Re: Power over ethernet and ePODs

["graham_howe" <graham@xxxxxxx>]

--- In ukha_d@y..., "Kevin Hawkins" <lists@u...> wrote:
>     Hi Graham
> 
>      - it's probably not a very good idea to
try and anticipate a 
voltage
> drop down a cable to reverse engineer the correct voltage at the 
other end.
> A few reasons spring to mind...
> 
>     1) The ePods and other devices take a varying
current and it is 
the
> current when passing through some resistance that causes a voltage
drop (yep
> ohms law) and so the voltage drop down the CAT5 will vary with what
the ePod
> is doing, the backlight contributing significantly.When the unit 
was in
> power save mode the voltage may rise causing risk.
> 
The idea is that the ePODs would be wall mounted and turned on all 
the time. They would (of course!) be running HomeAutomator as a 
control interface and I plan to incorporate some form of 'screen 
saver' mode that effectively cycles through pictures. The backlight 
would be set for a very short duration so would be off almost all the 
time.

>     2) Possibly the biggest resistance in the line
would be caused 
by any
> connectors that were in circuit but their actual resistance may be
a little
> variable depending on physical issues. (I haven't looked at the POE
design
> to see if there are any connectors in line.) Where there is 
resistance, heat
> is generated.
> 
All my wiring is structured and so is temrinated with RJ45 ports at 
either end (patch panels in node zero, wall ports elsewhere). I was 
planning to have an RJ45 plug from the power supply going in to the 
patch panel and then have the ePOD connected via a fly lead at the 
other end. So in terms of connectors we have :

PSU - CAT5 - RJ45 Plug - Patch Pannel Socket - CAT5 - Wall Port - 
RJ45 Plug - CAT5 - RJ45 Plug - ePOD socket

>     3) Not all runs to epods would have the same
voltage drop so 
you would
> need different source voltages anyway.
> 

The likely locations will be between 15M and 50M from node zero

>    The ePod appears to require 10V DC (an unusual
voltage) at up to 
2A.
> Presumably the 2A would be with the machine running full tilt, all
slots
> occupied and charging the battery as well. In typical use this may
be much
> less and you might not have a problem with the CAT5 anyway. What 
would be
> interesting to know is if there is further on board regulation in 
the ePod
> in which case it may well compensate for any small voltage losses 
anyway.
> Not sure how well regulated the ePod psu's are.
> 
As I said the ePODs will be turned on all the time and will have both 
PCMCIA (butchered internall NIC) and CF card installed at all times. 
However, as these will be 'permanent devices' I might be able to get 
away with not connecting the battery. The only down side would be in 
the event of a power failure where the ePODs would die and require re-
hacking. But that is now extremely easy and we hardly ever have power 
cuts here.

I have no idea about the regulation side of things.

>     The technical solution would be to have a
spare wire coming 
back from
> the ePod end that carried no current and was used to sense the 
voltage at
> the ePod end and automatically raise or lower the source supply to
> compensate but again this facility is only in the more expensive 
PSU's and
> would not work for multiple loads. e.g. several ePods.
> 
I was hoping to use two wires each for power, so that would not leave 
a spare wire to sense the voltage.

>     I would be tempted to try it and see, 
feeding the existing 
supply
> through a longish length of Cat5 and seeing if the ePod powered 
correctly,
> it is unlikely to do any real harm to the ePod (I guess there could
be some
> data corruption potential).What sort of length of run of Cat5 are 
we talking
> about ?? Avoid connectors if poss.
> 
See above, can be quite long and there are a few connectors involved. 
But trying would be quite easy I guess.

>     The cable will have a spec somewhere on
resistance per metre 
although
> there may not be a guideline to current carrying potential as it is
intended
> for data signals only. I know people use them for speakers and 2A 
into 8
> ohms would be 32Watts or 16W into 4 ohm speakers. 1A would be only
8 watts.
> The voltage drop would sort of be ignored though just producing a 
less
> accurate sound. All I am saying is that 32W of power is more than 
people are
> using them for so this is pushing it in comparison. How many 
conductors are
> available for the power ? Pairing them if poss could help - I 
haven't looked
> at the article.
> 
The article does pair the cables for positive and negative, I would 
do the same.

>     I am sure Keith would have a view on a current
handling 
capability. 2A
> is quite hefty though.
>

Keith?

>     Kevin
> 
Thanks Kevin, lots of good stuff

Yahoo! Groups Sponsor

Click here to find your contact lenses!

For more information: http://www.automatedhome.co.uk
Post message: ukha_d@xxxxxxx 
Subscribe:  ukha_d-subscribe@xxxxxxx 
Unsubscribe:  ukha_d-unsubscribe@xxxxxxx 
List owner:  ukha_d-owner@xxxxxxx

Your use of Yahoo! Groups is subject to the Yahoo! Terms of Service.