Engineers Welcome Project Automatic Fan Controller

[Mackintire]

Welcome,

Your mission is to provide ideas for building an automatic fan controller.

The control system is as follows

Thermaltake Tsunami VA3000 case
Athlon 64 3800
MSI K8T Neo2 KT800 Pro
Thermaltake silent tower with dual 90mm fans
(2) Seagate 200gig 7200.2 drives
(2) Western Digital Raptors 74gig
Radeon 9700 Pro with VGA silencer and Tweekmonster Heatsinks (350 GPU /360Mem)
(2) 120mm Fans silverstar 110CFM@12V



This is the goal I wish to accomplish.

The Case (system) thermal control will be taken by a thermistor (20K) from the top of the case for the case temperature.

A thermistor will be attached to one of the Western Digital HDs for hard drive temp control .

Another thermistor will be attached to the base of the Athlon 64 3800 for CPU temp control.


The room temporature is 32 degrees C. With quiet and cool running on a normal day; 34 C is the normal temp for the system temperature. Case fans running 1250RPM ~7.5V

Criteria (Rules for design)

The minimum speed of the intake fan must always be higher then the exhaust.

Both fans must scale up to maintain the temp of the case within 4 C of the outside room temperature. The room never heats up over 42 C

If the hard drives are heating up then the intake fan will scale in voltage untill the hard drive temp have fallen or stay below 42 C

The Cpu fans speed must be automatically regulated and its voltage input must add in part to the system fans overall voltage output. AKA if the CPU fans voltage spikes fast (from a fast temperature change) then both the intake and exhaust fans are given additional voltage.

All voltage control points must be tuneable. That is for example how much voltage the case fans recieve from the cpu getting X amount of voltage must be tuneable.

If your going to use resistors try to keep the current traveling through the resistors low as improper use creates heat


PS I have already designed most of this system. I am looking for your ideas. Who knows maybe I have done something wrong. My final design will be posted for others to use.

My parts list so far,

3 thermistors (20K) 54C is the 0 resistance point ( all from thermaltake products, I have more if needed )
2 JFETS PNP style
(assortment of resistors)
2 potentiometers 10K, 5K
5 (diodes)
2 .10micro F capacitors
Prefab circuit board
Lots of molex connectors
Wire


Good Luck,


Mackintire

 

[Talonz]

Mission Auto Fan Controller?

Seems to be missing the Payment section, how's 1/2 up front?

How do you have your system set up? People aren't here to do your work for you, sounds like you have a plan. Go through with it, tell us how it was, and we'll try to help you with any problems you have.

 

[Mackintire]

This is the short version of what I plan on building.

+12 volts is connected to the system thermistor in parallel with a 1.25 Meg ohm resistor, this provides the exact voltage 7.2+ volts to maintain a nice 1150 RPM minimum on the exhaust fan. The output side of the thermistor and resistor combo feeds another trace with a 120 ohm resistor and the 5K pot that is connected to the P type JFET in a IC socket. The Drain of the FET is connected to 12+V and the source is connected to positive intake fan terminal. Also connected to the positive intake fan terminal is again a combination of thermistor and 1Meg Ohm resistor and 10K pot combination. The negative terminals of both fans are connected to the 12 V ground.

This combination alone give me the ability to set the min fan speed for the intake and the amount and onset of voltage added by the system temp at any given time.

A similar technique will be used for the CPU fans , the output voltage will operate the Gate of another JFET which will speed up the system fan input voltage which will inturn supply voltage to the intake fan system. Since the one JFET will be biased so that they will only supply a higher potential at key transition times its heat output will be negligable. The other JFET however will at times be carrying up to .5 Amp. And will require a heatsink. Most of the time however the current through the FET will be around .10 Amps.

Therefore the IC I have chose to use will include 2 JFETS on one chip this will make the circuit design simpler to build. As I have been working alot recently I will not be able to finish this for a few weeks. If you see any flaws or conserns in my design please let me know.

Thanks,

Mackintire

 

[ocphangaz]

Or you can just use the thermistor as a reference voltage to do a voltage divider to control the fan speed... no need ICs, no need complicated circuits... just few passive components will do since you wanted it to be automatic

you can set the min voltage and max voltage (12V obviously) with voltage divider circuit as the thermistor acts as a variable resistor. and other components is just to act as controller.

just my 0.02

 

[iNsAnE_crAzY]

Or you can just use the thermistor as a reference voltage to do a voltage divider to control the fan speed... no need ICs, no need complicated circuits... just few passive components will do since you wanted it to be automatic

you can set the min voltage and max voltage (12V obviously) with voltage divider circuit as the thermistor acts as a variable resistor. and other components is just to act as controller.

just my 0.02

Uhh yeah what he said...lol

 

[bink]

first off, intake more than exhaust????

then as an engineer I would say do you want this system NEMA or IEC rated.. and charge your butt 300 bucks an hour for installation and other stuff, and another grand for a tech support package to tell you how to fix it when it breaks...

 

[Punchbugy]

first off, intake more than exhaust????

My guess is because he also has his power supply exhausting air, and he wants total intake and exhaust to be about equal.

 

[Fanerim]

Quote:
Originally Posted by bink
first off, intake more than exhaust????

My guess is because he also has his power supply exhausting air, and he wants total intake and exhaust to be about equal.

I do that to keep dust off my case..

 

[ocphangaz]

Originally Posted by iNsAnE_crAzY
Uhh yeah what he said...lol

But he is planning to use transistors .etc... what i meant was simplify the circuit furthur. just use the voltage divider circuit, patch it to the fan tail wires, and heatshrink it! Voila!

 

[HeThatKnows]

This is the short version of what I plan on building.
[snipped]

Got a schematic? Cause I tried to draw one based on your description and ended up with a big mess.

 

[potroast]

My guess is because he also has his power supply exhausting air, and he wants total intake and exhaust to be about equal.

bogus... if you're gonna go through this much trouble, then you better open that PS up and have those fans on control, too

 

[SUPERFLOP]

Personally, i like his idea with the Jfets, nice solid state operation, ina small package, get too much into analog and your project starts to blossom in the wrong ways. i have an idea but ive been out of school too long... someone remind me what those amp are calle dthat are on the input side of an A/D convertor, the ones that are nothing but a comparison circuit...set Vref (12v) and then give input and the out put is the difference between input and the Vref. then you take a few resistors, a thermistor and pot(for calibration), give it 5volts input (min) thru the thermistor giving you an output of 7Volts....schat this wont work at all im thinking thermistors work backwards....gaining resistance as heat builds...


i would delete this but maybe it will help you think of somethings...oh yeah im was thinking of Op-Amps, i believe (god i need to get back into electronics)


EDIT: why do you need all this circuitry? shouldnt you be able to splice in thermistors between the power source and fan? just get thermistors with narrow operating ranges and put in resistors to taste...

 

[101]

As an engineer I personally find thermally controlled fans to be very annoying acoustically. You brain will quickly adapt and ignore constant noise sources, but as soon as that noise changes you notice it again. Many people can hear when their 12V rail drops slightly under load as the they are quickly alerted to the RPM change of any audible case fans.

You will spend a lot of time and money (specially time) making a custom solution which in the end you will be unhappy with to some extent and replace it with something simpler and more reliable. If you main intent is to learn about circuits and some electronics then great, as that's what you will come away with.

 

[Mackintire]

Thank you for your feedback.

I would like to answer a few of your questions.

First the total sum of the current draw from all the fans in the system is higher then what I believe the thermistor is capable of handling without frying.

Second the speed of the fans in the system builds linearly not in steps as in a pulse width modulated fan controller.

Third I have never needed any of the system fans to exceed 9 volts. At this leve they are audiable but not loud.

The power supply is an Antec 550 and is already thermally controlled.

I did read somwhere that having a positive pressure system is prefered to having a negative pressure system. Which is why I choose to have the intake fan speed equal or higher then the exhaust fan speed. This is outside of accounting for the exhaust flow of the power supply fan.


Thanks,

Mackintire

 

[zer0signal667]

I did read somwhere that having a positive pressure system is prefered to having a negative pressure system. Which is why I choose to have the intake fan speed equal or higher then the exhaust fan speed. This is outside of accounting for the exhaust flow of the power supply fan.


Thanks,

Mackintire

This only helps when all your intakes are filtered. That way, air is not sneaking in through cracks and holes, so all air entering the case is dust-free.
It also bugs me when people say "positive" or "negative" pressure, because it's not really the right terminology to use. It would be more correct to say there is a postive or negative
pressure potential (difference) between the inside and outside of the case.
I guess that's just the engineer in me talking

 

[yusoweird]

This only helps when all your intakes are filtered. That way, air is not sneaking in through cracks and holes, so all air entering the case is dust-free.
It also bugs me when people say "positive" or "negative" pressure, because it's not really the right terminology to use. It would be more correct to say there is a postive or negative
pressure potential (difference) between the inside and outside of the case.
I guess that's just the engineer in me talking

hehe well when people say positive or negative pressure... as long as you know what they are talkin about... then, what they say is correct with respect to your understanding... I guess thats just the relativity in me talking ;D

 

[diredesire]

Both fans must scale up to maintain the temp of the case within 4 C of the outside room temperature. The room never heats up over 42 C

I'd hope not
that's 107 F

 

[Mackintire]

On the average day in the summer , with the air conditioning running, it is normally 86 degrees F in there. So when I say it has never exceeded 42 degrees C I mean't with the AC off and the door closed.


Mackintire

 

[MD_Willington]