Power Cables Question

[Guest]

Archived from groups: alt.comp.hardware.pc-homebuilt (More info?)

Hi,

I'm a bit paranoid. I have a 350W power supply, used to power a P4 2.4GHz
on 865PE based mobo, a DVDROM, a CDRW, 2 HDD and a 3.5" FDD. The power
supply have 5 sets of output cables in this cofiguration:
1 set for ATX power (20 pin header)
1 set for ATX +12V (4 pin header)
1 set for legacy mobo power (remnants from AT days)
2 set for general peripheral (i.e. HDD power)

The general peripheral set have the following headers and current usage:
1 set have 2 output headers (used for both HDD)
1 set have 2 output headers plus one for FDD power (used DVDROM, CDRW and
FDD)

I recently added a casing fan, with power header matching the general
purpose headers. I suppose I have to use splitter cables (Y-Cables) on
one of the headers which belongs to one of the aforementioned sets. So,
to which set do I connect it to (the set for both HDD or the one for
DVD/CDRW/FDD) to ensure power stability (prolong the lifetime of the
attached peripherals, in terms of power)?

PS: The mobo supports 1 additional system fan with 3 pin power header.
Strangely, I couldn't find any local dealers who carries such fans with
matching power input headers. They only carry those with general purpose
input headers. I have difficulties with buying online, not to mention the
shipment fees will outweigh the fan's price.

TIA

 

[Guest]

Archived from groups: alt.comp.hardware.pc-homebuilt (More info?)

Ricky Romaya wrote:

> I have a 350W power supply, used to power a P4 2.4GHz on 865PE
> based mobo, a DVDROM, a CDRW, 2 HDD and a 3.5" FDD. The power
> supply have

> 1 set for legacy mobo power (remnants from AT days)
> 2 set for general peripheral (i.e. HDD power)

> 1 set have 2 output headers (used for both HDD)
> 1 set have 2 output headers plus one for FDD power (used DVDROM, CDRW
and
> FDD)
>
> I recently added a casing fan, with power header matching the general

> purpose headers. I suppose I have to use splitter cables (Y-Cables)
> So, to which set do I connect it to
> to ensure power stability (prolong the lifetime of the attached
> peripherals, in terms of power)?
>
> PS: The mobo supports 1 additional system fan with 3 pin power
header.
> Strangely, I couldn't find any local dealers who carries such fans
with
> matching power input headers. They only carry those with general
purpose
> input headers. I have difficulties with buying online, not to mention
the
> shipment fees will outweigh the fan's price.

Don't restrict yourself to computer dealers since electronics dealers
often carry more products and charge less for them. A local technical
college can tell you where the better dealers are located because
students need to buy parts for projects. Some non-local dealers, such
as www.newegg.com, www.compgeeks, and www.svcompucycle, are not very
expensive on shipping but can be cheaper than local sources even when
shipping is included.

That "legacy mobo power" socket is for AT motherboards only if it has
12 pins in a single row. If it has 6 pins and 3 are black, 2 orange,
and 1 red, then it's for older Athlon or Pentium III motherboards,
especially those with dual CPUs.

If your current motherboard has a square 4-pin connector or an IDE-type
4-pin power connector, you should have a power cable plugged into it or
else the sole yellow +12V wire on the 20-pin ATX connector will be
overtaxed by the CPU voltage regulator. It's best to limit the current
on each wire to roughly 6A, to prevent its connector from dropping too
much voltage or overheating.

Generally, add connectors to the cable with the lowest load already on
it, but this shouldn't matter with a case fan because it won't draw
more than about .25A. But to be particular about it, get a cheap
voltage meter and measure the voltages at the connector pins, and use
the one with the highest voltages, indicating the lesser loads on that
cable.

What do you mean, "prolong the lifetime of the attached peripherals,
_in_ _terms_of_power_"? Lifetime is expressed in terms of time, not
power. What's your major?

 

[Guest]

Archived from groups: alt.comp.hardware.pc-homebuilt (More info?)

Ricky Romaya wrote:

> do_not_spam_me@my-deja.com wrote in news:1103947488.013959.64750
> @f14g2000cwb.googlegroups.com:
>
> > What do you mean, "prolong the lifetime of the attached
peripherals,
> > _in_ _terms_of_power_"? Lifetime is expressed in terms of time,
not
> > power. What's your major?
> >
> Yes I know that lifetime is expressed in terms of time, or AFAIK
> expressed as MTBF (Mean Time Before Failure) in number of hours (on
> average, or statistical mean value) before the device fails,
according to
> the tests performed by the vendor (usually in highly specialized
> environment).

MTBF may not be closely related to lifespan because MTBF generally
excludes aging. An example of this disconnect is with hard disk
drives, which now have MTBFs that are 10-20x as long as they were 20
years ago, yet the average lifespan rating is still only 5 years or so.

> What I meant by "prolong the lifetime ...." is to minimize power
problems
> as a factor for causing the device to fail. Insufficent power feed
can
> cause the device to fail more quickly than the said MTBF. This is my
main
> concerns.

Then you want something with low voltage shutdown. All power supplies
are supposed to have this, and my guess is it works more reliably than
the overvoltage, overpower, and overcurrent features. But I'd be more
concerned about the latter 3 because more failures occur with those
faults, and more supplies don't implement those protections adequately.

 

[Guest]

Archived from groups: alt.comp.hardware.pc-homebuilt (More info?)

do_not_spam_me@my-deja.com wrote in news:1103947488.013959.64750
@f14g2000cwb.googlegroups.com:

> What do you mean, "prolong the lifetime of the attached peripherals,
> _in_ _terms_of_power_"? Lifetime is expressed in terms of time, not
> power. What's your major?
>
Yes I know that lifetime is expressed in terms of time, or AFAIK
expressed as MTBF (Mean Time Before Failure) in number of hours (on
average, or statistical mean value) before the device fails, according to
the tests performed by the vendor (usually in highly specialized
environment).

What I meant by "prolong the lifetime ...." is to minimize power problems
as a factor for causing the device to fail. Insufficent power feed can
cause the device to fail more quickly than the said MTBF. This is my main
concerns.

AFAIK there are no direct number/rating systems which express lifetime
versus power feed/consumption ratio.

BTW, thanks for the advice on measurements. I always want to test it but
don't know the guidelines.