Cheap approach to down convert 19v to 12v to power a computer case-type fan?

[cyf]

I have a bunch of small computer case-type fans lying around, but they run at 12V and I wanted to hook one up to my router, which has a USB port (which provides 5V, but I need that for a USB drive and I don't have the $ to buy a 5V fan right now).

The power supply for the router runs at 19VDC, and I figured I could tap that to run one of the 60mm fans I have which is rated at 12V @ 0.17A--although some initial measurements with my multimeter shows that it draws about 40-50mA at ~12V (hooked up to the power supply that's integrated into my soldering iron station).

A long time ago I bought an electronics "components kit" from Microcenter on clearance for $33 http://www.elenco.com/search/searchdetails/basic_parts_kit=ODc= that includes a range of resistors, but the resistors are the the 1/4 watt size.  I've hooked three 470 Ohm resistors (in parallel) and then in series with the fan [edit: sorry, I forgot to mention that the resistors are in parallel because I figured they could dissipate the heat instead of using a single 150 Ohm resistor) and it seems to be about right, but I'm worried that they will get hot and burn out if the fan stops or gets jammed up for whatever reason.  There are a bunch of other components in the kit, including transistors, capacitors, etc., so I'm thinking I can throw together something simple that would be able to handle a range of currents and finally put some of these components to use.  It doesn't even have to adjust the speed of the fan--I just don't want to set my house on fire, as it were.

Any thoughts would be appreciated, or point me to something I can search.  The link above has the info that's better formatted, but here's the parts list:

Resistors:         1/4W - 5 each   10?, 100?, 470?, 1k?, 2.2k?, 4.7k?, 8.2k?, 10k?, 18k?, 27k?, 47k?, 82k?, 100k?, 220k?, 330k?, 470k?, 1M?
Electrolytic Capacitors:        50V Radial - 2 each    1?F, 4.7?F, 10?F, 47?F, 100?F, 470?F, 1000?F
Mylar Capacitors:       100V Radial DIP - 2 each    .001?F, .0047?F, .01?F, .047?F, .1?F, .33?F
Ceramic Disc Capacitors:       500V - 2 each    2.2pF, 4.7pF, 10pF, 1pF, 22pF, 47pF, 100pF, 220pF, 470pF, 680pF
Inductors:     1 each    4.7?H, 10?H, 47?H, 100?H, 470?H, 1mH, 10mH, 33mH
Trimpots:    Single turn - 1 each    500?, 1k?, 5k?, 50k?, 100k?, 500k?, 1M?
Switches:    2 each    Push-button, SPDT slide, DPDT slide, 8-position DIP
LED's:   Standard T 1 3/4 size - 5 each    Red, Green, Yellow
Displays:     2 each    7-segment common cathode, 7-segment common anode
Integrated Circuits:     2 each    7805, 7812 regulators, LM386 audio amp, 555 timer, 741 Op-amp, 74LS00, 74LS02, 74LS08, 74LS10, 74LS47, 74LS74, 74LS86, CD4013, CD4017
Transistors:   2 each    PN2222, PN2907, 2N3055, 2N3904, 2N3906, 2N5457
Diodes:      2 each    1N4148, 1N4001, 1N34, 1N4728, 1N4733, 1N4739, 1N4743, 1N4747
Miscellaneous:   25' 22AWG Hook-up wire, Alligator lead set (red and black), 5V DIP relay, Photo Cell, Microphone, Speaker, Buzzer, 9V battery snap, 4A 200V SCR

------Updated as of message 15 - Posted here because I like it when OP's update their original posts to make things easier to find.------------
https://www.eevblog.com/forum/beginners/cheap-approach-to-down-convert-19v-to-12v-to-power-a-computer-case-type-fan/msg530519/#msg530519------

Hey guys.  Wow.  Nice to see so many responses.  To kind of answer some of the questions:

1. Router - ASUS RT-AC66U (I've read many forums that state this thing "runs like a champ" regardless of how it is.  However, I figured why take the risk.

2. I connected two fans in series (the second of the fans was the same physical size as the one above, but rated at 12V and lower current 0.14A) and, as predicted in one of the replies, the voltage drop split was different.  Curiously, this 0.14A rated fan had the lower voltage drop (about 4.6V, with the first fan taking the rest).  I am surprised because the first fan is a new fan from a 2.5" SSD drive cooler and the second fan was from a really old PC (the kind with a third wire from a Hall sensor in the fan), and so I thought the first one would be closer to spec.  However, the second one actually did draw about 150mA at 12V and the first, like I said, only drew about 50mA at the same voltage. Not having experience with computer fans other than plugging the in and leaving it at that, I'm not sure if this is something completely unexpected.

Because the second fan worked at around 5V @ 50-60mA (though not at a high CFM, I'm sure), I'm going to just hook into the USB port--though I believe I will have to implement one of the suggestions in the thread with the first fan because, ironically, the second one is actually going to be too loud at 12V.

[Mr.B]

If the 19VDC power supply can handle the load of both the router and the fan then I would just use the 7812 regulator.
Power dissipation of the 7812:
19v - 12v = 5v
5v * 170mA = 0.85W
Might want to put a small heatsink on the 7812.

[johansen]

two of them in series.

[bills]

simple just series some si diodes .6 volt drop each. 1n4007's less than $1.00 and your done.

[cyf]

@bills--Forgot to mention that I wanted to limit the parts to the items in the kit.  Haha, yeah, I'm cheap, but I figure I will never use the parts unless it's for stuff like this--I've had the kit for over a year.

two of them in series.

Ya mean 2 x fans, I presume.  That could be another option.  I presume they don't have to be the same in current draw, just close would work best.

If the 19VDC power supply can handle the load of both the router and the fan then I would just use the 7812 regulator.
Power dissipation of the 7812:
19v - 12v = 5v
5v * 170mA = 0.85W
Might want to put a small heatsink on the 7812.

The power supply is rated at 1.56A@19V and ~33 watts, so I think it's okay.  I throw something together if I can't find another similar fan--and I  have a heatsink somewhere too.

[cowana]

I presume they don't have to be the same in current draw, just close would work best.

If they have different current draws, they will share the voltage unequally.

If you were powering two series fans from double the individual rated voltage (so say two 12v fans off 24v), it's essential they are the same model, or one could get 15v and the other 9v.

For your application, a slightly uneven voltage distribution is acceptable as long each sees 12v or less - so  12v:9v ratio would be okay. This corresponds to very roughly requiring fans with currents within 25% of each other.

[Tandy]

You may find that there is not enough power in the power supply to supply both the fan and the router. Especially if it is a cheap domestic router that is running hot that you are trying to cool. The manufacturers will have supplied a PSU that as cheap as possible and that means it is probably only something like 300mA and the router use close to that amount. Adding a fan that draws 100mA could be to much for the PSU.

[Monkeh]

You may find that there is not enough power in the power supply to supply both the fan and the router. Especially if it is a cheap domestic router that is running hot that you are trying to cool. The manufacturers will have supplied a PSU that as cheap as possible and that means it is probably only something like 300mA and the router use close to that amount. Adding a fan that draws 100mA could be to much for the PSU.

Nearly all these units will come with a 500mA or 1A supply for a device which consumes an average current in the tens of milliamps.

[Tandy]

Nearly all these units will come with a 500mA or 1A supply for a device which consumes an average current in the tens of milliamps.

Unfortunately the OP doesn't say if it is an ADSL router or Ethernet/Cable router but if it is an ADSL router the modem part is quite power hungry.

[Muttley Snickers]

The OP previously declared that the power supply was 1.56A at 19V, but we do not know what the router is drawing at this point in time, I would keep it simple and go with a single fan but would add a pot or resistor to limit current to the fan as well as a flow or protection diode.

Also a filter capacitor across the feed to minimise fan generated noise may be a good idea.

Good ideas had by all, nice easy thread.

Muttley

Important Update:

I just did a quick test with two fans in series across a battery all 12 volt, and my fans don't work in series, yours may not either, we have been using them singularly or in parallel in the past.

[Tandy]

The OP previously declared that the power supply was 1.56A at 19V.
But we dont know what the router is drawing at this point in time,

Well spotted, I missed that, I don't imagine that the router is using anywhere near that unless it has an onboard Hard Drive so probably reasonably safe to put one of the fans on it.

[Boltar]

How about using the 555 in a basic PWM configuration? 19V DC to 12V rms? I think that equates to a charge/discharge resistor ratio of about 60/40.

[mariush]

Just use a 7812 or a  *1083   ... linear regulators. You may get by with just a single regulator and a small output capacitor. 

19v in, 12v out with linear regulators gives you about 63% efficiency. Switching regulators would go up to 90-95% but there's more parts involved (diodes, resistors. capacitors, inductor) and expensive chips ... the cheap mc34063 and similar switching chips won't give more than 70-75% efficiency so you won't do much better than a linear regulator

Another option would be to simply source a 24v fan, there are such things available.

[mrkev]

Just use a 7812 or a  *1083   ... linear regulators. You may get by with just a single regulator and a small output capacitor. 

19v in, 12v out with linear regulators gives you about 63% efficiency. Switching regulators would go up to 90-95% but there's more parts involved (diodes, resistors. capacitors, inductor) and expensive chips ... the cheap mc34063 and similar switching chips won't give more than 70-75% efficiency so you won't do much better than a linear regulator

Another option would be to simply source a 24v fan, there are such things available.

I agree. For those currents its probably waiste of time.
But anyway. Even cheap mc34063 can go up to 85% efficiency if properly build. It has better efficiency in step-up mode, but that's because they tried to make it more versatile... 70% eff. is way below what that thing can do...

[sleemanj]

Using the parts only from your list given...

Very rough and ready, take one of the trimpots (50k probably best), a resistor and a PN2222.



Put a meter across the fan, tweak trimmer until you get 12v.  Provided the fan only draws your 50mA, it should be within the capabilities of a 2n2222 to handle that. 

Note in the unlikely event the fan short circuits, a single transistor won't be able to handle the dissipation it, but you could simply connect two transistors "piggybacked" in parallel which should work well enough for the purpose in a rough and ready style :-)

A diode antiparallel across the fan might be wise too.

[cyf]

Using the parts only from your list given...

Very rough and ready, take one of the trimpots (50k probably best), a resistor and a PN2222.



Put a meter across the fan, tweak trimmer until you get 12v.  Provided the fan only draws your 50mA, it should be within the capabilities of a 2n2222 to handle that. 

Note in the unlikely event the fan short circuits, a single transistor won't be able to handle the dissipation it, but you could simply connect two transistors "piggybacked" in parallel which should work well enough for the purpose in a rough and ready style :-)

A diode antiparallel across the fan might be wise too.

Thanks!  This looks like it may be the way to go if USB hack (see updated original post) doesn't bring the temperatures enough.

[amyk]

Depending on the exact fan it might be fine with running at 19V...

[tszaboo]

If the 19VDC power supply can handle the load of both the router and the fan then I would just use the 7812 regulator.
Power dissipation of the 7812:
19v - 12v = 5v
5v * 170mA = 0.85W
Might want to put a small heatsink on the 7812.

If you have problem dissipating power in a linear regulator, you just put a good sized resistor on the input of the linear regulator. In this case, an 1W+ rated 20 Ohm resistor will dissipate up to half a watt, which is not dissipated in the linear regulator. Much nicer than having a heatsink, usually cheaper also.
http://www.nandblog.com/just-a-few-more-milliwatts-please/

[Monkeh]

Depending on the exact fan it might be fine with running at 19V...

Or it might pop quite easily at around 15V. Voice of experience here..

[cyf]

Depending on the exact fan it might be fine with running at 19V...

Or it might pop quite easily at around 15V. Voice of experience here..

Believe me, the thought had crossed my mind more than once to see what would happen if I cranked up the voltage, but I also realized that even if it worked, it would be way too loud.

[mrkev]

Since sleemanj draw in fact a current limmiter, you may be able to change a speed of that fan.

To my mind came idea of basic voltage stabilizer, since you have those zener diodes. It will give you 12,5V at the output, so you could use more than one fan, but it won't hold if you short the output... R8 is optional.

[compet17]

Cheapest solution:

Hook your 12V fan to the 5V supply.

I did this with my Ethernet switch (additional fan) and it is great because the fan turns slow and produces no noise at all. If you take a big fan it will produce enough airflow even at 5V.
Some 12V fans do not work at 5V but most do perfectly.

[Zero999]

You can build a cheap SMPS using a couple of transistors:
http://www.romanblack.com/smps/smps.htm

[Simon]

------Updated as of message 15 - Posted here because I like it when OP's update their original posts to make things easier to find.------------
https://www.eevblog.com/forum/beginners/cheap-approach-to-down-convert-19v-to-12v-to-power-a-computer-case-type-fan/msg530519/#msg530519------

Hey guys.  Wow.  Nice to see so many responses.  To kind of answer some of the questions:

1. Router - ASUS RT-AC66U (I've read many forums that state this thing "runs like a champ" regardless of how it is.  However, I figured why take the risk.

2. I connected two fans in series (the second of the fans was the same physical size as the one above, but rated at 12V and lower current 0.14A) and, as predicted in one of the replies, the voltage drop split was different.  Curiously, this 0.14A rated fan had the lower voltage drop (about 4.6V, with the first fan taking the rest).  I am surprised because the first fan is a new fan from a 2.5" SSD drive cooler and the second fan was from a really old PC (the kind with a third wire from a Hall sensor in the fan), and so I thought the first one would be closer to spec.  However, the second one actually did draw about 150mA at 12V and the first, like I said, only drew about 50mA at the same voltage. Not having experience with computer fans other than plugging the in and leaving it at that, I'm not sure if this is something completely unexpected.

Because the second fan worked at around 5V @ 50-60mA (though not at a high CFM, I'm sure), I'm going to just hook into the USB port--though I believe I will have to implement one of the suggestions in the thread with the first fan because, ironically, the second one is actually going to be too loud at 12V.

Although you will confuse the hell out of people if you don't actually reply as well

[bills]

CYF.
I gave you the simplest and cheapest way to do what you need at the beginning of this thread.
If you and the op's wish to make it harder than necessary so be it. although some interesting designs have been submitted.   
no disrespect intended.
bill