Re: Current carrying capacity of CAT5?

["rectorydp" <dpullen@xxxxxxxxxxxxxxxx>]

When I (sort of) answered this question back in August, I couldn't
find a previous reference in UKHA_D. Since then it has been asked
several times!

My response at the time was as follows.


"I can't answer this question definitively but it is something that I
have struggled to find (only considering CAT5 here). I have used CAT5
to power relays and LED lights. I think this may be a good question
for some sort of reference sheet on Automated Home although I am not
sure if it would qualify as a FAQ since I haven't seen it asked or
answered before!

I started off by looking at Power Over Ethernet specs (standard IEEE
802.3af).

Voltage between 44V and 57V
Maximal current 550mA
Typical current 10mA-350mA
Maximal power 15.4W

My understanding is that PoE uses two pairs so we have some indication
of maximum power although this is using a higher supply voltage that
allows lower currents and therefore less power loss.

The site www.powerstream.com provides information on maximum current
for power transmission and chassis wiring (See section on wire gauge
and current limits). This suggests maximum current of 3.5A for chassis
wiring (applies to wiring in air not a bundle) and 0.577A for Power
transmission for 24AWG wire. These figures are supposed to be very
conservative. The page also points out that these figures are rule of
thumb which may indicate why it is so difficult to find a definitive
answer.

Just out of interest if you double the cross-sectional area the wire
gauge decreases by 3. This would suggest two cores of CAT5 is
equivalent to 21AWG. In terms of power loss and current capacity this
may not be strictly correct especially if there are higher frequencies
involved (not a problem if it is strictly DC) but it is probably okay
for rule of thumb!

Information on ePanorama.net
(www.epanorama.net/documents/wiring/wire_resistance.html) indicates
"Ampacity" of 1.3 for 24AWG copper wire.

There appear to be two important effects that are taken into
consideration when calculating maximum current: voltage drop and
heating effect. The figures for voltage drop are generally such that
the heating effect would not be significant so voltage drop figures
can be used to calculate maximum current. However, guidelines seem to
cover a wide range. I have seen recommendations ranging from 2-6%.

Voltage drop can be calculated on the PowerStream page or you can use
the formula

V(drop) per hundred feet paired run = 0.2 * I(load) * 1.26^(AWG-10)

Unfortunately I have found that many of the formulae in this area
haven't been touched by metric units!

Using this formula, a run of 10m (pair of cores out and back), and a
supply of 12V suggests a maximum current of 0.4A for CAT5. If you
double up the cores (equivalent of 21AWG) then you should be able to
get up to about 0.8A.

Having said all this there are two further problems:

1	Most of the information on this topic is really focussed on higher
power applications eg mains or automotive wiring.

2	None of this takes into account twisted pairs.

Twisted pairs will have some effect but I do not think it will be very
significant (hopefully someone more knowledgeable will add to this).
The figures I have used do cover 24AWG so hopefully we are okay and
all figures are quoted as being very conservative anyway.

So what does it all mean? Well I have decided that I am happy to carry
about 0.4A per pair for a 10m run (I am using the voltage drop
calculation with 6%). I am also setting a maximum of 0.6A per pair
even for shorter runs that would allow higher currents with the 6%
limit. I think it would probably be safe to go above this but I want
to make sure that there are no problems.

I hope this helps and that others will add to this with their
experiences."

BR
David



--- In ukha_d@xxxxxxx, ian.bird@c... wrote:
>
> I have a memory from a while back when this came up and it was about
300ma
> per pair (one plus, the other minus) depending on distance. I use
CAT5 for
> my watering valves which draw 230ma at 12v and I have set it up so
only one
> can come on at any time. Keith Doxey was the guru answering the
questions
> then.
>
> Ian
>
>
>
>
>
> ---------+---------------------------->
>                     Ake Hedman
>                     <akhe@b...
>                     om>
>
>                     03/01/2005 07:41
>                     Please respond to
>                     ukha_d
>
> ---------+---------------------------->
>
>-----------------------------------------------------------------------------------------------------------------------------
>

>          To:       ukha_d@xxxxxxx

>          cc:       (bcc: Ian Bird/CV/Novartis)

>          Subject:  Re: [ukha_d] Current carrying capacity of CAT5?

>
>-----------------------------------------------------------------------------------------------------------------------------
>
>
>
>
>
> Martin Howell wrote:
>
> > Just completed my pair of water leak detectors for under the
washing
> > machine and dishwasher, loads of coloured LEDs, ooh er missus.  I
plan
> > to feed them with power via the CAT5 cabling, which will also
carry a
> > copy of the status LEDs back to the central indicator panel. 
Anyone
> > know what current can be safely poked down the CAT5 cabling?  The
major
> > current is drawn by the water shut off valves, which are rated at
12
> > volts and 3 watts, by my rusty maths thats 0.25 amps each, and
the
> > washing machine has a pair of valves, so 0.5 amps.  Are my lovely
> > coloured lights going to be invisible through a cloud of acrid
smoke?
> >
> > Martin
> >
> >
> 500 mA should be fine.
>
> /Ake