Potenciometer for fan regulator

[Behemot]

Re: Potenciometer for fan regulator

I don't have bloody idea?! I guess the Super IO just sents some voltage and looks at sensor what speed comes back? Anyway, the speed usually tends to react on CPU temp so if it still increases, so the voltage could be encreased. Simple as that?

Look e.g. into , page 45, CPU Fan Control mode and following.

[budm]

Re: Potenciometer for fan regulator

see this about the cpu fans: http://pcbheaven.com/wikipages/How_PC_Fans_Work/
Page 24: The speed control function requires the use of a fan with fan speed control design. 3-pin Fan only has TACH output for monitoring.
The Voltage mode can be set for a 3-pin CPU fan or a 4-pin CPU fan. However, for a 4-pin CPU fan that is not designed following Intel PWM fan specifications. So basically, you will need to know what your fan voltage is and set the voltage to match your fan.
You can easily try hooking up Fan with 2 pin, there will be no TACH for the CPU, so more likely the system will shutdown, then try it with 3pin fan with Tach output, set the voltage for the correct fan Voltage (6, 9, 12V or what ever the fan is), then see if the b+ will vary from the set point when temperature goes up or down.

[Behemot]

Re: Potenciometer for fan regulator

I don't see your problem. I can get plenty of tach wires. Either from the fan cooling VRM and stuff arround CPU, or from ANY of the radiator fans. I have never been talking about 2pin fans if that is what you mean?

[budm]

Re: Potenciometer for fan regulator

Just monitor that red wire voltage when temp change and that will answer your question. You mean the two pin fan I had made the controller for in the beginning of the project?

[Behemot]

Re: Potenciometer for fan regulator

Nope, will off course modify it

There is no problem to lead the tach wire from any fan connected to it all the way back to the motherboard, is there? Basically the board will really drive all the fans, just with transistor powerfull enough to power several fo them. And also possibility to switch on fully manual control :-)

[budm]

Re: Potenciometer for fan regulator

Then just have to find what kind of control signal (PWM, variable DC Min to Max voltage) is sending out the by MB and build the interface to use it to control the fan controller. Should not be that hard.

[Behemot]

Re: Potenciometer for fan regulator

Well, basically, when you talk about it, using pulsed voltage should work as well, I don't think the pulses are so quick that the transistor would not react to it quickly enough, right? I will just have to abandon any capacitor on the regulator circuitry.

[budm]

Re: Potenciometer for fan regulator

Well if you read this: http://pcbheaven.com/wikipages/How_PC_Fans_Work/
If you pulse the B+ (the transistor will be turning ON/OFF so the b+ will be pulsing, will your fan still function?
You need to find out more how your mother board function, and also the type of fan you will be using if it can handle pulsing B+. Not that hard to figure out.

[Behemot]

Re: Potenciometer for fan regulator

What I mean is this: there are two ways of PWM control:

1. you make pulses, like a gaps, which will be cutting the whole fan of power for a fractions of seconds, right? that will also make him run slower; I believe the transistor can reapeat these pulses on emitter than (high level - power flows from emitter; low level - no power from emitter)
2. you make just „information“ pulses which drive circuitry inside the fan, and than this circuitry slows the motor itself; this is what the 4th wire does, and this is the option I omit

Basically I mean the first option. Than, it does not really matter if the motherboard regulatur is voltage level (varies voltage from low to 12 V continuously), or voltage PWM based (it pulses), is it? The transistor should handle both, should it not? I believe we are not talking about 100 kHz here, and even if, most transistors can handle that.

[ben7]

Re: Potenciometer for fan regulator

So, You are wanting to build a voltage-controlled PWM circuit?

Why are you trying to make this more complicated than it should be? You could just use a transistor to amplify the current from the fan connector (with an added voltage drop of 0.7v)

[budm]

Re: Potenciometer for fan regulator

It is the electronics of the fan that I am taling abot if it can hadle the B+ pulse.
From the link "The 3-wire fan powers the Hall sensor and the controller from the same line that the coils are powered. Thus, if someone tries to send PWM pulses to the coils of a 3-wire fan, the same pulses will arrive at the controller. The controller will then malfunction".

[Behemot]

Re: Potenciometer for fan regulator

ben7: but that it exactly what I want to do :-D

budm: if this applies to all 3wire fans and controllers, than it gives us automatically the answer, yes? MB voltage control than cannot be PWM, because it will damage itself. Than the only option available is voltage level control. Because MB controller has to work with every 3pin fan you can connect to it. And it does.

[ben7]

Re: Potenciometer for fan regulator

ben7: but that it exactly what I want to do :-D

budm: if this applies to all 3wire fans and controllers, than it gives us automatically the answer, yes? MB voltage control than cannot be PWM, because it will damage itself. Than the only option available is voltage level control. Because MB controller has to work with every 3pin fan you can connect to it. And it does.

Have you tried an LM555 yet?

[Behemot]

Re: Potenciometer for fan regulator

Umm, nope, I am prety much satisfied with circuits similar to budm's so far.

Actual PWM controller working with the 4wire fans is my next goal in the future.