Controlling three fans with one knob?

Page 2 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

MichaelD

Lifer
Jan 16, 2001
31,528
3
76
You explained that very well, Navid. Thank you. :) I understood it perfectly. :thumbsup: Now I'm going to try to figure the values for the stock 3 fans. We'll see how I do.

*EDIT*

Please check my math.

@NAVID: I got lost at this step

Then, the current is 12V / 2 / 16.7 Ohms = 0.36A.

Why are we dividing 12V by 2? We have 3 fans...I'm confused. But using the formulas you gave, above, I did the math for the 3 stock fans that take 9.6 watts each. I bolded the parts I'm not sure of b/c of the confusion.

9.6/12=.80A current
Resistance is 12/.80=15 ohms
3 fans in parallel is 15/3=5 ohms
At 7V minimum voltage, current is 7/5 = 1.4A
Variable resistor must be 7V/1.4A=5 ohms
Resistor power rating is 12V/2/5ohm=1.2A
Max power for resistor is 6V x 1.2A = 7.2W

So I need a variable resistor that's 5 ohms and more than 7.2W. <--if the bolded parts are incorrect, then these numbers are incorrect



So, did I do that correctly?
 

DerwenArtos12

Diamond Member
Apr 7, 2003
4,278
0
0
That really just doesn't seem right. It's my night off and I'm pretty boozed but, it just doesn't look right. 5ohms is nothing. your tongue is worth about 100ohm at 12v and .2amps. Don't ask how I know that.
 

MichaelD

Lifer
Jan 16, 2001
31,528
3
76
Thanks, DA. I've done the 9-volt battery on the tongue thing way too many times myself. :laugh:

It doesn't look right to me either...I think possibly Navid typed something wrong (accidentally, I'm sure) and I'm too ignorant to know for sure.

The first hint I got that it might not be right is that I cannot find a 5 ohm pot anywhere...
 

DerwenArtos12

Diamond Member
Apr 7, 2003
4,278
0
0
Originally posted by: MichaelD
Thanks, DA. I've done the 9-volt battery on the tongue thing way too many times myself. :laugh:

It doesn't look right to me either...I think possibly Navid typed something wrong (accidentally, I'm sure) and I'm too ignorant to know for sure.

The first hint I got that it might not be right is that I cannot find a 5 ohm pot anywhere...

I think 25 ohm is the lowest I've seen, radioshack has 25ohm 3 watt pots for $3-4 and I know from experience that they're not that great of quality. Had one melt down on me I was running about 5 watts through it but, thats not that great an overload for a ceramic pot. I only buy wirewound now, it's just safer. I actually just get them through work, the last 50ohm 10watters I got were in a box of 10 for about $10. They tend to frown on that but, I've been there long enough, they just let me do as I please for the most part.
 

Navid

Diamond Member
Jul 26, 2004
5,053
0
0
Originally posted by: MichaelD
Then, the current is 12V / 2 / 16.7 Ohms = 0.36A.

Why are we dividing 12V by 2?
Because we are looking at the resistance of the three fans in parallel as a single component. Even though those are three components, but, we can treat them as a single component with a resistance equal to 1/3 of each, in this case 5 Ohms.

When you have 2 components in series (added resistor and the triple fan), the power dissipated in one as a function of its resistance is maximum when the two resistances are equal.
If the resistances are equal, we have set the resistor to 5 Ohms. So, 6V is across the fans and 6V is across the resistor. That is why we divide 12 by 2.
So, did I do that correctly?

I see nothing wrong with your math.
However, I don't know why a fan would have such a high power of 9.6W!
Do you realize that a hard drive consumes 10W? So, this fan consumes almost as much as a hard drive?
Are you sure about that figure?

Connect your Amp meter in series with the fan and apply 12V to the combination and measure the current. How much is it?

Edit:
Be careful as 0.8A is enough to cause a small spark!

Edit 2:
Anyway, you cannot practically use a pot to control just anything.
If the resistance of the load is too low, you may need an active circuit, like a transistor, to buffer the current.
Many decent fan controllers contain voltage regulators. Obviously, that is a more complicated circuit compared to a single potentiometer.
 

MichaelD

Lifer
Jan 16, 2001
31,528
3
76
It's definitely 9.6w/.8a. Not only does the label on the fan say it, but the spreadsheet for the company does too. :) The fans are 2,800rpm 120mm monsters out of a 4U server. They are beyond loud at the full 12V.

See/hear for yourself. :D

I've replaced them with 3-speed Antecs, but having to open the case to change speeds is a PITA.
 

Rubycon

Madame President
Aug 10, 2005
17,768
485
126
Why should you need to change speeds? Go with a fan of at least 50CFM that you will not hear, strap it to your 12V line and call it a day.

EDIT: BTW those fans are NOTHING compared to a regenerative blower that's used to pressure the vent channels in our amp racks. Sounds loud as hell when the music is off but when the house is rockin' no fans could ever be heard. ;)
 

MichaelD

Lifer
Jan 16, 2001
31,528
3
76
Originally posted by: Rubycon
Why should you need to change speeds? Go with a fan of at least 50CFM that you will not hear, strap it to your 12V line and call it a day.

EDIT: BTW those fans are NOTHING compared to a regenerative blower that's used to pressure the vent channels in our amp racks. Sounds loud as hell when the music is off but when the house is rockin' no fans could ever be heard. ;)

Holy crap, it's a Harrier Jump Jet taking off! :Q Vent channels? Are you using the APC modular rack system?

Anyway, I'm dealing with the realities of trying to use a server case in a home office environment. So far, so good, actually. I'm just playing with cooling ideas. I currently have the Antec 3-speed fans on LOW, which is adequate, but I'd like better. You cannot hear the Antecs over the fans in the SATA cages. So, I'm going to crank them up to MED and see how that sounds.

It's been a little warmer in my office lately (spring's coming) and I've noticed a pretty big jump in case temp in relation to just a few degrees C more in the room itself. Once summer comes (and the AC doesn't run 24/7) it's going to be 29-30C in that room.
 

Navid

Diamond Member
Jul 26, 2004
5,053
0
0
Checking your calculations again, I found an error.

You are calculating for 7V across the fan as the minimum.
You are right up to where you calculate 1.4A for current at minimum RPM.
Then, the resistor will have 5V across it (12V - 7V = 5V). So, you should divide 5 by 1.4A to get 3.6 Ohms. So, you need a 3.6-Ohm pot!



You can get a good fan controller.

If you want to do it yourself, you can get several resistors of right value and power and a switch that let's you switch between them. Then, you can change speed in 2 or 3 steps.

As an example, you can get a single pole triple throw switch and a 3.6-Ohm and a 1.66-Ohm resistor. You need to build these resistors by putting other resistors in parallel and series. You have to make sure the power rating of the resistors will not be violated.

Then, you can have 12V or 9V or 7V across the fan using the switch.
 

Rubycon

Madame President
Aug 10, 2005
17,768
485
126
Ambient has a big effect and even 1000cfm is not a cure for high ambients. You can only approach it - never match it. Asymptotics 101. ;)

The fans on medium should still be very quiet. Heck I'm willing to bet high is going to be barely noticeable. If it is, you should think about replacing those fans with Scythe fans like this one.

The racks are custom made for our Digam Powersoft K10 1U amps (10,000W). The application is unique because they have an airflow requirement and the air around here (usually in the middle of the ocean hehe) is corrosive. Those blowers are neat sounding and super reliable like most things that cost lots of pounds.
 

MichaelD

Lifer
Jan 16, 2001
31,528
3
76
Thanks for all the help, Navid. I appreciate it. If I decide to go back to the stock fans I'll probably just buy a Sunbeam rheobus now that I see what's involved. I could do it; I'm handy with the soldering iron, but the expense of the parts wouldn't be worth it, unfortunately.

Rubycon: You've got a really interesting job! :thumbsup: I had originally set the fans to LOW b/c I wasn't sure how loud the SATA cage fans were going to be and I didn't want to add noise unneccessarily. But being that the SATA fans are loud (more whiney than "whooshy loud") I'm pretty sure I can leave them on MED and not hear any diff.
 

DerwenArtos12

Diamond Member
Apr 7, 2003
4,278
0
0
+1 for the slipstreams. your application is pretty much perfect for them. I find the 1900's a little noisy though, unless on 9v or even 7v. The 1600's are near the top of what I like and also work well on 9v.

Also, depending on how much spare room you have in your case, I'd reccomend the Sunbeam Theta, got one in my HTPC and the level of control available is top notch and it comes with thermal probes so you can monitor whatever temps you want and even vary the fan speed based on those temperature read-outs. I haven't tested it to see exactly what the lowest votlage available is but, it's shut off some of the cheaper fans I've hooked up to it so I'm thinking it bows all the way to 5v. It does require an external USB and an internal molex though, kind of a pain. All the adjustments are made through their program which is pretty lightweight and intuitive though not top noth on either category, I'd say 7/10 for leightweight and 8.5/10 for intuitive. Main reason I suggest it is that it's completely OS operated so no need for finding nobs, either on the front of back of the system so adjustments can be made of the fly, without having to fumble around your case.