[CUWiN-Dev] LMR400 too industrial grade ...Re: CU-Wireless-Dev Digest, Vol 11, Issue 16

[Patrick]

Rick et al,
   See my comments below your message: 2
-Patrick

Message: 2
Date: Wed, 16 Mar 2005 19:01:54 -0800
From: Rick Elnor <Rick at nixbased.com>
Subject: [CUWiN-Dev] Cable type Re: CU-Wireless-Dev Digest, Vol 11,
        Issue 15
To: cu-wireless-dev at lists.cuwireless.net
Message-ID: <4238F322.2020004 at nixbased.com>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed

>
>This is with regard to the message below (Message: 1). I agree with you
about the
boards but saying LMR-400 is too industrial grade maybe an oversight. LMR-400
(depending on type of dielectric,the size and type of the conductor,
length of
cable and so on) has about a 7dB loss for 100FT. Now this may not be an
issue with
the person you were replying to but I know from experiance it can be. We
actually
have a 96FT tower here that we were thinking of adding a failsafe to and
doing the
cable run up the tower.  If I missed something I apologize for being off base
(probably am as I havn't seen the location).
>

In general, and as we said in our original note, the distance between the
radio card and the antenna is a consideration. For short runs (a few
feet), LMR400 is too industrial grade in our opinion.

The issue is RF performance, especially on the receive sensitivity, which
involves more than just dB loss per foot ... it's a double whammy because
the signal strength is reduced AND the noise temperature is increased more
and more with cable length until your signal hits the LNA (which is in the
radio card).

If you've got a 100 foot run, we suggest you do a 100 foot PoE run to the
"box" with a short run of between the radio card and the antenna (say 1 or
2 foot ... or zero feet if you are attaching the antenna directly, which
would give best RF performance).

If for some reason, you must separate the "box" from the antenna by such a
long run, you should probably consider placing an RF device at the
antenna. The system receive sensitivity will then be dominated by the
performance of the RF device.  hese devices probably cost between $100 to
$500 (depending on PA & LNA quality ... both gain & noise factor). You
will likely need to get power to the RF device too. We're not doing this
so we can't tell you what it entails for sure. Also, not sure about the
length of run wrt duplexing and timing issues (i.e., the device switches
between Xmit & Rcv somehow because the radio cards are half duplex).

-Patrick