Amplify 5-12VDC 5W to 50+W

Discussion in 'General Electronics Discussion' started by mike_, May 24, 2013.

1. mike_

4
0
May 24, 2013
I'm trying to control several computer fans. I have a reference circuit, but it is only capable of 5W. I plan on having 8 fans, and they all need about 5W each.

Is there a circuit similar to a LM317 voltage regulator that I could build to get what I'm after? I don't mind a simple linear regulator and can build a heatsink I'm sure I'll need. The only thing I care about is that the fans all run the same speed, based on the input voltage.

I honestly have no idea what something like this would be called.

2. (*steve*)¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥdModerator

25,505
2,849
Jan 21, 2010
What you need is a power supply.

If those are 12V fans, then you need one rated at 4A or more.

If they're 5V, it needs to be rated at 10A or more.

3. mike_

4
0
May 24, 2013
Yeah, I sorta figured that would be obvious.

Do you have a recommendation for a circuit that will be fed 12V by the power supply I already have? I don't want constant 12V, I want the circuit to reference the analog voltage signal of my fan controller and vary the output accordingly.

In other words, when I don't need a lot of cooling power my fan controller will reduce it's output voltage down to 5V automatically. It's an analog output signal, so the fans will have an infinite amount of speeds. A power supply does me no good, it always outputs 12V so the fans run loud all the time. So, I have a 12V power supply. I want to make something that keeps the output voltage of my fan controller and fans the same. This circuit wouldn't be powered by the weak 5W output, it would use the 1500W+ of 12V I have available.

Is there a circuit like a lm317 that uses a voltage as a reference instead of a resistance? I would have no problem building several of those, if I knew how.

4. (*steve*)¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥdModerator

25,505
2,849
Jan 21, 2010
Yes, the LM317 Ref pin is the reference voltage. The output is 1.25V higher than this pin.

If you place a 470 ohm resistor from the ADJ pin to ground, then take your reference voltage, pass it through 2 diodes (to drop approx 1.25V) and then connect that to the adj pin, the output of the LM317 will track (pretty closely) the input voltage.

Note that the voltage tracking will stop as soon as your reference voltage hits 1.25V -- because that's as low as the LM317 goes.

This isn't very efficient, and I'm surprised your current power source is not using PWM to control the fan speed.

5. mike_

4
0
May 24, 2013
Ah, perfect! Thanks!

The specs for the controller specifically says "Of course without the annoying noise from PWM a highly precise analogue voltage is put out." I assume this is because not all fans tolerate PWM well.

As the voltage of the controller will never go lower than 5V I think this should work perfectly. I'm assuming it'll be less efficient the further output voltage is from the 12V powering the LM317, right?

Would that mean that the whole setup would only draw slightly more than the fans at full 12V speed? In this case about 50W? Or do the LM317s actually waste more power to lower the voltage?

I'm assuming that my current fan controller is doing the same thing, because this is also listed in the specs.
" If a too high load is drawn from a fan socket the electronics automatically detect it. A switching of the output voltage to 100% reduces the thermal power loss in the system."

So switching output voltage to 100%, or 12V, means the LM317's waste no voltage and therefore the circuit cools down.

Are there any IC's like the lm317 but with a higher current capacity?

6. (*steve*)¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥdModerator

25,505
2,849
Jan 21, 2010
1) analog voltage -- perfect!

2) yes, the further it goes from 12V the more power the 317 drops. Note that there will be a minimum drop of about 1.5V across the 317 (it depends on load current). So you're never going to get the full 12V to the fan. If you can find a low dropout version of the 317 you could do better, but I'm not sure that such a beast exists.

3) The fans will draw less current at lower voltages, so the power consumption in total should fall with voltage. Yes, there's a slight quiescent current required for the 317, but that should be insignificant in this application.

4) yes, 100% is lowest loss in the speed regulation circuits.

5) higher current parts exist, but I'd stick to the lower current devices and use one per fan if you need to.

Here are a couple of examples which might work better:

http://www.digikey.com/product-detail/en/AP1186T5-L-U/AP1186T5-L-U-ND/1301237
http://www.digikey.com/product-detail/en/BA00CC0WCP-V5E2/BA00CC0WCP-V5E2CT-ND/3768954
http://www.digikey.com/product-detail/en/BA00CC0WT/BA00CC0WT-ND/722247
http://www.digikey.com/product-detail/en/BA00DD0WCP-V5E2/BA00DD0WCPV5E2CT-ND/3663742

I haven't looked at the specs (I'll check the first 2...)

The AP1186 requres a higher "control" voltage. This will work ig you have access to a slightly higher voltage rail (14V would be OK, 16 would be the max I would apply). This will give you at least 11.6V from a 12V rail (with a 16V control input)

The BA00CC0WCP-V5E2CT-ND operates more like a traditional regulator (no need for a higher voltage control input) and can operate pretty much as I've describes the 317 will. It will give you up to 11.5V from a 12V rail.

7. mike_

4
0
May 24, 2013
Oops, somehow I got myself confused and thought I'd need several LM317's per fan. At 1.5A I should get 18W per LM317 circuit, right?

I think I'll make one for each pair of fans, since they probably surge a bit on starting.

I'm still trying to decide if a low drop IC would be worth it. The BA00CC0WCP-V5E2CT-ND has a few extra pins, but I'm assuming I'd hook it up the same as the LM317? Can I just ground out the control pin to keep it on all the time?

Thanks for all your help Steve, I really appreciate it!

8. (*steve*)¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥdModerator

25,505
2,849
Jan 21, 2010
Yeah -- kinda.

The LM317 will limit the current to a value that won't kill it, and this won't really hurt the regulator or the fans, but keeping the dissipation as low as possible on the regulators will certainly keep them happier. (note that leaving the regulator overloaded can kill it, short overloads are safe)

The extra 1.5V or so that the fans will receive will make them go much closer to their normal "full" speed. I would certainly seriously consider them, perhaps even trying one to find out what the difference is (although, the cost of buying 1 will probably be about the same as buying 4 or so, considering shipping).

Pretty much. The datasheet gives you some reference designs. The major differences are the additional ground pin and the control input.

These regulators are also probably more sensitive to the requirement for an output capacitor. (The datasheet for this one does not make a song and dance about it though).

You can tie it to the appropriate value to keep it on all the time. The control voltage needs to be between 2V and Vcc to make the regulator operate (tie it to Vcc)

[/quote]Thanks for all your help Steve, I really appreciate it![/QUOTE]