Computer Power Supply and Switch Wiring Questions
 

Hey guys, ever since I got my new Antec P182 case, I've been itching to make this baby silent.

The 3 Tri-Cool fans I have in the back and front can get pretty loud at times when it's simply idling... so I want to make a switch.

So....

Here's my plan, I want to make a simple SPST switch for each and every fan, and since the fans are simply just + and - wires, wiring up a switch should be fairly easy. So I want to grab a few connectors, and wire the switches, however, I'm unsure of what and how much of a current the switches would need.

I know that the fans are running are probably running at 12V (max), and so most SPST switches are 125VAC rated. (Which I assume would be more than enough), however should I be worried about the current?

Also, as far as the wiring, would it be as simple as splicing the positive wire to one of switch's terminal, and another wire from the other switch terminal to the final positive wire?

You should be fine as long as the fans dont get full of dust putting lots of load on the motors. Also yes just cut the red wire and connect each end to each pole on the switch.

However why not get a fan controller like this: http://www.newegg.com/Product/Produc...-808-_-Product

I was looking at fan controllers, however, my current fans, Antec Tri-Cools, are controlled through a little switch which is already connected to the fan.
The only power it gets is from the molex.
Some have modded them for the RPM regulator, however it's more trouble than it's worth.
Plus, fan controllers don't have an "off" switch :p
Well, some do, but...some don't


I was also thinking about picking up some Rosewills that are RPM controlled, variable (I had a blue 120 before, and it was freaking awesome) which also come with their own PCI controller.
I was thinking of just cutting the PCI bracket up, and mounting it on a 5.25" bay drive cover... But, well, money doesn't grow on trees!

Got it, so any switch should be fine, as long as it's rated over 12V?

Any fan should have the operating voltage and current on the center hub sticker. Current draws for normal speed fans is around 0.25A or less. Those really loud and fast "vacuum cleaner" fans can take ~0.5A. Either way, any switch capable of 1A is more than enough to control 1 fan. RadioShack has a micro switch which is rated for 6A; I use these because I like the stiffness of the lever actuation (no comments linc!).

If you want a rudimentary speed control, simply use a SPDT "center off" (3 position) switch. Then use 2 or 3 diodes in series when the switch is thrown one way (reduces the voltage to the fan by 0.7v per diode), and direct power when thrown the other way. When in the middle, the fan is off. The "circuit" is really simple, but I can throw together a schematic if you need...

And to answer your question; yes, a 125v switch will work fine. The volt rating just tells you that the voltage will not "arc" between the contacts.

So... I guess this is overkill?
http://www.radioshack.com/product/in...ductId=2062529
:intello:

I'm digging that one because it matches my case theme, black. :yes:
So anything that can handle more than 1A, and 12V, should be perfectly fine?

Wow, a circuit is that simple? Yes please on the schematic!
I might just make one for kicks!

How does the 125V rating tell you that it doesn't arc?

I'm a huge noob.

lol, 25A is certainly overkill, but wouldn't hurt anything. I think you'll find it quite large. The toggle lever is about an inch or so long, and the switch case is rather large. But if you have room, use it.

This is a simple drawing of how to arrange the diodes:

http://scriptasylum.com/forumpics/fanspeed.gif

This type of switch is what you need (SPDT center off): http://www.radioshack.com/product/in...uctId=2062486#

If you don't want to be able to turn the fans off at all, don't get the "center off" type (you can only toggle left or right).

If you are controlling only one normal speed fan per switch, any 1A diode will work, such as 1N400x (x is any number), like this: http://www.radioshack.com/product/in...ductId=2036268

If you are controlling multiple fans per switch, or a single high speed fan, use 3A diodes like these: http://www.radioshack.com/product/in...ductId=2062577

If you are controlling multiple high speed fans, I'd go with even larger 6A diodes, like these: http://www.radioshack.com/product/in...ductId=2062591

The 3 diodes as pictured drops ~2.1v total, which means the fans get ~9.9v from a 12v source. Add more or less diodes to suit. Just be aware that if you add too many, the fan may not want to start. Usually, nothing less than ~8v is recommended. Also note that if the motherboard is monitoring the fan speed via the rpm signal wire, reducing the fan speed too much may cause the motherboard to think something is wrong with it. If you are hooking the fan directly to a simple 12v source, this doesn't matter.

You could also use a regulator to get any speed you want, but depending on the number and/or power level of the fans used, heat could be an issue. That's why most (if not all) fan speed controllers use a more complex PWM setup. This could also be made with some 555 timers, but if you are going to get that complex, you might as well just buy a pre-made setup and be done with it.

Hmm, now you've really sparked my interest.
How hard would setting up a potentiometer be?
I would assume it'd be much simpler than this, as the potentiometer would just step down the voltage itself?

The tri-cool's in the back of my case are controlled through the switch, which I've mounted to the back of my case.
However, now that you've brought this switching method up, I may have to just do this since it'd be much easier to flip open the front door, tick a switch, and be done with it then reaching to the rear (linc..) and blindly move the switch.

Does such a switch exist with 4 throws?
Possibly wire it to a low, med, and high (12V), and off...
In my case, I'd just set the manual switch in the back to HI, and control through the switch up front?

A typical potentiometer doesn't have the power rating needed - they are usually rated for only about 1/8 watt or so. Let's say a fan is rated 12v and draws 100mA. If you set the pot so the fan sees half the voltage, that means the pot is dropping the other 6v, and therefore has to dissipate 0.6w (100mA x 6v). That of course assumes that the fan will still draw 100mA at the reduced voltage, which it probably won't, but even so, the pot isn't strong enough. You'd have to use a power rheostat, and those can get more pricey, are larger (to dissipate the heat), and you'd have to order them from someplace like Digikey, Mouser, etc.

Yes, there are switches with 4 or more "throws"; those are typically rotary switches. RadioShack has a 6 position rotary switch, but is only rated for 0.3A. This could work if you are positive each fan will pull less than that.

If this is done right, you could mount your switches to a blank 5.25" expansion bay plate so they are all within easy reach.

I see, so what are they using in the fan controllers?
I know I got one that looks like a very, very large potentiometer..

So Radioshack doesn't stock any switches that have 4 positions..
Hmm.

If I were to get a hold on a 4 "position" switch, would the wiring also be the same?
Just, with diodes in different positions?

Also, the schematic you linked to, would be a SPDT switch (3 positions so I can remember :p), Low, Off, High?

And if I had the fans set to "High", and wired the switches to my blank 5.25" cover, would that work equally as well?

Thanks so much BrianG...

Here's a rotary switch with 6 positions... No specs, but based on the case size I think it could handle a few PC fans. :smile:

http://www.allelectronics.com/make-a...-SWITCH/1.html

Most fan controllers do use pots, but in the PWM control circuit, so they aren't directly controlling the current flow to the fan.

No, RS doesn't have any 4 position slide/toggle switches that I see, the best they have is a 6 position rotary switch.

Yeah, the wiring would essentially be the same. If you were to use the rotaty switch, the common terminal would go to the fan, and each of the other contacts would go to different diode banks (one diode, two diodes, etc). I can provide an exact schematic if you end up getting a rotary switch.

Yup, the 3 position switch would the low, off, and high. Actually, the center position just doesn't contact either of the two internal contacts, so it is off. It's still technically a 2 position switch, but with an added "center off" detent in the middle that hooks to nothing internally.

Sweet, this could be my project for the long 3 day weekend coming up!

Would a simple 3A diode work over the 3 1A ones in series?

Also, wouldn't the lowest rating of the fan be right around 5V? As most fans "low" would be be ~5V

I think BrianG suggested the 3 diodes because each one has a 0.7V drop, for a total of 2.1V drop, the amp rating is still 1A. Using a single 3A diode would only get you the single 0.7V drop.

Just when I thought I was getting a hang of things...

If I were to try and get it to drop to 5V... how would I find how much a diode will drop?

I haven't read the thread, but the wall of text says that you guys are going at this waaaay too much...

hook them to the 5v rail and be done with it...

Well I'm trying to setup a switch that would differentiate from 5V to 12V, and Off.. and such.

It's a wall of text because I'm electrically stupid... :oops:

--

EDIT:
Found a cool article while surfin' da webz.

http://www.silentpcreview.com/article6-page1.html

Pretty much talks about how to get 5V, 7V, or 12V through a switch, without using diodes..

Are there any cons to that version over Brians?
Thanks guys, and sorry for the excessive questions!

Hey guys, was kinda bored in math class earlier today, so thought why not draw out what I'm gonna do?

Anyways, I still have a question; how many and what diodes will I need to drop a 12V connection to 5V?
Or how do I find how much voltage a diode will drop? So that way I can figure out which diode I will need...

Here's what I have so far...
A rough diagram:
http://i53.photobucket.com/albums/g6...4/th_Scan1.jpg

Normal silicon diode is 0.7V forward voltage drop, so 12V to 5V is 7V drop, you'd need 10 diodes in series.

Why not just have the switch connected to the 5V rail in one position and the 12V rail in the other, and detent middle could be off.

Being completely illiterate when it comes to components + electrical stuff, how do I wire that up on a normal SPDT switch? :oops:

Here are a couple more diagrams for 12v/7v and 12v/8.7v, whichever you prefer:

http://scriptasylum.com/forumpics/fanspeed1.gif

http://scriptasylum.com/forumpics/fanspeed2.gif

That will run the fan at 12v or 5v. The 5v setting may be too low for a 12v fan to even start, and even if it does, airflow will be so little it will probably be all but useless. 7v is about the lowest I'd personally go...

I was thinking about that- 5v might be too low to run the fan...

You can actually get 3.3v, 5v, 7v, 8.7, and 12v from a PC supply, but from experience, the 7v, 8.7v, and 12v are the usable voltages. The diodes are only needed if you want something different than those values.

I've seen numerous tests running fans @ 5V, 7V, and 12V.
SilentPCReview.com being on of them.

I'm sure it wouldn't matter if I used + or - for the wire that goes through the switch, right?

Just in case you're wondering where the 3.3v wire is...

http://atxg4.com/images/ATX.jpg

Aaaaacctualllyyy....... :lol:

It does. Either diagram would need to be hooked up exactly as pictured to work properly. :smile:

Well, 5v may work, but may not. You could temporarily wire the fan up to 5v and see if it starts reliably (may "grumble") and has the airflow you want before you make a more permanent switch setup.

That's correct!

:lol: Where's my gold star sticker?

http://www.pixabella.com/wp-content/.../gold-star.png

Good enough for you? :smile:

Yes, that will do. :mdr:

Well......
It's a dang good thing I asked then... :whistle:

I was gonna just not ask and just hook it up the other way, what would go wrong if I did reverse? Sparks and sparks? :lol:

I have one of the fans that I'm gonna be picking some more up later, so I'm gonna try and test that out this weekend.

If 5V doesn't do much, may just stick to 7V then..

I'm not sure what you mean by "hook it up the other way", but if it wasn't hooked up properly then, yes, you might see some sparks- or your computer just turns off and won't turn on again, or nothing happens at all. :lol:

Edit: There is a possibility of smoke being involved as well.....

Wired up a temporary molex connector for 5V instead of 12V, the LEDs are dim, the fan doesn't spin, turn it about 100%, then to 25% (to get the fan started), and it'll spin, however there's no noticeable airflow until around the 75% mark... (You can say it BrianG... you told me so! :oops:).

Looks like I'll be switching between 7V and 12V... maybe.

So, in order to wire up the "temporary molex" thing, I would have the negative black wire from the fan to the red wire (5V), and the positive to the yellow wire (12V), right?
And this will just push out a constant 7V?

:lol:

Yeah, a constant 7v setup would be the fan - going to the 5v line, and the fan + going to the 12v line.

But at least it's quiet! :lol:

So... should I be worried if I am using 7V?
I've read where 7V may be "stressful" on the PSU...

I'm not sure... I think BG would have mentioned it if it was a problem, though. If it's bothering you, just do the diode thing and use 7 diodes in series to drop it to 7.1 volts.

I'm mainly worried since this PSU isn't exactly a.... "nice" PSU... :whistle:
Coolmax 600W cheapie PSU, so I wouldn't know how it'd feel if I used the 7V on it.