PWM to Voltage bridge

Control: management of fans, temp/rpm monitoring via soft/hardware

Moderators: NeilBlanchard, Ralf Hutter, sthayashi, Lawrence Lee

Hypernova
Posts: 124
Joined: Sat May 31, 2008 2:33 am
Location: Australia

Post by Hypernova » Sun Nov 15, 2009 3:15 pm

Some what different topic

I have been planning on building my own fan controller for a while. But I am not well versed in the power systems side (my studies are in embedded systems and their programming).

So basically the specs I'm trying to achieve is:

-No PWM switching noise, so operating the MOSFET's in their linear region
-Handles up to 5A at 120C
-Takes 3.3V TTL PWM input @ about 5~10KHz

The Maxim application note AN3149 looked like a good start. So what criteria do I have to look at to replace the MOSFET with something that can handle 5A? The one (FQP27P06) Trident chose claims ID of 19.1A but he says it will only practically handle 2A. How does that work?

Mr Evil
Posts: 566
Joined: Fri Jan 20, 2006 10:12 am
Location: UK
Contact:

Post by Mr Evil » Sun Nov 15, 2009 4:42 pm

Hypernova wrote:...Trident chose claims ID of 19.1A but he says it will only practically handle 2A. How does that work?
It's probably to do with power dissipation.

Anyway, if I was making a fan controller from scratch, I would make it in the style of a DC-to-DC converter. This is much more efficient than a linear design (theoretically 100% efficient, with >90% being easy to achieve in reality), which massively reduces the requirements of the MOSFET.

Hypernova
Posts: 124
Joined: Sat May 31, 2008 2:33 am
Location: Australia

Post by Hypernova » Mon Nov 16, 2009 3:02 am

Buck converter looks even more complicated... It appears to be affected by load size a lot. So you need different sets of L and C for different load

Would RC filtering the output of High Side-Drivers work? It sure looks a lot simpler
I'm thinking of this one

So at 15KHz you have R of 0.02 and a cap of 250uF. Taking Ron, Rr total voltage drop and internal diode forward of 0.6 should be 0.9V at 4A.

Mr Evil
Posts: 566
Joined: Fri Jan 20, 2006 10:12 am
Location: UK
Contact:

Post by Mr Evil » Mon Nov 16, 2009 4:31 am

An RC filter would not be efficient, so there is no advantage to doing it that way.

A buck converter doesn't have to be complicated. It could look something like this:
Image
You just need to choose values of L and C that give acceptable ripple with both the minimum and maximum loads you expect.

Hypernova
Posts: 124
Joined: Sat May 31, 2008 2:33 am
Location: Australia

Post by Hypernova » Mon Nov 16, 2009 4:39 am

Well that's not in wikipedia...

I'm having trouble on how to get the LC values. All I know is that D should be fast. MOSFET need to to handle the current load with cooling. And that's it.

Mr Evil
Posts: 566
Joined: Fri Jan 20, 2006 10:12 am
Location: UK
Contact:

Post by Mr Evil » Mon Nov 16, 2009 4:52 am

It would probably be a good idea to start a new thread if you want to explore the idea further.

Hypernova
Posts: 124
Joined: Sat May 31, 2008 2:33 am
Location: Australia

Post by Hypernova » Mon Nov 16, 2009 5:23 am


Shaggy
Posts: 13
Joined: Wed Jan 03, 2007 5:44 am
Location: Spain

Post by Shaggy » Fri Nov 20, 2009 2:04 pm

Hi all,
i have a question about the Maxim AN3149 circuit, is it ok for using it with a low frequency (two pin) pwm controler?

Thanks.

trident
Posts: 56
Joined: Sat Jan 13, 2007 10:16 am
Location: Indiana, U.S.A.

Post by trident » Sat Nov 21, 2009 9:36 pm

Shaggy
Your question doesn't have enough information for a definitive answer. What do you mean by low frequency and 2-pin? I am going to guess that you have a controller that outputs high current (500 to 1000mA) PWM that is capable of directly powering a typical fan. If this is the case you may want to look at one of the commercially available "fan attenuators" that connect between the fan and the controller.
_______
trident

Shaggy
Posts: 13
Joined: Wed Jan 03, 2007 5:44 am
Location: Spain

Post by Shaggy » Sun Nov 22, 2009 10:33 am

Hi trident,
thanks for your answer, and sorry for the not enough information, i thought the less i wrote, the less mistakes i made, my english is not as good as i wish.

Your guess is right, i would like to use it to convert PWM output, of my gpu fan controler (i mean the one the video card has), in order to control accelero turbo module fans.

I´ll try a fan attenuator from mcubed or alphacool, but isn´t there any way to use the maxim circuit?, i´m planing to use 2N2222 and 2N2907 as transistors.

Edit: I supose your circuit isn´t able to handle a 12v PWM output, isn`t it?

Thanks.

trident
Posts: 56
Joined: Sat Jan 13, 2007 10:16 am
Location: Indiana, U.S.A.

Post by trident » Sun Nov 22, 2009 8:20 pm

Shaggy
You are correct, the MAXIM circuit was designed for a 3.3 volt PWM signal, however you could add a voltage divider configured to divide buy 3.6 (100k and 39K) to scale the 12 volt signal to a 3.3 volt signal.
_______
trident

Shaggy
Posts: 13
Joined: Wed Jan 03, 2007 5:44 am
Location: Spain

Post by Shaggy » Mon Nov 23, 2009 5:40 am

Ok, thanks trident.

So, i only have to add a voltage divider between PWM output of my video card and the MAXIM circuit; it sounds easier than redesing all the circuit with oder components, as i have planed.

Thanks a lot, you helped me very much.

trident
Posts: 56
Joined: Sat Jan 13, 2007 10:16 am
Location: Indiana, U.S.A.

Post by trident » Mon Nov 23, 2009 9:11 am

Shaggy
My apologies, I did not consider the graphics card could be using low side drive. If this is the case, using the divider, the signal would be 180° out of phase. That is 12 volts @ 0% duty cycle and 3.3 volts @ 100% duty cycle. If you use differential sensing, low side drive won't be an issue, but the circuit will be more complex. Don't know how skilled you are with electronics, can explain further if needed.
_______
trident

Shaggy
Posts: 13
Joined: Wed Jan 03, 2007 5:44 am
Location: Spain

Post by Shaggy » Mon Nov 23, 2009 12:10 pm

Hi trident,
i´m not skilled at all, the only thing i could do were making the circuit and pray for nothing get burned when it were plugged, thats because i ask about it.

This is too dificult for me, i simulated the circuit on LTspice application, and everything seemed easy.

Your last post is chinese for me, if you want to explain it, i´ll be very gratefull , but its very probable you waste your time.

I think i´ll go for a fan attenuator, or i´ll plug fans directly (i supose i can because turbo module fans draw less power than stock fan).

Thanks for your time.

Post Reply