[manchesterwireless] Effective Radiated Power?

[G.W. Haywood]

Hi there,

On Tue, 30 Aug 2005, Dave Page wrote:

> Can anybody point me to some reliable documentation about Effective
> Radiated Power calculations? Some things I've read suggest that it's
> just a matter of adding the transmitter power to the gain of the
> transmitting antenna; others suggest that that only applies to
> omnidirectional aerials and it's different if you use directional ones,
> which I'm planning on doing.

ETSI 300-328.  But don't go there, there are dragons.  AFAICT you're
already under something of a misconception.  An omni IS directional
(otherwise it couldn't give you any gain:) but it's directional in the
sense that a diagram of the radiated power looks like a doughnut.

Yes, you should keep your output to 100mW (or 20dBm) EIRP but I've yet
to see any reliable way of measuring it and I've never heard of anyone
being taken to task for emitting a few milliwatts into a 15dBi Yagi at
2.4GHz.  Now if you talk to Darren I'm sure he'll have some stories...

Incidentally dBi means referred to a fictitious isotropic radiator
(think lightbulb) whereas dBm means referred to a miliwatt.  Sheesh.
So 20dBm means that 10 * log(ratio) = 20.  That means log(ratio) = 2.
Or in English, that means 100 times since the logarithm of 100 is 2.
These are logarithms to the base ten of course.  Dunno if they do this
in schools any more.

> I think that rather than trying to wirelessly network two
> properties, I'm just going to get a second ADSL line and use VPN

Having had some experience of the kind of problems that you can run
into with radio links I wouldn't necessarily argue with that, but
let's talk about it first.

> If anybody fancies casting an eye on
> http://www.livejournal.com/users/diffrentcolours/584492.html and
> offering any advice, I'd be very grateful.

<quote>
"Finding a robust solution to this is also important"
</quote>

Then I think you should seriously consider Linux boxes, even if it's
only for the links.

<quote>
It's a bit flakey, particularly between floors, but I've not been able
to work out why - I'd suspect interference or just a lack of vertical
signal strength in the built-in omnis
</quote>

Think polarization.  Try moving the kit around - particularly rotating
it, since the stuff probably was not designed to work in a lift shaft,
or try a different way of doing it, such as using separate antennae.
Or access points which permit the connection of external antennae.  It
really is a case of experimenting, with some thought applied to the
experiments (which is obviously one of your strong points:).

<quote>
I am assuming that the building frame of Zen will also block wifi
signals much as the one at Old Skool Daze does
</quote>

Wild assumption.  Using WRT54GS boxes with the supplied whip antennae,
I can easily get a 50metre link when the WRTs are inside two aluminium
portakabins on either side of a steel-framed shed.

<quote>
At the wireless protocol level, I'm looking at switching from the
ad-hoc network to one revolving around an access point, running off a
Linux server at Old Skool Daze wired to the external aerial.
</quote>

Oh, you _are_ considering Linux boxes.  Good man.  I've used hostap
for about three years and it does the job just fine.

<quote>
I'm thinking of doing away with IPSec, and using VPN instead.
</quote>

I've used OpenVPN for a couple of years.  For communication between
Derbyshire in England and Languedoc-Roussillon in France I use VOIP
over OpenVPN/UDP with good results through, inter alia, two access
points and three IPCop firewalls (two of which are wireless clients).
I run probably twenty or thirty ssh sessions (one of them is writing
this mail) and a number of other network applications concurrently.
It all works fine unless there's a lot of snow on the aerials.

It's too easy to imagine dragons.  Often when a link doesn't work as
well as was hoped you start to think about all the unfathomable things
when in fact it's the fathomable ones that are causing the problems.
Radio waves can get here from distant galaxies fercryinoutloud, I'm
sure they'll make it up the street.  It's just what you do with them
next that matters.  FWIW I wonder if you might be better off using
something like Yagis for your 220m link, since you suspect that there
might be interference.  I suspect you're probably right, but in my
experience interference isn't likely to be a big issue.  Take a look
at the little plastic box up the tower in

http://www.jubileegroup.co.uk/radio/tower.html

There are two access points in the box, about 40metres away from my
desk at 45 degrees to the horizontal.  From the wireless card in the
PC on my desk I'm able to choose which of the access points that I
connect to.  Clearly they aren't interfering much with each other.

<quote>
This would also mean that we could offer public wifi access to
non-VPN-authenticated clients, which would be nice.
</quote>

Check on the legality of what you're planning there.

<quote>
so basically I need to match the impedance of the card, the cable and
the aerial? ... it's been bloody ages since I did this AC stuff.
</quote>

Don't worry about it unless you start building your own.  All the kit
you'll be buying will have the same characteristic impedance (50 Ohms,
give or take) unless you start getting creative with satellite stuff.

<quote>
Most coax at 2.4 Ghz is hugely lossy, so keep the RF cables as short
as possible.
</quote>

Good advice.  Very good.  I use RG213 in lengths of 5-8 metres but I
wouldn't recommend it.  The main reasons I use it are that it's cheap
and easy to terminate.  If you have LMR400 then that's great, if it's
terminated well you can more or less forget about it.  I've seen some
pretty terrible terminations done 'professionally' (by which I mean it
cost me money) and now I do them all myself.  FWIW I'd trust Harry. :)

<quote>
An old satellite dish will probably give you more gain than an
expensive yagi.
</quote>

True enough.  But it might not have the right impedance and it will be
a bitch to point it in the right direction.

----------------------------------------------------------------------

Executive summary:

Measure things.  Make sure all the bits are doing what you think they
should be doing.  Write it all down in a book.  Draw pictures.  It
will take time, but from what I've seen you'll get the hang of it.

73,
Ged.