as you can between them.

So what do you mean by 'stuck' for the temp sensor : Not enough space to slip a blade, glued to the heatsink, or no moving allowance ?

chazza · Post by **chazza** » Thu Jun 01, 2006 3:17 am

I've cut out the grills at the back. This did not make any noticeable difference to noise. I regret doing it as bugs might creep in.

I've improved the airflow by blocking a useless vent on the side of the PSU using sticky tape. Now more air is coming through the back vent (where the heatsink is) and this has reduced the fan speed noticeably!

It is still loud overall though compared to my other fans. I don't know what to do now.

I couldnt get the temp sensor off because it was not only glued, but there was a plastic clip over it which was screwed into the heatsink. I couldnt unscrew it as there was a transformer thingy (a ring with lots of wires around it) in the way.

chazza · Post by **chazza** » Fri Jun 02, 2006 2:46 am

More good news. I've suspended the hard drive and built a PSU air duct.

I've suspended the Samsung hard drive in the 5.25" bays using elastic string. The HDD idle noise is gone and I can barely hear seeks now. I'm very happy with that.

I also built a air duct for the PSU to pull air from the 5.25" bays. The 5.25" bays on my Centurion 5 case can are well ventilated due to the front having lots of tiny holes. I think this has drastically reduced the noise of the PSU when idling.

I'm very happy with the results now. Maybe when summer comes in Australia it will get louder again, but for now it's great.

The HDD suspension idea and PSU air duct idea both came from reading this forum. Thanks guys.

Le_Gritche · Post by **Le_Gritche** » Fri Jun 02, 2006 3:05 am

chazza wrote:It is still loud overall though compared to my other fans. I don't know what to do now.

What kind of noise is it : a wind noise, caused by the airflow, or a mechanical noise caused by the fan's motor ?
Depending on that you can either reduce the airflow and improve the airpath or swap the fan.

Is the exhausted air hot ?
If not you have room to further slow the fan (by soldering a resistor or swapping the fan for a slower and quieter one, as you can't act on the sensor)
If it is hot you will have to focus on reducing the power consumption and/or providing fresh air to the intake.
There are other mods, as adding heatsinks, puting them out of the PSU casing, or maybe even crazier mods but I can't help you with that.

Blocking a PSU vent to direct the airflow on the heatsinks might reduce the intake openings to a point where the fan will be fighting with a resistive airpath and creating negative pressure.
You say it reduced the fan speed noticeably; I hope it's because it improved the cooling by redirecting the airflow, not because the fan can't suck enough air, because these resistive airpaths create noise (maybe the same kind of cavitating noise happening with water propellers)

chazza · Post by **chazza** » Fri Jun 02, 2006 4:00 am

It is a wind noise, not a motor noise. The PSU exhaust air is cool so there is room to slow down the fan. I may add some resistors/diodes in future.

The intake vent on the PSU shell is pretty big, so I assume there is not too much negative pressure.

How to slow down Coolermaster 380W fan

How to slow down Coolermaster 380W fan

I would like to know more about this too!!

Re: How to slow down Coolermaster 380W fan