While these guys are offing good help (some have forgotten what its like to be "inexperienced" IMO though and are being a bit "jerkish".. pretty common here sadly.. good on you for letting it slide right past you and not taking offense)..

There is one thing that I think you need to tell us before it can go any further..

You stated "Smart" fan controller... Nothing about manually turning a pot is "smart" to me...
Why don't you define how you intended this device to be "smart" so "better" suggestions can be provided..

"smart" in that its sensing temperature and adjusting fan speed automatically?
"smart" in that it can be digitally controlled via some signal?

The LM317 circuit is done/perfect for manual control though..

 

While these guys are offing good help (some have forgotten what its like to be "inexperienced" IMO though and are being a bit "jerkish".. pretty common here sadly.. good on you for letting it slide right past you and not taking offense)..

There is one thing that I think you need to tell us before it can go any further..

You stated "Smart" fan controller... Nothing about manually turning a pot is "smart" to me...
Why don't you define how you intended this device to be "smart" so "better" suggestions can be provided..

"smart" in that its sensing temperature and adjusting fan speed automatically?
"smart" in that it can be digitally controlled via some signal?

The LM317 circuit is done/perfect for manual control though..

Thanks! some comments here were great help and I am sure everyone means well.

I have a c# software that sends thermal sensor info via serial to an arduino that displays it nicely on an LCD. The arduino also controls the whole operation with relay, I just had trouble with the voltage control circuits sine I am new to them, In programming I am better hehe. I will use this circuit as a "manual override".

On another note, why do I need to regulate the voltage that comes out of the PSU? To my understanding it is highly accurate else some components would not work in my PC.

 

Back in the beginning the reason you toasted a few pots was due to the current a fan draws. Here is the problem, I did some test years ago using 12 Volt 80mm fans. At 12.078 Volts the fan current was 157.4 mA. What that means is the fans power was 12.078 Volts * .1574 Amp = 1.888 Watts of power. You would need a pot rated for about 5 Watts to work comfortably. This is why you were burning up those little 10K pots. Dogy provided a good circuit which would work just fine. While it looks a little complicated it really isn't.

However, rather than try and understand all of this there are dozens of inexpensive fan speed controllers on the market ready to do what you wish to do. I would just buy a simple plug and play solution.

Ron

 

Should I supply it with a higher input voltage to avoid the dropout?

Every linear regulator will have some loss of voltage. That's why you can never get the whole 12.0 volts out when 12.0 volts is your input.
There are ways around that. If you give the LM317 regulator 13.8 volts, you can get 12 volts out of it.
You might find a LDO (low drop out) regulator that only loses a few tenths of a volt. You might use a fast switching converter to boost its output up to 12.0 volts for any input from 6V to 12V.

That's 3 I can think of immediately. There are probably a few more.

 

I guess I could add a bypass to just get the 12 volts, thanks!



I asked since the Vmax in the schematic is lower than 12 volts, but I guess I can bypass as I stated above?

Should I supply it with a higher input voltage to avoid the dropout? Will that help me get a 12vmax output? (Can I use the 5v rail for that somehow?)

If you want to increase the output voltage, change the 1k pot for a 2.2K, OR change the 220 resistor for a 100 ohms.

 

If you want to increase the output voltage, change the 1k pot for a 2.2K, OR change the 220 resistor for a 100 ohms.

Thanks again, do you calculate it by hand? Since I do not know how to do math on a circuit with a component like the lm317.
I tried circuitlab.com to try and simulate it but they do not have that component to test with, only simpler.

 

Thanks again, do you calculate it by hand? Since I do not know how to do math on a circuit with a component like the lm317.
I tried circuitlab.com to try and simulate it but they do not have that component to test with, only simpler.

It's the resistors R1,R2., i always make R1 120ohms, this gives a static curent of 10mA.

http://www.reuk.co.uk/wordpress/electric-circuit/lm317-voltage-calculator/

 

Cheers I will use that!

After speaking to a corsair rep in my local hardware store, the fans are capable of receiving 13.2v.

It's the resistors R1,R2., i always make R1 120ohms, this gives a static curent of 10mA.

http://www.reuk.co.uk/wordpress/electric-circuit/lm317-voltage-calculator/

After messing around with the resistors and my PSU I am just not getting enough voltage. The fans are rated for 12v and anything under that is lowering their performance. My goal is to regulate between 12 and lower while reaching 12v is very important.

Is there any way I could get more voltage from my PSU leads so my Vin would be higher and this will be easily solvable?

To clearify, I have access to one molex power connector and nothing more, meaning I have a 5v and 12v lead and their respective grounds.

 

You cannot series connect an ATX's 5V & 12V.
If you want higher voltage then you need to boost the 12V

 

Using any adjustable voltage regulator you will have a "drop out" voltage. Even using for example a LDR (Low Dropout Regulator) you can not start with 12 VDC and have 12 VDC out. The typical output of a 12 volt ATX PSU is generally right at 12.00 VDC.

One option is to use a DC to DC converter to convert between for example 9 to 18 VDC to a regulated 15 VDC and then use the LM 317 for your adjustable output. Starting with 15 VDC would leave room for the dropout voltage. Here is an example of a DC to DC converter which would afford a little over 2.5 Amps @ 15 VDC output. Calex 12S15.2666NCM would work.

The simple method and less costly would be just buy a simple PWM turn key solution from a distributor like NewEgg. These devices make for inexpensive and simple fan speed control solutions.

Ron

 

Here's the peanut gallery take on this:

 

Quick and easy solution: Go to the junk yard and pick up (scavenge) the light dimmer from a modern vehicle. It's a PWM control that controls the intensity of the dash lights. I have one laying in my stock cabinet. I don't recall exactly how it's wired, but if memory serves, it acts as a sink; meaning the fan sees the 12 volts and the PWM takes it to ground in moments of on and off. Look up "Pulse Width Modulation" to learn more about how they work.

The module will likely control your fans far better than if you build a circuit of your own. Then you have time to study up and learn more about electricity and the math involved. Learn about power (watts in DC circuits) and how changing resistance or voltage affects amperage. Amps times volts is the calculation of watts. Don't be afraid of it - tackle it as best you can. I'm 60 and still haven't gotten the hang of it. It takes me 20 minutes to figure out what others know in just seconds. Or a minute if it's a complicated calculation.

 

If you want an adjustable voltage that goes nearer to 12V maximum you can use a low-dropout regulator such as the LT3083.
That will go to nearly 11V output with a 12V input.

 

Big pots ( rheostats ) were long used for lamp dimming & motor control. My sister described high school stage lighting controls, SI, NY, 1939, as 10 in. dia. with perforated metal covers & got really hot. Next came thyratrons, then SCRs & triacs.
Other possibilities are boost converter operating from 5 V but pushing 5A input, or boost- buck converter like XL6009DU on ebay, US, $4.49. No need for a V regulator.

 

Here's another vote for an off-the-shelf cheap PWM controller like the ones you can find on eBay. It would allow full 100% power at 100% duty cycle and, depending on the rating, would be fine for controlling several fans in parallel. You can likely even find one that can be controlled by your Arduino, for instance by a 0-5V input voltage to produce 0-100% duty cycle. Don't mess around with linear (voltage) controllers.

I know it sounds daunting at first and the intuition is that a linear controller is somehow simpler. Try to get over that. Maybe linear is a little easier to understand at first, but PWM will be a simpler solution in the end because it will be so much more versatile and powerful. For one thing, PWM can drive a fan to a lower RPM reliably whereas a low voltage can cause the torque to fall so low that the fan won't start and run.

 

Here's another vote for an off-the-shelf cheap PWM controller like the ones you can find on eBay. It would allow full 100% power at 100% duty cycle and, depending on the rating, would be fine for controlling several fans in parallel. You can likely even find one that can be controlled by your Arduino, for instance by a 0-5V input voltage to produce 0-100% duty cycle. Don't mess around with linear (voltage) controllers.

I know it sounds daunting at first and the intuition is that a linear controller is somehow simpler. Try to get over that. Maybe linear is a little easier to understand at first, but PWM will be a simpler solution in the end because it will be so much more versatile and powerful. For one thing, PWM can drive a fan to a lower RPM reliably whereas a low voltage can cause the torque to fall so low that the fan won't start and run.

Ok then, I will look at PWM too! But my fans are 3 pin fans that do not support PWM, can I just ignore that fact?

 

Your fan can work using PWM.
Just apply PWM signal to the 12V Vcc pin

 

Ok then, I will look at PWM too! But my fans are 3 pin fans that do not support PWM, can I just ignore that fact?

All of the three pin computer type fans I have worked with run on PWM, it's how they change speed. Three wire fans just use a third wire for a pulse out or tachometer out signal (2 pulses per revolution is standard in home PCs). Anyway hoem PC fans like the 80mm and 120mm Corsair, Antec and the lot all run off a PWM signal. This is how the smqart ones vary their speed for component temperature.

Ron