Path: utzoo!utgpu!jarvis.csri.toronto.edu!cs.utexas.edu!swrinde!zaphod.mps.ohio-state.edu!think!ames!apple!motcsd!hpda!hpcupt1!hprnd!pat
From: pat@hprnd.HP.COM (Pat Thaler)
Newsgroups: comp.dcom.lans
Subject: Re: Re: Why Ethernet-over-twisted-pair RFI emissions.
Message-ID: <2230038@hprnd.HP.COM>
Date: 5 Dec 89 18:16:31 GMT
References: <505@wcc.oz>
Organization: HP Roseville Networks Division
Lines: 33

> tom@wcc.oz (Tom Evans) asks:
> 
> 1. Has anyone actually _MEASURED_ the RFI emissions from UTP - 10BaseT?
> 
> 2. Isn't this the FCC's business? Shouldn't they have something to do
>    with it, or is it "you installed it - your problem"?
> 
Yes, I have measured it (on real product).  The emissions are about the
same as 10BASE2 (ThinLAN) products that I have measured.  They meet
the requirements of FCC and VDE (VDE is roughly the German equivalent
of FCC).  In my experience, any problems come from common mode signals
that are allowed to couple to the cable.  The source is usually processor
clock harmonics or the switching power supply rather than the data 
signal.  If you allow noise to couple to it, the shield of the coax
cable radiates as does the twisted pair.

If a product violates the FCC limits, the manufacturer may be required
to fix it.  In some cases, the manufacturer may also be fined.  VDE
has even stricter enforcement which in some cases may include jail.
If you are not using the product as the manufacturer intended (for
instance, if you operate without the covers in place) you may be 
required to fix the problem.

Many products were not subject to FCC limits until several years ago.
(I think it was about 10 years ago.)  Products in production at
the time were "grandfathered."  The noisy 1200 baud modem I mentioned
in a previous posting was made before the FCC limits applied (and was
not an HP product).

If you are worried about emmissions, 10BASE-T is not likely to be a
problem.

Pat Thaler