Decouple fans and hard disks supplies

Do you have a question? Post it now! No Registration Necessary.  Now with pictures!

Threaded View

Hello respected engineers!
I've built myself a frame where I've put six hard disks for a RAID5
system, and six 8x8 cm fans to cool them down (they live in a quite
hot environment!). The PSU is a 500W one however I was concerned that
the fans may pollute the hard disks' supply too much.. since they're
motors after all!
I've accidentally put a finger into a fan (what a pain!), and this
stopped the fan for a moment and also caused hard disk errors, and
thus convinced me even more that I need to decouple the fans' motors
supply from the hard disks supply.

But how? I'm just a hobbyst when it comes to electronics, so I may
even intuite correctly, but I won't be really sure of my thoughts.
Right now, they are that I need some big inductors (with as low DC
resistance as possible), some (big?) capacitors and maybe also a big
rectifier as well.

The best scheme I've thought so far is to link all 6 fans together in
parallel, then apply power through a circuit like this:

6 fans--*-ind-*-ind-*-ind-*-ind-*-ind-*-ind---diode------*-----PSU +12V
        |     |     |     |     |     |                  |
       cap   cap   cap   cap   cap   cap               6 HDs---PSU +5V
        |     |     |     |     |     |                  |
Gnd ----*-----*-----*-----*-----*-----*------------------*-----Gnd

of course in the above scheme I don't even need them to be six,
it's simply the more the better, I just happened to have six of
them on my bench (more or less a coincidence).

But now the big question: what's the right size of the caps, also
to not overload the PSU at start? Will the diode really help the
hard disks to not get some bad bad bad back EMF?
Notice also that the wire that feeds the fans is AFTER all the wires
that feed the hard disks (otherwise the voltage drop caused by the
fans would have affected the hard disks). Yes, at least I know this. ;)

Is my scheme ok? What can I do to improve it? I really care about
the hard disks and the data they contain!

Thank you very much,

Re: Decouple fans and hard disks supplies

I don't think I would worry too much, unless of course the fans are
the 'carbon
brush'  types which could create a very noisy invironment
due to the arching,
these type's could cause un wanted magnetic fields
which could cause hum in the
audio department or wipe  data from the
hard drive ....!  

'The squirrel cage'
motor type's are better in this respect but tend
to create a high magnetic field
which could also cause havoc as above
due to their large magnetic field, at
least these don't create arching
as they are brushless.

The type's normally
used in computers are the 'brushless DC fans',
true all all  motors require a
magnetic field to convert power into
torque, the the filed in these are self
contained within the rotor,
they tend not to generate any external field, no
sparking being
brushless henve no RFI (Radio Frequency Interferance) which could
into the logic  and audio circuitry.

Modern computers fans require very
little power to run, a typical fan
at the most might only draw between 80 ~
120MA or 0.08 ~ 0.12 Amp, so
we can say VxI = P in Watts, being operated at 12V
works out 1 ~ 14 W
appx.  I know the logic circuitry is the main power zapper,
but feel
1 amp is peanuts....!

Does seem strange stopping a fan with a finger
would cause data error,
I would imagine a stalled fan isn't gonna going to do
anything apart
from an increase temperature rise and doubt that a hard drive is
that sensitive..... a cpu could be, a thought could be given to the
chips in this respect, think I'd be looking elsewhere...

unless of course the
hard disc are close to melting in it's normal
environmental conditions..!


Re: Decouple fans and hard disks supplies

Well, many 8 cm fans have high rpms make lots of noise to push the air.  The
problem with cooling is it is often better to bring in cooler air to heat
the the drives from outside of the case.

Heat can build up inside a case so to properly cool the drives, you need
room for air to circulate between the drives.  This might require a special
case configuration that allows more drives that are spaced so air can flow
between them.  Drives create a lot of heat.  When they are stacked one on
top of each other, they are like a giant metal brick.

You may acutally need more than the wattage you have.  Some power supplies
can not actually create their max voltage without burning up.

I think you can create more air flow with 120mm fans.  They spin at slower
RPMs and move a larger volume of air per fan.  You really need a case that
has a good air flow pattern that can move air unimpeded from the front to
the back.

Water cooling might help in a really hot system.  Another option is using a
motherboard with a Pentium M Motherboard.  There are not many, but some
companies make them.  A-Open makes a few and some cube cases have
motherboards with Pentium M motherboards.  They run cooler and use fewer
watts and need less cooling.  They may run slightly slower but still can
handle a lot of load.

If you have heat problem it may not be the drives.  The chipset or the cpu
or the video cards can all overheat as well as the memory.  Often memory is
real near to the CPU so that can cause it to overheat.

Power supplies can appear to be functioning and actually be fluctuating.
They can also create more heat if they dont operate at peak efficiency.

This is why the total picture is important.  Motherboards sometimes have
problems with chipset fans failing.  This is a common problem, especially if
the system is exposed to high Dust conditions.  One option could be to put
the computer in a booth and cool the booth.  Another option could be to
build the box as a server and have a small client machine access the server
which does most of the work, which is referred to as Thin Client.

Quoted text here. Click to load it

Re: Decouple fans and hard disks supplies wrote:

Quoted text here. Click to load it

Motors have a stall rating, referred to as their "locked
rotor amps". When you stick your finger in the blades of
one of your fans, you cause the current consumption to
climb. If could be that by drawing a lot of current,
you actually overloaded the 12V output of the power

You might consider sticking a fuse on the fan. That
way, the fuse will open in the event of an overcurrent

You can achieve the same results, by using a three terminal
regulator. This device is a 10 volt regulator, and if you
power it from the 12 volt rail, will regulate down to 10
volts. The fan will run a bit slower by using 10 volts,
but not by much.

If you stick your finger in the fan now, the 10V regulator will
cut out when the current limit is exceeded. These regulators
also typically have overtemperature protection, and the
regulator cuts out if it gets too hot.

Another device that can achieve this end, is a simple
Polyfuse. That is a polymer fuse that opens when it gets
hot, and closes again when it cools off. It can be used
over and over again, and these are used on motherboards to
protect the USB, Firewire, and PS/2 power signals.

So, perhaps what you need, is a simple means to limit the
current available to the fans.

The regulation properties of the switching power supply, and
the small amount of capacitance present, should be adequate
to handle the current waveform of the fans. Remember that the
processor current, for example, can vary radically in a small
period of time, so something like gaming can cause more current
variation, in a shorter time, than the fan can.

Another thing to consider, is regulated power supplies have
phase margin stability criterion like any other circuit. If you
add too much capacitance to the output of the ATX supply, it might
begin to oscillate. The following text is from the ATX 2.01 power
supply spec (

  "3.2.8. Capacitive Load
   The power supply should be able to power up and operate
   normally with the following capacitances simultaneously
   present on the DC outputs. This capacitive loading should be
   used to check stability and should not be included for noise

   Table 12. Output Capacitive Loads
   Output       Capacitive load (F)
   +12V1DC      5,000
   +12V2DC      3,000
    +5VDC       6,000
   +3.3VDC      6,000
   -12VDC         350
    +5VSB         350"


Re: Decouple fans and hard disks supplies wrote:

Quoted text here. Click to load it

I seriously doubt you need such an elaborate L-C filter, and a single L
and C for all 6 should be enough for all but the worst motor noise.
Also the diode is unneeded.

Its possible that stalling the fan caused disk errors simply because
the +12V line was overloaded and the voltage dropped, not because the
fan created excess electrical noise.

Quoted text here. Click to load it

Re: Decouple fans and hard disks supplies

On 15 Jun 2006 08:15:24 GMT, wrote:

Quoted text here. Click to load it

It would be nice to know the approximate current the fans
use, running and stalled... as well as any adjustment for
current if you will undervolt the fans to reduce noise.

Note that the current rating on the fan label is not typical
(RMS) current, they will have to be measured with at least a
filter (RC?) in front to get a reasonable reading.

Quoted text here. Click to load it

... or put fan guards on the fans and secure all loose
wires, if not both.

Quoted text here. Click to load it

I doubt you need inductors, nor that if you did they would
need be "as low DC resistance as possible".  In the end the
key is whether the fans provide enough cooling, it is easy
to go overboard on fans, choosing faster than needed which
just accelerates wear, increases dust and noise buildup.  It
is also easy to choose higher (RPM) spec fans and if they
voltage is below 12.0V, their RPM drop can be acceptible.

Quoted text here. Click to load it

Do you know that you have bad bad bad bck EMF?
A capacitor or two alone might suffice.
If you had the current requirements of the fans, a regulator
might do the job too, as might the typical fan bay
controller which is ready made and thus, might not only be
quicker but as inexpensive too.

Quoted text here. Click to load it

Something  else a bit simplier is to use a different lead
off the PSU, and a cap or two if needed.

Quoted text here. Click to load it

In that case, do all of the above.  It's not  THAT expensive
or time consuming.

Re: Decouple fans and hard disks supplies wrote:

Quoted text here. Click to load it

Looks to me like you're trying to solve a 'problem' without knowing what
caused it. For example, when you whacked your finger in the fan blades you
might have also bumped the cabinet and that could be what caused the drive
errors rather than any electrical incident.

Even if it were electrical, you're guessing. Was it a drop in voltage
caused by over current or an inductive spike, or what?

I don't know what size fans you're using, and their current draw, but it's
hard to imagine how one going into stall could over current a 500 watt
power supply and just as hard to imagine how an inductive spike could get
past the PSU's output filter plus the drive's input filter. Not to mention
they're usually current limited with some filtering in the fan itself to
protect the internal electronic brushless motor controller.

If you want to attack this then the first thing would be to characterize
what it is you're attacking. I.E. Put it on a test bench and see what the
effect of a sudden fan stop is.

A simpler solution might be to not stick your fingers in the fan, and I
don't mean that tongue in cheek. I'd think it falls into the same category
as don't kick the case or pull out the power plug while it's running, and
numerous other 'improper operating conditions'. You don't expect fingers in
the fan to be routine, do you?

Re: Decouple fans and hard disks supplies

Quite right Paul......

Just checked for curiousity... a fan I tried draws a
100mA when
stalled drew 220mA, of course all fans will be different - something
I overlooked, sorry folks, read on and you'll see why.

The reason why I did not
consider the point that Paul made, a stalled
fan is not going to draw one heck
of a lot of current, yes there will
be an increase that is obvious with every
stalled motor.... since we
are not dealing with a 'wound' rotor with it's carbon
brushes these
will draw much more current than the DC brushless types, so much
the armateur will burn up..

If the computer is so effected by a such a
minute change in current,
obviously leaves the PSU to question,  but I would
have imagined it
would have been able to cope with this increase..... since ~

no effect had been mentioned when booting up, ie, the hard drive
running or any
problems with the added load when any disc drives are
running or how about when
the CPU and RAMs are fully loaded....
that's when the PSU is stressed
.............. not with a measely

Oh,  and what about the 12V feed for the
motor drives..?


Re: Decouple fans and hard disks supplies

One thing I just thought about and no one ever asked, including me
....... was
the hard drive spinning when you stopped the fan... need
not say more..?

about dodgy connections..... did you try the proverbial
engineering test....  
try tapping around with a insulated object for
bad connectoions etc.

A good
clue could be had by the above question.


Re: Decouple fans and hard disks supplies

On 15 Jun 2006 08:15:24 GMT, put finger to
keyboard and composed:

Quoted text here. Click to load it

Back EMF is usually suppressed by connecting a reverse biased diode
(eg 1N4001) in parallel with the motor.

     _____________ +12V
    _|_    |
   /   \  _|_
   | M |  /_\
   \___/   |
     |_____|______ Ground

When the supply to the motor is interrupted, the current continues to
flow in a loop through the diode until it decays.

- Franc Zabkar
Please remove one 'i' from my address when replying by email.

Site Timeline