7 volt trick?

Control: management of fans, temp/rpm monitoring via soft/hardware

Moderators: NeilBlanchard, Ralf Hutter, sthayashi, Lawrence Lee

Post Reply
hjm
Posts: 12
Joined: Sat Dec 30, 2006 11:48 am
Location: Finland

7 volt trick?

Post by hjm » Mon Jan 15, 2007 11:27 pm

Being adviced by SPCR article http://www.silentpcreview.com/article6-page1.html I thought I would like to try the 7 volt trick for Nexus to make it completely silent while still blowing same amount of air that the TriCool is now doing.

This article discusses about black and red fan wires. But my fan has three wires, black, red and yellow. So what's that yellow wire for? Do I need to connect that also to somewhere?

I just bought this little thingy, it's Zalman Multi Connector 1 (or ZM-MC1).

Image

The black 3-pin connector gives 12V and the white gives 5V. Now, I was thinking to mod one of the black 12V connectors to become 7V by reconnecting the black wire to red 5V wire output of the Molex connector. And then connect the Nexus straigtly to that 3-pin connector.

I just wanted to double-check this would work.

Also, I should remember to have some other loads on the 5V of the same cable to avoid damaging the PSU. Connecting the hard drives to the same cable should do it I believe. Also, I have connected the Scythe Ninja's fan to one of the 5V outputs of the ZM-MC1 so it's also helping there. Just to be sure.

Any comments?

Devonavar
SPCR Reviewer
Posts: 1850
Joined: Sun Sep 21, 2003 11:23 am
Location: Vancouver, BC, Canada

Post by Devonavar » Tue Jan 16, 2007 1:49 am

Yes, that should work. The third (yellow) wire on the fan header is an RPM sensor cable; it doesn't need to be hooked up, but you do lose RPM monitoring by not running it off the motherboard.

Bluefront
*Lifetime Patron*
Posts: 5316
Joined: Sat Jan 18, 2003 2:19 pm
Location: St Louis (county) Missouri USA

Post by Bluefront » Tue Jan 16, 2007 3:22 am

Since you're splicing wires.....take the yellow wire from the fan, and connect it by itself to the MB. You'll be able to read rpms then, but the power for the fan will be coming directly from the PSU.

FWIW.....I do not recommend that 7v mod. It's electrically questionable. Fanmates are cheap, and can be dialed in to whatever voltage you want.
[size=75]"At the core of liberalism is the spoiled child - miserable, as all spoiled children are, unsatisfied, demanding, ill disciplined, despotic, and useless. Liberalism is the philosophy of sniveling brats." - P.J. O'Rourke[/size]

cmthomson
Posts: 1266
Joined: Sun Oct 09, 2005 8:35 am
Location: Pleasanton, CA

Post by cmthomson » Tue Jan 16, 2007 5:27 pm

RPM sensing will not work at all with the "7V trick", even if you connect the third wire to the motherboard header. This is because the sensor circuit compares the tach input signal to ground, but the fan is running 7V above ground, and the speed always reads as zero.

For about $10, there are several ways to slow down a Nexus. One is a Fanmate, another is an inline resistor, such as the Zalman RC56.
i7 4790K CPU@4.6 GHz, ASUS Z97-PRO, 16GB G.Skill 2400C10, Intel 335 240GB SSD + WDC EFRX 1TB, Internal i7 graphics, Antec P180 case, Seasonic X-400 fanless PS, Megahalems CPU HS, Nexus 3-pin & AC PWM fans < 600 RPM, AcoustiPack foam, homemade ducts.

hjm
Posts: 12
Joined: Sat Dec 30, 2006 11:48 am
Location: Finland

Post by hjm » Thu Jan 18, 2007 4:27 am

cmthomson wrote:RPM sensing will not work at all with the "7V trick", even if you connect the third wire to the motherboard header. This is because the sensor circuit compares the tach input signal to ground, but the fan is running 7V above ground, and the speed always reads as zero.

For about $10, there are several ways to slow down a Nexus. One is a Fanmate, another is an inline resistor, such as the Zalman RC56.
I can't remember too much school physics anymore, but I'm not a solder-shy, so I could actually mod my ZM-MC1 to have a resistor.

Should the resistor be put on the yellow wire of the 12V connector or?

But how to measure the resistor? E.g. for running the Nexus at 7V or 9V (should be still silent enough while blowing bit more air)?

Qwertyiopisme
Posts: 237
Joined: Tue Nov 11, 2003 6:48 am
Location: Gothenburg, Sweden
Contact:

Post by Qwertyiopisme » Thu Jan 18, 2007 7:00 am

In general it's better practice to put the resistor on the positive wire, though in this case it doesn't really matter.

The resistance can be found with ohms law, there the resistance is V/I, where V is the voltage lost over the resistor and I is the current though the resistor/fan at that speed, so if for instance you want to run the fan at 7V, you'll have 5V over the resistor (12V total, 7V over fan, leaves 5V over the resistor). The current will probably be something like 60% of fullspeed current (I'm assuming that current is proportional to voltage, afaik it should (nearly) be proportional), say for instance that it's 60mA, and since R=V/I => 7/0.06 = 115 ohms, or thereabouts.

In order to determine the power rating of the resistor you can calculate the power in the resistor by the fact that P=R*I^2, which in this case is 115*0.06^2 = 0.4W, so try to get something at least for 0.5W, prefferably 1W or above, as when running near their rated power they get very, very, very hot.
[size=75][color=purple][list=1]
[*]3800+ (X2), 2gb ram, 1390Gb storage, complete homemade watercooling
[*]3000+ (Winchester), 1gb ram, 120Gb storage, home-made chassis[/list][/color][/size]

hjm
Posts: 12
Joined: Sat Dec 30, 2006 11:48 am
Location: Finland

Post by hjm » Fri Jan 19, 2007 8:47 am

I did the 7 volt trick for Nexus, not changing the wirings of ZM-MC1, but Nexus' itself. It was far easier that way, just swap the 5V and GND wires between the male and female Molex connectors that are connected with Nexus. What I did is exactly as shown at http://www.cpemma.co.uk/7volt.html. The good thing with this one is that the Molex itself is not changed (despite the swapped wires), and I could actually piggy-back the ZM-MC1 to the Nexus just normally, and have the stock fan of Ninja @5V consuming the "5V leftovers" from Nexus. I also had hard disk and optical drive connected to the same power cable so I assume everything was safe from PSU point of view.

And Nexus became totally silent. Ninja's stock fan @5V is silent, too. So noisewise perfect. But, the airflows came down too. System idle temps rose about 8C, being 44C.

So, after all, 7V was maybe too little. Or what do you think?

Next try is either get a fanmate or start playing with resistors. Or just stick with 12V despite it's bit noisy, just only a bit. Luckily the noise character of Nexus is very smooth, so it's not too disturbing.

cmthomson
Posts: 1266
Joined: Sun Oct 09, 2005 8:35 am
Location: Pleasanton, CA

Post by cmthomson » Fri Jan 19, 2007 10:19 am

Back before I had a motherboard that supported SpeedFan, I used inline resistors to slow down Nexus 120mm (orange) fans.

Here are some free-air measurements I kept notes on:
12V 1150 RPM
56 ohms 8.9V 915 RPM (this one doesn't agree with the others...)
78 ohms 7.9V 815 RPM
103 ohms 7.1V 780 RPM

From this you can conclude that the RPM is roughly proportional to the supplied voltage, and that the resistance of this particular fan sample was about 150 ohms.

So, if you want about 7V, use a 100 ohm resistor. For 6V, use 150 ohms.
i7 4790K CPU@4.6 GHz, ASUS Z97-PRO, 16GB G.Skill 2400C10, Intel 335 240GB SSD + WDC EFRX 1TB, Internal i7 graphics, Antec P180 case, Seasonic X-400 fanless PS, Megahalems CPU HS, Nexus 3-pin & AC PWM fans < 600 RPM, AcoustiPack foam, homemade ducts.

bags
Posts: 2
Joined: Thu Feb 01, 2007 1:57 pm

Post by bags » Thu Feb 01, 2007 2:00 pm

cmthomson wrote:Back before I had a motherboard that supported SpeedFan, I used inline resistors to slow down Nexus 120mm (orange) fans.

Here are some free-air measurements I kept notes on:
12V 1150 RPM
56 ohms 8.9V 915 RPM (this one doesn't agree with the others...)
78 ohms 7.9V 815 RPM
103 ohms 7.1V 780 RPM

From this you can conclude that the RPM is roughly proportional to the supplied voltage, and that the resistance of this particular fan sample was about 150 ohms.

So, if you want about 7V, use a 100 ohm resistor. For 6V, use 150 ohms.
thanks for the great info..

I wanted to undervolt my fans but i'm not comfortable with the standard 7V mod wire switcheroo...

so i followed your specs...
coupla noob questions tho...

can you let me know if i picked up the right resistor?

"100-Ohm"
"1/2 Watt"
"5% tolerance"

and if that is correct, are resistors directional?

btw - these are for 120mm Yate Loon Fans...

bags
Posts: 2
Joined: Thu Feb 01, 2007 1:57 pm

Post by bags » Thu Feb 01, 2007 3:31 pm

hmmm... after using some voltage calculators I found that a 12V fan with a 0.3A current actually needs like a 24 - 28 ohm resistor to get it down to 7-7.5V...

if i'm using these calculators correctly, a 100-ohm resistor as you mentioned, on a 12V / 0.3A fan would bring it down to like 3.5V

star882
Posts: 27
Joined: Fri Jan 26, 2007 1:15 pm

Post by star882 » Thu Feb 01, 2007 9:31 pm

cmthomson wrote:the resistance of this particular fan sample was about 150 ohms.
DC fan motors are more or less AC induction motors with oscillators built in (except for ECM fans but those are generally made for high power instead of low noise). As such, decreasing the voltage increases "slip", decreasing efficiency and causing the motor to heat up. The oscillator frequency will decrease to counteract this, so the current went down. (Use a small, high value inductor connected to a sound card to pick up the frequency of the oscillator.) However, this doesn't happen with every fan (although it's true for most of them), so you might encounter a fan that draws more current with less voltage. Just something to keep in mind when dealing with a fan...

Note that ECM fans often dislike low voltages, but some can be put into a "voltage governed" mode. I have a 120mm ECM fan in my Athlon 64 system. It developed a lot of static pressure (and as many would have guessed, was quite noisy!) at 12v. By wiring the control wires to set the fan to voltage governed mode and installing a 12 ohm, 5w resistor to reduce the voltage to 7v, it actually became quieter than the dual core P4 server (Dimension 4600) in the same room, yet it still moved more air than the regular fan in that server.

cmthomson
Posts: 1266
Joined: Sun Oct 09, 2005 8:35 am
Location: Pleasanton, CA

Post by cmthomson » Fri Feb 02, 2007 7:03 am

bags wrote:so i followed your specs...
coupla noob questions tho...

can you let me know if i picked up the right resistor?

"100-Ohm"
"1/2 Watt"
"5% tolerance"

and if that is correct, are resistors directional?

btw - these are for 120mm Yate Loon Fans...
Those specs are fine, although a 1W resistor will stay cooler. Resistors are not directional.

The Nexus is a Yate Loon variant with a lower RPM controller. As far as I can tell, the motor is the same. This means that the effective resistance of the Yate Loon will be lower than that of the Nexus, so you should use somewhat lower resistor values. I've never measured a Yate Loon, so I can't be any more specific, sorry.
i7 4790K CPU@4.6 GHz, ASUS Z97-PRO, 16GB G.Skill 2400C10, Intel 335 240GB SSD + WDC EFRX 1TB, Internal i7 graphics, Antec P180 case, Seasonic X-400 fanless PS, Megahalems CPU HS, Nexus 3-pin & AC PWM fans < 600 RPM, AcoustiPack foam, homemade ducts.

cmthomson
Posts: 1266
Joined: Sun Oct 09, 2005 8:35 am
Location: Pleasanton, CA

Post by cmthomson » Fri Feb 02, 2007 7:04 am

bags wrote:hmmm... after using some voltage calculators I found that a 12V fan with a 0.3A current actually needs like a 24 - 28 ohm resistor to get it down to 7-7.5V...

if i'm using these calculators correctly, a 100-ohm resistor as you mentioned, on a 12V / 0.3A fan would bring it down to like 3.5V
The Nexus is also rated at 0.30A, but it does not actually draw anywhere near that much once it is running. Fans draw more current during startup than during normal operation.
i7 4790K CPU@4.6 GHz, ASUS Z97-PRO, 16GB G.Skill 2400C10, Intel 335 240GB SSD + WDC EFRX 1TB, Internal i7 graphics, Antec P180 case, Seasonic X-400 fanless PS, Megahalems CPU HS, Nexus 3-pin & AC PWM fans < 600 RPM, AcoustiPack foam, homemade ducts.

cmthomson
Posts: 1266
Joined: Sun Oct 09, 2005 8:35 am
Location: Pleasanton, CA

Post by cmthomson » Fri Feb 02, 2007 7:14 am

star882 wrote:
cmthomson wrote:the resistance of this particular fan sample was about 150 ohms.
DC fan motors are more or less AC induction motors with oscillators built in ...
Yep. It is an oversimplification to say the fan has X ohms resistance, but for practical purposes, when running it with an inline resistor to reduce the RPM, it is fairly accurate to treat the fan as a constant resistive load. Of course, this discussion is restricted to 12V PC case fans.

I only measured two fan types: the Nexus orange Real Silent 120mm, and the AcoustiFan DustProof 120mm. The Nexus had an effective resistance of 150 ohms, while the higher-RPM AFDP was 92 ohms.
i7 4790K CPU@4.6 GHz, ASUS Z97-PRO, 16GB G.Skill 2400C10, Intel 335 240GB SSD + WDC EFRX 1TB, Internal i7 graphics, Antec P180 case, Seasonic X-400 fanless PS, Megahalems CPU HS, Nexus 3-pin & AC PWM fans < 600 RPM, AcoustiPack foam, homemade ducts.

Felger Carbon
Posts: 2049
Joined: Thu Dec 15, 2005 11:06 am
Location: Klamath Falls, OR

Post by Felger Carbon » Fri Feb 02, 2007 8:58 am

cmthomson wrote:...for practical purposes, when running it with an inline resistor to reduce the RPM, it is fairly accurate to treat the fan as a constant resistive load.
It is constant at a particular RPM, but the equivalent resistance drops as the RPM increases. The "resistance" of a fan is least at 12V, and is maximum at the minimum running voltage. For 12V fans, of course.

The reason for this variance is that the power required to push the air varies with the cube of the RPM, not the square.

cpemma
Posts: 351
Joined: Sun May 11, 2003 2:31 pm
Location: S Yorks, OK
Contact:

Post by cpemma » Mon May 21, 2007 11:42 am

Felger Carbon wrote:It is constant at a particular RPM, but the equivalent resistance drops as the RPM increases. The "resistance" of a fan is least at 12V, and is maximum at the minimum running voltage. For 12V fans, of course.
I get the opposite, doing measurements on an 80mm Panaflo L1A. Columns are volts, mA and V/I=impedance.

Code: Select all

   5.30   36.4   145
	5.32	36.6	145
	5.63	39.8	141
	6.32	43.8	144
	6.57	44.5	148
	7.23	47.1	153
	7.78	49.3	158
	7.99	50.0	160
	8.88	53.6	166
	8.97	54.1	166
	9.91	58.4	170
	9.99	58.0	172
	10.79	61.1	177
	11.12	63.8	174
	11.46	65.0	176
	11.48	65.2	176
However, a Vantec Stealth 80mm measured as almost constant impedance.

Code: Select all

	5.32	31.7	168
	5.70	33.6	170
	6.20	36.2	171
	7.13	41.3	173
	7.67	44.2	173
	8.57	49.3	174
	8.71	49.8	175
	9.44	54.7	172
	9.58	55.0	174
	9.95	57.7	172
	10.13	58.6	173
	10.54	61.4	172
	11.28	65.7	172
	11.41	67.3	169
	11.42	66.6	171
Must check some 120mm. :?

Post Reply