Potenciometer for fan regulator

[budm]

Re: Potenciometer for fan regulator

The Zener is for regulating the output voltage if the input voltage fluctuates, nothing to do with limiting current, but if you need 12v out (input has to be higher than 12V, 15V will be good) the zener voltage will have to be 0.6v higher than the require output, I pick 1K base on beta of 100 on the transistor but I will not run dead short between Base and Collector. Voltage output will always be 0.6V less than the Base to Ground voltage because it is Emitter follower circuit.
Motor start up current can be high, so if 1A base current can damage the EB junction, not good practice to run transistor that way.
Yes, cap is 0.1uF.

[Behemot]

Re: Potenciometer for fan regulator

I am limited to ATX power supply, so 12 V is maximum obviously. Minimizing voltage drop is always good in here. As for fluctuating, according to ATX specs it should be 100 mV max. Nothing what motor is capable of reacting to anyway…

[tom66]

Re: Potenciometer for fan regulator

I don't see why you care if the fan is ~10% slower at full speed. Will it really matter?

[budm]

Re: Potenciometer for fan regulator

He can put the bypass switch to bypass the circuit and get full 12V to the fan.

[Behemot]

Re: Potenciometer for fan regulator

Yeah that is option too of coruse. Just that with 6-fan controller, it is difficult to squeeze two switches and potenciometer per fan into 5,25" position…

[budm]

Re: Potenciometer for fan regulator

Do you want to have ONE master switch to turn all 6 fans in full mode (with about 0.6v less than full voltage) when you need it? You can also have one of those potentiometer with switch built in to turn each fan on/off but can be forced on full with one master switch.

[budm]

Re: Potenciometer for fan regulator

Here is another minimum part counts circuit with low side controller (sink) instead of sourcing (High side drive) with Master bypass switch.
I show proof of concept in the pictures.

[Behemot]

Re: Potenciometer for fan regulator

Would like separate bypass switch for every fan and on/off switch for every fan, which both takes some space :-) Components on board are not problem, but ATM a guy wants six-channel controller into 5,25" bay…

[budm]

Re: Potenciometer for fan regulator

Well on/off switch can be on the pot, that will eliminate 6 switch on the panel. so you will have 6 pot/switches, 6 bypass switches.

[Behemot]

Re: Potenciometer for fan regulator

Interesting, didn't know these small pots with off switch. I think I can get some around here. And as long as this scheme of yours actually regulates outcoming power, having infinite resistance on base means transistor is closed, nothing going through. Hmm, seems nice :-)

As for fans, I have actually measured some and there is no strange volt-ampere characteristics going on when having it on low voltage, current goes down as well. Together with placing termistor in front of the fan, it whould be safe to mount like 20 fans on output (considering typical 150 mA power drain and 4 A transistor). Also depends much on used fans, but I mean like, it's just if somebody mounted more fans to one output than it is healthy, it should still survive, yes? Termistor would eat all the starting motor spikes so you could really connect many fans.

[budm]

Re: Potenciometer for fan regulator

The fan will have minimum voltage requiremnet to spin, if you make the output voltage go to 0 then slowly turn the pot up there will be blank area on the pot until minimum voltage for the fan to start spining again, you will have to play with the pot range. The pot at the lowest setting should be just a bit below min voltage requirement.
"And as long as this scheme of yours actually regulates outcoming power" but I thought you do not care about voltage variation. That is why I have Zener Diode in the first circuit to regulate the output voltage.

"having infinite resistance on base means transistor is closed" when base is @ Ground there will be no bias for the transistor.

"Together with placing termistor in front of the fan" What do you want the thermistor to do? You can use the thermistor in place of the pot to control how the fan should react with temperature.
Safe practice when building circuits is to double the rating on what your needs are, especially the parts that will generate heat, or expose to high voltage. For home use, that is up to you.

[budm]

Re: Potenciometer for fan regulator

Start with simple 1 transistor driving 1 fan, now you are talking about 20 fans controlled by 1 transistor, bypass switch for each fan, thermistors, etc? It is not going to be simple project any more. if you want to keep the transistor cool, you can mount it on the back side of the fan so the air will flow pass the transistor first and out to the exaust side of the fan.

[Behemot]

Re: Potenciometer for fan regulator

I am already counting with heatsink

Thermistor - NTC, for surpresing current spikes when the fan motor starts.

[Behemot]

Re: Potenciometer for fan regulator

I will resurrect this. A guy wanted me to make him regulator for radiator fans which would „repeat“ speed of CPU fan.

I think it should be possible under two conditions with this circuitry:

1. he has voltage regulation set in BIOS
2. the fan's voltage wire (red) would be used to drive base of transistor on the regulator

I think a motherboard varies the voltage from some lower (like 8 V or so) all the way up to 12 V. So with the 10k pot and 1k resistor in budm's scheme that may work directly, right? In the worst case scenario I would have to just slightly change the values, or replace the pot with 47k or 100k to lower the current flowing through base (if the motherboard will pull high too high voltage thus opening base more).

[ben7]

Re: Potenciometer for fan regulator

I am still confused at what you are trying to say, but I do my best.

1. The bios is controlling a connector on the MB which has one fan connected.
2. This guy wants to connect multiple fans to that connection.
3. You want a circuit that can amplify the current enough to power more than one fan, keeping the same voltage on for the fans.

Is that correct?

-Ben

[budm]

Re: Potenciometer for fan regulator

Yeah, I am confused too, since most cooling fan with speed control will have 3 wires, 12VB+ (FIXED), Ground, Speed control line by PWM or DC depends on the fan.

[Behemot]

Re: Potenciometer for fan regulator

Yep, more or less what ben said. I think I could use the red wire to drive base. Red is not fixed, is it? When you have voltage control (3pin header), it has to change voltage on the red wire. Ground is ground all the time, speed sensor (yellow) is jsut for feedback.

PWM control is the 4th wire, that is not what I want at all. On PWM you can just buy PWM fans and drive 50 of them, there is only signal on that for circuitry inside the fan, the means tiny current values.

[budm]

Re: Potenciometer for fan regulator

The red b+ is fixed voltage which is the B+ to run the electronics inside the fan, the control wire is used to control the electronics that drive the fan, fan the drive circuits is PWM controlled so it will not loose the torque. The fourth wire is the RPM feedback for monitoring if the fan has lock up. CPU does not control the B+ line for the fan since controlling fan by dropping the voltage will have torque problem, there is small circuits inside the fan to drive the coil and it is done using PWM. They do not use brush type DC motor any more or cooling fan.

http://www.cypress.com/?docID=17102

[Behemot]

Re: Potenciometer for fan regulator

There is ground (black), voltage (red) and tachometer sensor (yellow) on the fan. Since the voltage supplies power for BOTH electronics and motor, that is also the reason it cannot be too low. Electronics does not operate properly than (you can notice squeezing and so). Enermax has custom fans in their PSU's where they supply electronics with full 12 V and motor with lower voltage. But let's talk about ordinary fans.

Now, since there is nothing besides GND, voltage and sense, you have to vary voltage to lower fan speed. That is exactly what motherboards do EVEN ON CPU FAN HEADER when you have voltage regulation selected in BIOS.

What I want to use is this supply voltage also to drive transistor's base. Low CPU temperature = low fan speed = low voltage on red wire = gate less open = lower speed of radiator fans. The guy will most likely use ordinary fan to cool down CPU VRM and he wants to drive fans on his radioators automatically if possible. I see no reason why motherboard's controller can't control fans in water circuit as well. If he will need lower CPU temperatures, radiator fans will just increase speed than.

[budm]

Re: Potenciometer for fan regulator

"voltage regulation selected in BIOS" What voltage is that you refer too? CPU Voltage regulation:CORE VOLTAGE or the DC fan voltage, how would that work if they do not know what kind of min and max voltage and current the fan will require to run?
http://pcbheaven.com/wikipages/How_PC_Fans_Work/
Look at your PC CPU Cooling fan. Look at your BIOS.