Hi,

I'm using an old PC (AT form factor, I think) power supply to power a
device that uses 12V at 3.7A, since I couldn't find anything else cheap
that would provide 12V at that level of current. All is working very
well, but the power supply fan is quite noisy, and I'd like it to be
silent.

Does anyone know if it's safe to disconnect the power supply fan? Is
the purpose of the fan to cool the supply itself, or is it (as I hope)
included as a convince to cool the case of the PC it's intended to
power?

Thanks for any help,
Jamie.

On 14 Jan 2007 20:05:41 -0800, wrote:

Hi,

I'm using an old PC (AT form factor, I think) power supply to power a
device that uses 12V at 3.7A, since I couldn't find anything else cheap
that would provide 12V at that level of current. All is working very
well, but the power supply fan is quite noisy, and I'd like it to be
silent.

Does anyone know if it's safe to disconnect the power supply fan? Is
the purpose of the fan to cool the supply itself, or is it (as I hope)
included as a convince to cool the case of the PC it's intended to
power?

Thanks for any help,
Jamie.

Going way out on the proverbial limb... g

If all you are doing is drawing 12V at 3 or 4 amps, you could probalby
do away with the fan. (Maybe a big probably...)

A much better solution would be a low noise fan!

As to your question: in the AT, the fan was to keep both the power
supply and the system cool. These systems generated virtually no heat,
except for the few systems that had a lot of added on cards and such.

Does the air coming out of the fan feel warm at all? (bet it does
not...)

wrote:
Does anyone know if it's safe to disconnect the power supply fan? Is
the purpose of the fan to cool the supply itself, or is it (as I hope)
included as a convince to cool the case of the PC it's intended to
power?

No way for anyone else to tell you because power supplies are
designed without providing such details and because some supplies are
built inferior. Any computer supply should do just fine without a fan.
Heatsinks are properly sized. Some power supplies save that $0.75 by
using no (or too small) heatsink and count on a fan to keep it cool.

So you experiment. Power it. Then measure component temperatures.
If it is hot enough to burn skin, then add a heatsink. Simple? Not
for one minute. There is this massive human safety problem. Those
parts can be electrically lethal - some charged to 300 volts even when
power cord is removed. You better damn well understand why that 300
volts can exist even when powered off AND take appropriate precaution.

Cooling - a little bit of airflow makes a major increase in cooling.
Then airflow must increase 4 times to only double cooling. Massive
cooling occurs with airflow so gentle as to not even be felt by a hand.

Fans of same size come in all types of options. For example, it can
move more air (more CFMs) and then increase noise. Or a same fan with
less CFM and little noise (and heatsinks) can be purchased from
electrical supply houses such as Digikey, Allied, Newark, Mouser,
Jamesco, etc

PeterD wrote:

If all you are doing is drawing 12V at 3 or 4 amps, you could probalby
do away with the fan. (Maybe a big probably...)


Unless its the worlds worst designed PSU I suspect he'll be fine up to
around 10A - ive powered circuits off old PSUs (that were removed for
noisy/dead fans) up to that current for hours at a time with no problems.
Cant vouch for it lasting months like that but seems no problem on mine
(old AT PSUs, not so sure about modern ones that seem extremely
temperamental and I get through them at the rate of 1 a year even just used
in a PC!).

J

Just to let everyone know, after monitoring the temperature of the the
PSU and the air emanating from it, and seeing that they were booth
essentially room temperature, I removed the fan (realizing the safety
implications astutely pointed out by w_tom).

The PSU is a /little/ warm to touch now, but there's very little heat
actually radiating from it, and it's still completly comfortable to
hold. It's been on and powering the device for over a week now, so I
think it'll continue to be fine.

Thanks for the advice,
Jamie.

On 27 Jan 2007 16:30:15 -0800, wrote:

Just to let everyone know, after monitoring the temperature of the the
PSU and the air emanating from it, and seeing that they were booth
essentially room temperature, I removed the fan (realizing the safety
implications astutely pointed out by w_tom).

The PSU is a /little/ warm to touch now, but there's very little heat
actually radiating from it, and it's still completly comfortable to
hold. It's been on and powering the device for over a week now, so I
think it'll continue to be fine.

Thanks for the advice,
Jamie.

SOunds good then. Were you to increase the load, you might want to
recheck the temperature. BTW, some PSUs have thermistors to monitor
temperature and shut down when overheat conditions are detected, but
not all do this. (Bet your's does not... g)

Heat from the power supply system is not relevant. Heat on
individual components (that should have heatsinks) may be excessive.
Surface temperature of major transistors and diodes should be
measured. But again, those part surfaces may be many hundreds of
volts. Temperature of case is not relevant. Temperature of major
internal parts is relevant. You cannot safely touch them with a hand
or conductive temperature probe. But if a part is exceeding maybe
150 or 180 degrees F, then consider installing a heatsink on that
part.

On Jan 27, 7:30 pm, wrote:
Just to let everyone know, after monitoring the temperature of the the
PSU and the air emanating from it, and seeing that they were booth
essentially room temperature, I removed the fan (realizing the safety
implications astutely pointed out by w_tom).

The PSU is a /little/ warm to touch now, but there's very little heat
actually radiating from it, and it's still completly comfortable to
hold. It's been on and powering the device for over a week now, so I
think it'll continue to be fine.