[UK-CONTEST] IOTA Logging Advice Please

[Ian White, G3SEK]

G3WGV wrote:
>
>
>Of LANs and Wirelesses and things...
>
>UTP is, IMO, a menace to Amateur Radio in anything longer than about 3m
>lengths.  It relies upon the balanced pair posessing common mode 
>rejection, which
>in turn is totally dependent on the accuracy with which the pairs are 
>twisted.
>Otherwise, it's basically another antenna, with all its EMC susceptibilities
>in both directions.
>
Or basically another ground radial. In a multi-TX setup, a star network 
is the perfect radial ground system. Each station will try to shove as 
much ground current into the network cables it as it can... and into the 
mains cables... and... and...

In UTP network cables, these are common-mode currents flowing on both 
wires in phase. As John says, they get converted unto differential 
(signal) mode due to the poor balance of UTP and its PC terminations. 
This is what causes trouble in t' computers.

You can improve the common-mode rejection by simply winding some turns 
on a ferrite core, at both ends of the run. For HF, you need quite a lot 
of inductance. Because of the size of UTP and its pre-formed plugs, you 
need a large toroid, or else a *very* large (probably impractical) 
number of clip-on ferrite beads. By far the best solution is to use 
ferrite scan cores from scrapped TVs - about 10 turns will stop just 
about everything, down to Top Band... and they're free for the asking.

In fact, it may be better to put this RF choke on the interconnections 
between the computer and the radio, at each station. Given a large 
enough core, more than one cable can be bundled together and wound on 
the same core. That puts all the computers on the "non-RF" side of the 
system.


>At D68C we used 10Base2 (UR43-like coax to us radio types) and by and large
>this worked well, though we did get some sprogs on 30m, which we fixed by
>burying the cable a few inches underground.  We also occasionally had 
>the LAN drop
>out, which we attributed to lots of QRO transmissions all coming together at
>the same time.  These days 10Base2 is becoming tricky to implement as suitable
>terminating hubs, NICs, etc are getting hard to source.
>
CPC may still have them (www.cpc.co.uk - they'll deal by credit card, 
but don't trust either the website or the catalogue as an indication of 
actual stock).

As longer-term insurance, I've picked up a few extra 10-base-T PCI 
cards, and also a UTP hub which has a 10-base-T port as well. The plan 
is to use UTP for short runs, but continue to use coax for the long runs 
that are more likely to act as aerials/radials.

However, technology marches on...

>For 3B9C it seemed to me that WiFi was the way to go, so as the DXpedition
>technology guy, I started experimenting.  So far I have been completely unable
>to detect any EMC troubles in either direction.  Once you get it set up
>properly, the LAN is stable and, more or less, works just like a wired 
>LAN.  It may
>also help with good RF engineeering practice, because the LAN no longer 
>creates
>an electrical connection between the stations.
>
Avoiding that hard-wired connection seems like a *very* good idea!

Temporary multi-TX setups are an extreme high-risk situation for RFI, 
because the interconnections are hellishly complex and it's impossible 
to debug the full layout before you get there. It certainly is worth 
taking every possible precaution in advance... and no surprise at all to 
see that John's already on it.



-- 
73 from Ian G3SEK