If I wanted to do a fan installation in my player, but didn't care if it was thermistor controlled, could I make it very simple? For instance, like this:

Fan_pin----Variable_resistor-----Fan-----Ground

Then I could use the variable resistor to set the desired fan speed and it would just run quietly whenever the player was on. There's no voltage regulator or anything in this design.

Anyone see a problem with doing it that way?

Seems ok to me. Where are you putting the pot.?

Stu

Most variable resistors are low wattage, and this could be a problem:

Assuming that the fan header is 12V, and that the fan draws 30mA at 6V. (Purely hypothetical figures here...) The the resistance needed would be:

(12 - 6)/ 0.03 = 200 ohms.

The power dissipated in that resistance would be IR = 0.03 * 200 = 6W.

Another calculation. Assuming you want your fan to run at 8V, and it draws 40mA at 8V:

R = 4/0.04 = 100 ohms. power= IR = 0.04 * 100 = 4W.

Where are you putting the pot?

I keep it stashed in a jewelry box in the hall closet. HEY WAIT A MINUTE....

All kidding aside, I haven't decided on where I would put the pot. Just wherever would fit, I suppose. If it's small enough, perhaps the security slot would be the right place.

Since all the power for the fan will be going through this pot, does it matter if it's a tiny little thing? Does it have to be a heavy-duty pot or anything?

Okay, what if I just messed with some ordinary resistors and chose a value that I liked? Would it melt the resistors eventually?

If power is a problem, go with a 5 watt wirewound pot. or even a rheostat. I don't know how small the stats come, but they are usually higher power handling than pots.

Stu

Yeah, you could do that, but actually this discussion adds an interesting slant to cooling the empeg in general. The empeg consumes a little over 6W when operating with disks spun down. All of our slow fan solutions so far discussed are putting 4-6W of additional heat into the case. Does anyone else find it a bit perverse that 40-50% of the heat being removed is generated by the cooling solution?

Maybe it's time to look at DC-DC switch-mode converters, or switching regulators.

All of our slow fan solutions so far discussed are putting 4-6W of additional heat into the case.

Are they really doing that?

The power dissipated in that resistance would be IR = 0.03 * 200 = 6W.

No it wouldn't. P = U x I = R x I x I = U x U / R

So in this case we have 0.03 x 0.03 x 200 = 0.18 W

Sanity check: the pot drops 6 volts (12-6) with 30 mA current going through it, which gives U x I = 6 x 0.03 = 0.18 W

Beware that a fairly small (depending on what value the pot is) part of the pot might be generating the heat - not the full track of the pot... Select pot size with that in mind.

/Michael

Aarrgh. I'm losing my mind. You're so right. How could I get such an elementary equation so utterly wrong?

I've obviously spent too many early mornings and even earlier mornings recently watching the World Cup...

Heh, easy enough to slip up... Not like I've never done it

I'd probably use a pot initially, to find a value where the fan is acceptably quiet and still moving enough air, then replace it with a fixed resistor for final wiring. Then a 1/4W resistor would be enough in the example above - might go with 1/2W for better margins...

/Michael

This might sound stupid but, if you could find a 24v fan that size it would run quietly on 12v. You could also use a fixed 8 or 10 v regulator transistor, but then your back to the generating more heat prob.

Ben

This is just the setup I had immediately I discovered the fan was so noisy when directly connected to the 12v supply. As the others have said you just have to bear in mind the power that will need to be dissipated by the resistor.

Something else to bear in mind is that the quoted current consumption of your chosen fan will be at it's designed operating voltage. As you're going to be running it at a lower voltage your current drain will also be lower. You will probably find that you'll do all the maths (or math for you lot on the other side of the pond ) and it won't work out exactly how you thought. I would expect the fan's resistance to be reasonably constant and independant of operating voltage so if you run the fan at half voltage then the current drain will be around half of that quoted (or thereabouts). However, you might want to skip the maths and just experiment.

eg.
Fan = 12v, 50mA
R(Ohms)=V/I
= 12/0.05 Ohms
=240 Ohms

Say you want to be able to vary the voltage between 5V and 12V.
Current drain @5V = 5/12*0.05 Amps (assuming constant fan resistance)
= 0.021A (21mA)

The voltage drop across the resistor at minimum speed will be 7V (12V-5V).
R=V/I
=7/0.021 Ohms
=333.3 Ohms

W=I*V
=7*0.021 Watts
=0.147 Watts

Now then, near max speed say you're supplying the fan with 11V (ie dropping 1V across the resistor):
Current drain of fan = 11/12*0.05
= 0.046A (46mA)

Required resistance = V/I Ohms
=1/0.046 Ohms
=21.7 Ohms

Power dissipation = V*I
=1*0.046 Watts
=0.046 Watts

So the maximum resistance you'll need in this case is 333 Ohms and a max dissipation of 0.147 Watts.

Hope this is of some help picking a suitable pot.

Okay, here's what I did.

- I hooked up the fan to a variable resistor and 12v power.

- I found that I wanted the fan running on anywhere from 300-500 ohms resistance.

- I dug through my resistors pile and found a 470 ohm 1/4 watt.

- It worked, but the resistor itself felt rather warm to the touch after running a few minutes.

- I didn't have anything in 1/2 watt resistors that was close to the 470. But I did have a bunch of 1000-ohm 1/2 watt resistors. I tried both a pair and a trio of these wired in parallel, and they seem to be doing what I want. The trio seems to do the best.

- The trio of 1/2 watt resistors gets only faintly/barely warm to the touch after running a while.

Does this sound like the proper solution, or do I have anything else to worry about?

That sounds fine... 333ohm, 1.5W power capacity. Even I can do that math

I think 333 ohms was too little resistance, but 500 was too much. I added an 87-ohm 1/2 watt in series to bring it to 420 and I think that's about right. I'll give that a shot...

Crap.

/me discovers that the voltage on the "Fan" pin is different than the voltage coming straight from the main power plug.

Okay, once I got everything working the way I wanted it and the cover on the player, I decided that the fan moves too little air, even when the noise is above my tolerance level. So if I'm going to do a fan, it's not going to be in the player itself.

The only reason I'm messing with this is because I'm working on a docking station, and the player tends to get hot when in the docking station. I'll try some other options and see what I come up with.

In reply to:

---

bring it to 420 and I think that's about right

---

Still talking about pot?

Stu

I used a 24V fan from a product at work with no other components. It works perfectly and is inaudible in the car or if the empeg is more than 3 feet away at home.

Edit: Oh and it does work too!

Reminds me of peltiers. I once hooked one up to a heatsink and CPU only to realize that the peltier put out more heat than the CPU. The poor heatsink had three times the work to do, not to mention the extra 45W the PSU had to find.

Needless to say it didn't help.

(I know there are cases where they actually help, but this wasn't one of them.)

Cheers,

Rex.

That's pretty much why I wanted to do a temp controlled fan. That way when it's sitting on my desk in free air (ie not getting all that hot) the fan's running slowly and quietly and when it's in the car (getting somewhat hotter) it's running faster to give more cooling effect. With the player embedded in the dash it can be almost as noisy as it wants to be without me actually being able to hear it.

When I eventually get round to building a docking station I'd like to duct the air out of the box. Ideally I'd like it to go out of the back but I'm not sure whether the fan would have enough power to push the air through a long duct. I expect I'll end up just venting directly through the top panel.

Thing was, with my chosen fan, even at the lowest possible rotation speed, it was louder than I wanted it to be. So a thermistor wasn't going to help.

I'm considering doing a thermistor version in the docking station. That might be tolerable.

In reply to:

---

Thing was, with my chosen fan, even at the lowest possible rotation speed, it was louder than I wanted it to be.

---

Hmmm... that sucks (or blows - depending on which side of the fan you are ). Sorry about that, I couldn't resist.

That must be one helluva noisy fan. Either that or you have super acute hearing!!

FYI the fan I'm using has a quoted noise level of 23 dB(A)/m. How does that compare to yours?

Why not do temp controled PWM?

You could nail the fan with constant 12v for a second (or less) and then use PWM to keep it rotational speed way way down. Ramp it up if the temp goes up.

There are single chip (and near single chip) PWM fan controllers.

This should be able to turn the fan slower than simply lowering it's input voltage. Plus you aren't wasting power (not like with a resistor anyways).

Could do it with a simple micro and jack it into the serial port. Then you could have a userland app or hijack mod which would allow you to control ramp up, thresholds, etc.

I was going to look into it more in my copious amounts of free time. At the least a fan controller IC should be a good option.

-Geoff

I just wanted to mention that the closer the fan is to the holes in the top the louder it will be. If you have any clearance try spacing it away from the top even a little will help. Old sirens used to be made by blades rotating past holes, this could cause some extra noise by the fan being so close.

Ben
mk2 20gig
hijacked

I just wanted to mention that the closer the fan is to the holes in the top the louder it will be.

I was totally aware of this. The thing is, even at the fan's lowest setting, and even with it being moved away from the holes, it was still too noisy for my tastes.

Add to that the fact that, if you move it away from the holes, less air actually gets blown out of the casing. When I moved it away from the holes, I couldn't feel any air coming out of them (fan was oriented to move air up and out of the case).

You may want to check out a couple of links in the other thread. There's a few 8 pin PWM fan controllers that look *very* simple to hook up.

What size are these fans? 40mm x 10mm? I have found some pretty quiet (14dBA) ones in larger sizes. Perhaps I can locate a suitable replacement. Sunon makes a 40mm x 6mm fan that moves 3.5 cfm at 16 dBA and operates on 5 volts. If my calculations are correct (probably not) that would be sufficent to exchange the volume of air in the Empeg ~41 times a minute at full speed.

Stu

I've just commented on the PWM aspect in another thread. I did consider using PWM but I was a bit concerned that it might introduce some electrical noise into the system. Maybe someone with better knowledge of electronics could give a definitive answer. I wouldn't mind doing some experimentation when I get the time though.