Which fan speed reduction cable should I get?

[code65536]

Background:

I recently bought an In Win K1 case. It's really great, except for one problem: the included fan is louder than I'd like.

It's a PWM fan, but the lowest it will go is around 2200RPM (which is way overkill for the Ivy Bridge Celeron that I have in there, based on the temps that I see), there isn't an option to control the fan speed in the BIOS (aside from turning off the speed control and having it run at the maximum RPM all the time), and none of the various fan control software that I've tried has worked.

I'd just like to get it to run slower, at around 1500RPM or so when the system is idle. It's been suggested that I get one of those fan speed reduction cables, with a resistor in the middle, but I'm new to all this, and I was never any good at electrical engineering, so I'm not sure which of the many options I should get.

My questions:

The fan in question is a 4-pin PWM fan that says 12V, 0.4A on the label.

I did some reading about this, and based on my calculations (which may very well be totally wrong), I need something with a 15 ohm resistor capable of dissipating 1W.

1) Are my figures correct? Or am I completely off-base? And does this fan being PWM change anything?

2) When I go looking for these speed reduction cables, the lowest I could find is around 22 ohm. Many are much, much higher, like 30 ohm or 47 ohm. I couldn't find one in the ballpark of 15 ohm, which makes me suspect that my calculations are wrong. And then there are a few that list only a percentage reduction (e.g., 30% reduction or 15% reduction), but if the reduction is dependent on the fan's rating, wouldn't those percentages vary based on what fan you use? And none of them list how much the resistor is capable of dissipating.


Anyway, I'm considering getting these. They're 23 ohm, and I guess if it turns out I need higher, I can just daisy-chain them. So... am I about to make a horrible mistake here?

 

[BonzaiDuck]

Background:

I recently bought an In Win K1 case. It's really great, except for one problem: the included fan is louder than I'd like.

It's a PWM fan, but the lowest it will go is around 2200RPM (which is way overkill for the Ivy Bridge Celeron that I have in there, based on the temps that I see), there isn't an option to control the fan speed in the BIOS (aside from turning off the speed control and having it run at the maximum RPM all the time), and none of the various fan control software that I've tried has worked.

I'd just like to get it to run slower, at around 1500RPM or so when the system is idle. It's been suggested that I get one of those fan speed reduction cables, with a resistor in the middle, but I'm new to all this, and I was never any good at electrical engineering, so I'm not sure which of the many options I should get.

My questions:

The fan in question is a 4-pin PWM fan that says 12V, 0.4A on the label.

I did some reading about this, and based on my calculations (which may very well be totally wrong), I need something with a 15 ohm resistor capable of dissipating 1W.

1) Are my figures correct? Or am I completely off-base? And does this fan being PWM change anything?

2) When I go looking for these speed reduction cables, the lowest I could find is around 22 ohm. Many are much, much higher, like 30 ohm or 47 ohm. I couldn't find one in the ballpark of 15 ohm, which makes me suspect that my calculations are wrong. And then there are a few that list only a percentage reduction (e.g., 30% reduction or 15% reduction), but if the reduction is dependent on the fan's rating, wouldn't those percentages vary based on what fan you use? And none of them list how much the resistor is capable of dissipating.


Anyway, I'm considering getting these. They're 23 ohm, and I guess if it turns out I need higher, I can just daisy-chain them. So... am I about to make a horrible mistake here?

You mention that your processor is an Ivy Bridge Celeron, but you don't mention the motherboard. So that's a question.

Another one: Is this fan which troubles you installed on the case rear exhaust port?

Without being able to look at the mobo, I would still bet that the CPU_FAN header (some have a second CPU_OPT_FAN PWM) is PWM. After all -- it has to be a socket 1155 board . . . isn't that right?

Before you go off to find your solution in wiring so as to emasculate the fan's top-end speed, do yourself a favor and inventory some things:

1) BIOS should have a fan control feature which supports thermal control of fan speed.
2) Mobo manufacturer should provide software for Windows allowing you to define a "fan-curve" for the CPU_FAN header (at least) -- and just as likely or often -- CHA_FAN (chassis-fan) control.
3) Even if there is only one PWM header for CPU_FAN, you can use that header to control up to eight 12V fans/pumps of varying amperage and top-end RPM without powering any of them from the motherboard.

Check out the ~$9.00 Swiftech 8W-PWM-SPL-ST device at any of several resellers, but I guarantee you'll find it at www.sidewindercomputers.com

All the PWM devices connected to the Swiftech are powered directly from your PSU, but the duty-cycle or fan speed is determined by the CPU_FAN PWM signal wire.

 

[code65536]

I forgot to mention, the case came bundled with a motherboard. It's a Gigabyte MSH61DI, which Gigabyte lists on their website as an "embedded" board.

And perhaps I was unclear with the word of my post, but the board does support PWM for the CPU fan. It can go as fast as something over 4000RPM.

The problem is that the board will let the fan go is 2200RPM, which is pretty overkill for this setup. As I mentioned, there is no fan speed control in the BIOS settings (I know most boards expose these settings, but this particular board does not), and the board does not appear to expose any means for software to control it. Probably because this is an embedded board and not a more mainstream product.

Anyway, without a way to control the speed via software or BIOS, a resistor seems to be my only option.

 

[BonzaiDuck]

I forgot to mention, the case came bundled with a motherboard. It's a Gigabyte MSH61DI, which Gigabyte lists on their website as an "embedded" board.

And perhaps I was unclear with the word of my post, but the board does support PWM for the CPU fan. It can go as fast as something over 4000RPM.

The problem is that the board will let the fan go is 2200RPM, which is pretty overkill for this setup. As I mentioned, there is no fan speed control in the BIOS settings (I know most boards expose these settings, but this particular board does not), and the board does not appear to expose any means for software to control it. Probably because this is an embedded board and not a more mainstream product.

Anyway, without a way to control the speed via software or BIOS, a resistor seems to be my only option.

Looks as though native mobo thermal control isn't part of your options. The remaining options are either expensive -- as with a special fan-controller and thermal sensors -- or simpler and "less dynamic" -- as with low-speed fans running at their top-end.

People still use Almico's SpeedFan, as did I several years ago to control the CPU fan. Just not my favorite choice for this problem, but it may be the only one providing thermal speed control. And now I'm wondering how it might work with a splitter device like the Swiftech.

Or -- there's the idea you have for modding the hardware wiring. Not my favorite choice, either . . . Wish I could offer more ideas . . .

 

[code65536]

I ended up ordering the cables shortly after starting this thread, and I got them a few days ago. They work reasonably well.

My calculations were wrong, though. A single 23 ohm resistor brought the minimum fan speed from 2200 down to around 1850, and two of these resistors in series took it down to 1700, which is good enough for now. If I turn off the PWM in the BIOS and let it run at full tilt, it's down to around 2700 instead of something north of 4000. So it does what I want--shaving a percentage off at every speed setting. They are barely warmer than ambient when when running at full speed, so the resistor overheating is definitely not a concern.

The cooling for this thing definitely was overkill for an Ivy Celeron--even maxing out both the CPU and GPU at the same time in a stress test, the CPU temps reached only 40C (room was round 21C), and the fan remained at 1700RPM the entire time.

 

[BonzaiDuck]

I ended up ordering the cables shortly after starting this thread, and I got them a few days ago. They work reasonably well.

My calculations were wrong, though. A single 23 ohm resistor brought the minimum fan speed from 2200 down to around 1850, and two of these resistors in series took it down to 1700, which is good enough for now. If I turn off the PWM in the BIOS and let it run at full tilt, it's down to around 2700 instead of something north of 4000. So it does what I want--shaving a percentage off at every speed setting. They are barely warmer than ambient when when running at full speed, so the resistor overheating is definitely not a concern.

The cooling for this thing definitely was overkill for an Ivy Celeron--even maxing out both the CPU and GPU at the same time in a stress test, the CPU temps reached only 40C (room was round 21C), and the fan remained at 1700RPM the entire time.

Looks as though you've worked out a solution.