Hi everyone
I'm new to the the forum and to electronics in general, I am used to deal with computers but when it comes to single electronic components I am almost completely ignorant..
That said, the problem I'm facing now is:
I have a few server fans rated 12V and 3.24A, I need them to run at around 8V and 10V (no matter to be accurate, any voltage within this range should be fine), I have a lot of resistors, and about 20 wirewound ones rated 7W and 10W, but since I also have around a hundred of various diodes, I was wondering if it was possible to use them (if necessary i will post the codes, but it's gonna be a quite long list) instead of the resistors to have a more efficient voltage drop.
So, ideally, which diode would I need to drop 2/4V in a fan rated 12V and 3.24A?

 

I guess something like this: http://www.digikey.com/product-sear...=1&stock=1&quantity=0&ptm=0&fid=0&pageSize=25

Drops 4.04 volts. Lets 15 A though. Since you need to pass only 3.24 A, it should work just fine.

There are probably dozens of others.

 

What ever you use to drop 4V at 3.25A will dissipate 13W, which will get stinking hot. If you use a wire-wound resistor, it better be rated for at least 15W.

Can you mount it in the airstream so that whatever it is is cooled by its own airflow?

 

What value resistors? You may need to put a few of them in parallel to accommodate the wattage. I've used dropping resistors to quiet noisy fans, before, and I was surprised at how much wattage was needed.

Less sure about the diodes, unless you've got some really big rectifiers, and then you'll probably need about six of 'em.

 

well, thanks all for the quick replies
about the diodes, I apologize if I haven't been enough clear, I posted the question just because I have bought this "lucky bag" (second picture,s011) and, since I know that they could be used to drop the voltage I was hoping if there was any usable, but considering that I am completely ignorant about and the current is quite high, I'm probably gonna use the resistors instead.

I was thinking to use one 0.56Ω 10W to drop ≈2v, but also I wanted to try and put 2 of them in series to get ≈3,5/4v drop...will the power rating be enough in this second case?
Anyway, the resistor will be placed in the airstream (that's quite consistent), also note that the fan is controlled via pwm with an external pwm controller, basically the problem is that when I reduce the duty cycle of the pwm to ≈30% the fan suddenly switches off, instead I need it to keep a slightly lower speed, that's why I thought to reduce a little the voltage..I don't know if I have been clear..

 

I don't see that dropping resistors would help if the problem is that the fan stops running at a low duty-cycle. Unless you're doing something different than it sounds like, you're just reducing the amplitude of the pulses by including the resistors (or diodes).

 

I don't see that dropping resistors would help if the problem is that the fan stops running at a low duty-cycle. Unless you're doing something different than it sounds like, you're just reducing the amplitude of the pulses by including the resistors (or diodes).

trying to be quick I'll list the infos collected until now:
1. The fan is Reverse PWM rated 12V, 3.24A and at full speed spins at 5500rpm.
2. It works at 8V, keeping a speed of ≈4000rpm, further lowering the voltage makes it run in a strange speed up-slow down cycle, probably because the embedded PWM DSP circuitry is not designed to work with a too low voltage.
3. Controlled by PWM it can be slowed until ≈1500rpm, after it stops completely.

So, I was thinking to combine both the PWM control and a slightly lower voltage to get a speed of 1000/1200rpm, at a bearable noise level.
As I said before, my electronics knowledge is quite poor, so I was trying to figure out an intuitive solution with what I have and I know.
Sorry if, considering the title of the thread, I'm actually going a bit OT.

 

Can you put the diodes/resistors between the PWM and the fan? That's the best place for them, and you might be able to get slower than you think, since you won't be depriving the PWM module of its supply voltage.

 

Can you put the diodes/resistors between the PWM and the fan? That's the best place for them, and you might be able to get slower than you think, since you won't be depriving the PWM module of its supply voltage.

mmm, I haven't thought about it, but consider that the fan embedded circuitry is REVERSE PWM, so at 0% duty cycle it runs at full speed, at 100% it stops (actually at around 70%, when I said 30% before I was simplyfying the reverse thing)...also, I don't know which is the current in the pwm signal, but probably one or two 1n400x diode (the only one I have ever used that I know drops 0.7v) would do...still, anyway, even if the fan would work with a PWM of, let's say, 4v, I think it would probably increase the speed..

 

Since PWM effectively controls the average supply voltage that the motor experiences, you may find (depending on the smarts inside the fan) that with the supply dropped to 8V by resistors/diodes the PWM still won't let you go below the 1500rpm cut-off point.
Btw, the 1n400x diodes are rated for 1A, so you'd need to parallel them to handle the ~3A motor current.

 

coming in a bit late here, 3.24a seems a little high to me unless the fans are bigger that what i'm currently thinking about

3 scenario's

- are they powered by the motherboard? (i know some hp server boards that have them - very noisy) slowing them or shutting them down turns off the server.

- individual chassis mount taking power off the motherboard? may be able to control speed inside motherboard bios. most current motherboards have that facility.

- stand alone, speed probably controlled by a temperature sensor on the fan

 

#Alec_t

I was suspecting something like that, but is the only chance I have to reduce the speed below 1500rpm at this point, I hope the fan circuitry is not smart enough!
Anyway, I was considering the 1n400x diode option to eventually reduce the 5v of the PWM signal (I guess lower than 1A) as to3metalcan suggested, but considering the Reversed Pwm problem is not a viable alternative.

#tom_s

The fans are some huge Delta 120mm X 35mm , and the current is actually 2.6A at full speed measured by multimeter (but when I put the multimeter in series to measure the current the speed is actually a bit lower than the full speed without multimeter)...for further information here there is the link to the original thread I started on Overclockers.com, it's a quite long and messy thread, but you should get a clear idea reading the last post.

 

coming in a bit late here, 3.24a seems a little high to me unless the fans are bigger that what i'm currently thinking about

3 scenario's

- are they powered by the motherboard? (i know some hp server boards that have them - very noisy) slowing them or shutting them down turns off the server.

- individual chassis mount taking power off the motherboard? may be able to control speed inside motherboard bios. most current motherboards have that facility.

- stand alone, speed probably controlled by a temperature sensor on the fan

Probably fans from automotive industry. Not computer fans. 7

 

I have tried several resistor, and the best option is a 10ohm (24W, actually I have put 2 X 7W and 1 X 10W in series because I didn't have a 20W 10ohm res), simply because if I put one with higher resistance the server gives me an error at the start up..the problem is that with a 10ohm resistor the fan is still loud (even if a little bit less), the max speed is considerably reduced, but the min speed change only of a few rpm...
I have noticed that with a 30ohm resistance the minimum speed is quite acceptable (so, to answer to Alec_t, it's possible to go far below the min cut-off speed imposed by the pwm), and once I am in windows the server doesn't care anymore of the speed of the fans (I have even completely switched off the fans, no problem at all, in idle of course)...
does anyone know some kind of switch that allows me to bypass the three indipendent 12V wires through three parallel wires with the 30Ohm resistors (I forgot to say it before, the server controls 3 fans)?
in this way I can keep a medium speed through the pwm during the start up (or during full load operations), and lower it to a minimum level when I am in windows and not under heavy loads...I can draw a quick sketch if it's not clear..

 

Sounds like a 3PDT switch would do it, really. Or do you need electronic control?

 

no, I just need a very simple switch, that with one click opens the three 12v lines, so the electricity is forced to pass through the resistors...the problem is that the 12v are indipendent, so i can't use a simple on off switch...
looks like the one you said is fine, the three lines doesn't interfere each other, right?

 

Way late to be saying this but, a centrifugal fan is a lot quieter than a propeller fan for the same CFM. If you're dead serious about noise, use a centrifugal fan.

 

I've been assuming we're stuck with the fans we have, because the OP hasn't mentioned the possibility of replacing them...adelage, is that the case?

As to your question about the switch, yes, it switches three independent lines.

 

sorry for the absence guys, 7 to 22, long shift at work today...anyway, I can't replace the fans (but the suggestion is still well accepted, as I often deal with cooling stuff), the server is quite fussy and gives troubles even if I replace only one fan, I had to do a very tricky job to obtain the current configuration, I posted a link for whoever is curious yesterday

...for further information here there is the link to the original thread I started on Overclockers.com, it's a quite long and messy thread, but you should get a clear idea reading the last post.

Anyway, thanks for the 3pdt suggestion, is it perhaps the switch normally used for the guitar effect switch? I can't find it on the maplin website (place were I use to buy electronic stuff, note I live in the UK)...

 

Haha, yes, but that vacuum-style push-button isn't the only form-factor available...you know, unless you just want to control fan speed via a foot-pedal. Try "3PDT Toggle."