View Full Version : High wattage fan controller options

05-14-2003, 01:45 AM
Alrighty, I tried searching but no luck.

Anyways, does anyone sell a really basic PWM kit? Either that, or a rheobus kit capable of handling VERY high wattage. I've got a pair of Comair Rotron 172mm fans, 26Watts each. I can run these beasts at 7v and they are pretty quiet, while sounding like [and pretty much producing] a tornado at full power. I'd like to get the very most power outta these as I can, but nobody seems to make a product that can handle all the wattage they take.

I've seen the reviews of Aardil's PWM, but I'm not too interested in temp control, I'd much rather be able to manually control. If you can manually control fan speed with his, I spose I'd get that. I really don't want to make a PWM, cause my soldering skills are pretty much worthless. It took me about 5min to solder two LED's together, and even longer to solder a couple split leads to a pair of those LED's. Any complex circuit would probably take me all summer to assemble, and I've got better things to do than that =]

Thanks for your help everyone!


05-14-2003, 03:21 AM
Try Hobbytron if you're looking for a PWM with 'balls'. Their CE-CK1400 will handle 5A @ 100VDC and has a pot for speed control. You would have to solder it. But, at $15 US (If I remember correctly) it includes a PCB and all the parts (which number about 15) and would be a good project to learn soldering on. Its got discrete components, an IC, a couple jumpers, all kinds of different leads for you to learn on! Besides, to buy it pre assembled would probably be about $20-30 US, and methinks you aren't going to spend that much on this simple of a project.

06-29-2003, 07:42 PM
Re PWM, ideally choose a circuit where you can vary the
PWM frequency - Pulse Width Modulation frequency.

PAPST make 2 PWM controllers:
o PCM001
o System 3000

They are expensive, and neither have the facility to
vary frequency - which in some cases is *vital*.

Well, the frequency relies on slippage of the fan, to keep
turning whilst power is removed. That is fine, but obviously
this is a characteristic which will vary between fans & makes.

As the fan gets slower, on some the PWM frequency ends up
creating a "growl" which completely negates fan noise reduction.

For example the PCM001 does this with PAPST 6212NM fans,
a 174mm 12V fan of 220-235cfm. It growls horrendously at
lower speeds completely negating the benefit of PWM.

Similarly, the PCM001 whilst ideal for 120mm papst fans is
utterly useless with panaflo 120mm - they growl terribly.
Different bearing & weight, different slippage etc.

Many DIY circuits DO allow you to vary the PWM frequency.
Sadly a 8-pin chip did exist to do most of this, for ~2.50$US
but it was discontinued at least in Europe if not USA also.

Could have made a very good unit with variable frequency.
Just something to consider in achieving real silence.
Dorothy Bradbury
http://homepage.ntlworld.com/dorothy.bradbury/panaflo.htm (Direct Prices)

07-05-2003, 02:51 PM
Well, I did find something that I thought would work: A voltage regulator circuit. I bought an LM338K, which is a 5A, 1-35V regulator, similar to the famous LM317 series except it handles more amperage. I used the following guide to assemble it, but left out the capacitors, as I didn't have those, and as best as I could gather, they aren't necessary.


Anyways, I finished the circuit, and plugged it in. It DOES vary the voltage, from about 3-13V [plugged into a 14.2V PSU] but it only produces mA, so I can't power the smallest fan I own with it, let alone my Comair Rotrons. Anybody here have an idea? I've got another spare LM338, so I was going to buy those capacitors the guide recommended and try making another circuit.

Thanks for your help!


07-05-2003, 03:25 PM
You need special fan controlers to handle high watts?

I'm only running 6x80mm fans through mine ATM but I'm going to be upgrading the fan count to 6 SETS of fans (3x120, 2x120, 2x80, 4x80, 2x80, 2x80).... When I get back home.

The thing looks pretty simple swaping from the 12v line (adjustable) the the 5v line(non adjustable) so it should be OK.... right?

Incidenat;lly you'll never know I'm on water cooling with all these fans ;),

07-05-2003, 08:33 PM
You're going to be fine if your controller is just a 12/7/5V alternator, unless they used some really shoddy switches. Even small switches are rated for a lot more current than all those fans put together.

07-05-2003, 11:10 PM
Aye I thought it would be fine... but with you saying that high wattage switches got me worried a bit cause my fans would load my PSU by about 80W in total at full power.

07-11-2003, 02:36 PM
Originally posted by Boli
Aye I thought it would be fine... but with you saying that high wattage switches got me worried a bit cause my fans would load my PSU by about 80W in total at full power.

Most switches I've seen handle 10A at 110V, which is way more than what you are using. Incidentally, I soldered on my other LM338K and now my circuit works, controls my fan's voltage from 1.25-12.5V, exactly as I wanted it to work. I've just got to attach a heatsink on that bad boy [I cut an OEM Athlon HS in half for that] and I'm set.


07-11-2003, 03:32 PM
Incidenatlly I have routed all my fans through my controller (total 50W infact) and the only thing I have noticed is it gets rather warm. But since I have run it for about 60hours none stop I've given up worrying about it.

07-11-2003, 03:49 PM
Stick the PWM heatsink in the airflow.

50W PWM solutions are around 50-90$US, so nice solution :-)

o If you run it off a PC PSU tail, fuse it but add in a relay with
diode protection to run a 12V beeper for fan (fuse) failure.
o Yes fans going silent should be a good cue, could be missed.

Also, if you can vary your PWM frequency, do so because you
may be able to knock a few more dB(A) off your noise levels.
It can certainly affect the qualitative aspect of PWM fan noise.
Dorothy Bradbury
http://homepage.ntlworld.com/dorothy.bradbury/panaflo.htm (Direct Prices)

07-11-2003, 04:17 PM
You CAN directly vary the speed with Aardil's controller and just unplug the thermistor. It's certainly a decent controller, but as punisher and many others have mentioned, there are alternatives.

For a 26W fan at 12V, that's just a hair above 2A. That's not a huge amount of current for hobby kits to handle. At 60% efficiency, you'd be wasting somewhere around 10W in heat, so keep that in mind. You can get higher efficiency PWM solutions, but I'm guessing that most of them trade off efficiency for "cost effectiveness".

Oh, and I'll give a hearty second to the suggestion to require the ability to vary the frequency. A frequency that works fine on some fans won't work at ALL on other fans, or it might merely sound like a banshee on crank.

07-26-2003, 02:38 PM
In case anyone is interested in making a fan controller like mine, here's some info about it. I basically followed Uller's guide to make the circuit, but since I didn't have any capacitors, I left those out [it works fine anyways] The only real problem is the fact that the LM338K is only available in the metal package, not the plastic, smaller one that the LM317 is. At least I couldn't find the LM338 in that package.

Also, you would need to actively cool the heatsink for this circuit. I left my 26W Rotron on that circuit, running at 7V, for about 2 hours, and the heatsink was burning hot. It's not the biggest HS, just half a cheap Athlon OEM one. Seems odd to have to add more fans to your case when you end goal is to reduce the noise of your existing fans =]

Here's some links to things:

Uller's Guide: http://casemods.pointofnoreturn.org/vregtut/
Info on LM338K: http://www.national.com/pf/LM/LM338.html
Link to buy an LM338K: http://sales.goldmine-elec.com/prodinfo.asp?prodid=8586

Hope somebody finds this useful =]


07-27-2003, 11:31 AM
The reason you needed a heatsink is that linear regulation is extremely wasteful of electricity. Running at 7V, you waste almost as much electrical energy in your regulation as you are using to run your fan! This is why I wanted to use PWM.

BTW, props to Uller. He's the one who helped me the most when I was learning electronics. He wrote that linear regulation article, then I wrote my fan regulation article with some input from him, and then he wrote another article on just PWM fan regulation.

Smart guy, and it's too bad I don't hear from him more often. He posted a lot on the HardOCP forums, but (last I checked) he stopped being such an active poster. He must have an ultra-busy RealLife(tm) schedule.

07-28-2003, 02:38 AM
I realised that linear regulators would require a heatsink, which isn't that big of a deal for me. I went with VR's because that was my first real soldering expedition. I wanted something easy to do so I could learn how, before doing something more tricky like a PWM. Plus, I think I heard somewhere that Comair fans growl a lot when you use a PWM. Don't know if that is really true, but they seem pretty nice and quiet with my LM338 circuit at lower voltages.


07-28-2003, 06:54 AM
Which fans growl & don't growl with a PWM is almost a black art:
o It will vary between speed specs, between makes & sizes
o It will vary between bearing types
o It will vary with PWM frequency used

Most of all, mix fans on the PWM rail, and some will growl and
some will not - finding a balance when all don't can be impossible.

You can't extrapolate from 1 fan to another even next to it in
the range with regard to noise on VoltageReg or PWM control.
There are too many discontinuities - very frustratingly :-)

So if a PWM is bought...
1) test it with your fans, don't assume it will or will not be silent
2) try to get one with variable frequency if you can't do "1)"

PWM is chopping voltage up into bits, and relying on the very
mechanical aspects of the fan to keep turning AND the very
individual IC motor drive circuitry to not add other noise when
it essentially has a non-DC power applied to it (not designed).

Amusingly (or not) it's quite common to find a companies own
PWM solution utterly unusable with fans it cites as usable. For
example the PAPST PCM001 is *unusable* with 6212NM fans,
as it gets to ~60% running it growls louder than full-rpm noise.

So you can't assume anything about an entire fan range, or a fan.

Have fun :-)
Dorothy Bradbury
http://homepage.ntlworld.com/dorothy.bradbury/panaflo.htm (Direct)

08-16-2003, 03:10 PM
This 15W Rheostat (http://www.allelectronics.com/cgi-bin/category.cgi?category=470015&type=store) is the one I use for my 172mm fan. Mine dosn't draw as many watts as yours but I got it just to be safe, and so I could add more fans to it.

08-16-2003, 04:20 PM
Some notes when using 171mm fans.

o Their asset is not in their diameter
---- the cast alloy shroud is 171mm
---- the blade area is considerably less - below 150mm
---- hence the "shroud shrunk" 171x150mm versions

o Their asset is in depth & so static pressure
---- most are 51-55mm deep, some NMB are 38mm deep

Their cfm may not be impressive re size - 200-235cfm,
what is impressive is their ability to *maintain it*.

o Whilst 2x 120x25mm fans can match 171x55mm in cfm
o It takes 4x 120x25mm fans to match 171x55mm in pressure

Integration issues must be considered:
o 171mm fans are built for 1) Life 2) Pressure 3) CFM
---- noise considerations are mute to economic buyers
o Whilst 40-120mm fans dB(A) scales with rpm, 171mm do not
---- 171mm at 100% duty 51dB(A), 50% cycle 35-43dB(A)

Noise is thus a fundamental concern when using big fans.
o Least - motor IC noise (DC switched to make fan spin)
o Notable - depth for static pressure increases noise
o Major - bearings for heavy rotor magnets
---- DC fans have stationary motor coils
---- motor IC switch the DC around at N frequency
---- very big magnets in rotor hub chase the DC

Bearings are your major concern for PC applications:
o PAPST / Comair Rotor = "Rough"
---- major component of noise up to 65% Duty
---- an increasingly loud rumble, noisy stainless bearings
o NMB = "Smooth"
---- hard to get hold off
---- 6820PL-05W-B10 for 24V & B10-B50 if bothered

The solution for silence afficinado's is voltage:
o Choose the 24V fans
o Then run them at 12V to get the noise & load down

Remember for big fans the wattage is non-linear:
o 24V fans often have a wide range 12-32V DC
---- note the range is as *your* fan however :-)
o 16V = 4.9W, 32V = 17.0W
---- so running at 12V is actually under 4W
---- pressure is good, major noise is bearings

An unfortunate note about bearing noise is that it is
rarely "white-noise". Most A/C units you hear are designed
so the predominant noise from a 300-400cfm+ fan is airflow.
Airflow noise is a whitenoise, favoured because human
perception can most easily mask that at the perception level.

PWM with big fans (big = depth) gets problematic:
o Bearing noise is high
o Slippage is high due to bearing & big magnets & static load
o PWM frequency that is 1-2Khz may be objectionable
---- however it will be quite efficient
o PWM frequency that is 10-20Khz may not be noticed
---- however it is often less efficient

Efficiency comes down to the fan's motor design & loading.
o So wattage need not be great if you run a 24V fan at 12V
o Run bigger fans at 60-70% anyway re Noise & Longevity

08-16-2003, 11:43 PM
Great info jabf2000. It seems you know a LOT about this stuff =] Just so you know, my fans don't seem to have any growl or excessive noise besides the noise from all the air they push when I run them on my voltage regulator.

Cheese: Did you get the topmost of those two rheostats, or the bottom one? That bottom one is a real monster, I wonder how big that is, and if it would require active cooling =]


08-17-2003, 08:20 AM
Running on a regulator, you should have no problem.

They are useful fans for high-resistance enclosures, eg, radiators.
A pity there are no plastic housing variants, re sheer weight.

Just watch the fingers :-)

08-24-2003, 01:30 PM
Originally posted by CheeseBall This 15W Rheostat (http://www.allelectronics.com/cgi-bin/category.cgi?category=470015&type=store) is the one I use for my 172mm fan. Mine dosn't draw as many watts as yours but I got it just to be safe, and so I could add more fans to it.

I only wish I could find a 20watt Rheostat, 25watt Rheostat (http://www.surplussales.com/Potentiometers/PotsRheost.html) (RWA-NH25100250L; 25/100/250 Ohm 25watt 1/4"Shaft $25 P57) is as close as I could find, I wanted to hook up two 80mm Orange Smart Fan 2 fans to 2 Rheostats(2-fans to a Rheostat) as The fans use double the wattage on startup (4.8/9.6x2=19.2watts) I've been told. And then mount the 2-Rheostats in a blank that already has 2 switches for My UV lights(www.case-mod.com/). The Rheostats would be for when I finally water cool really.

08-25-2003, 05:45 AM
Wow, those are really expensive! You'd be better off just building a circuit like the one I created. You could get the parts for that for about 10 dollars.


09-17-2003, 02:37 AM
here's the PWM circuit i designed that handles 48W. they are thermally controlled too:

Thermal PWM Circuit (http://www.strongshock.com/pwm.htm)

if you're circuit savvy, you could replace the thermal control portion with a poti for manual speed control. the frequency is adjustable, and if you use the thermal stuff, the voltage per degree value can be set sky high so it actually works for PC cooling. btw, you could also keep the display if you go manual, so you can have a neat throttle indicator.

09-17-2003, 09:38 PM
Those look very nice, and good job explaining your thought process too! Sounds like you put a LOT of time and energy in coming up with these, but they sound like they can pretty much do whatever you might need a fan controller to do =]