[Flexradio] Odd RFI Problem

[Scott Myers]

Brian,

I'll get the problem whipped since I am now armed with this new info.
Thanks again for all your input.  And of course Tim and Dudley, who are
always helpful.

73,
Scott AC8DE

-----Original Message-----
From: brian at lloyd.com [mailto:brian at lloyd.com] On Behalf Of Brian Lloyd
Sent: Saturday, December 05, 2009 5:29 PM
To: Scott Myers
Cc: FlexRadio at flex-radio.biz
Subject: Re: [Flexradio] Odd RFI Problem

On Sat, Dec 5, 2009 at 12:57 PM, Scott Myers <ac8de at ameritech.net> wrote:
> Brian,
>
> OK, NOW we're getting there.

Ah good.

> I did try to put few chokes up at the switch to no avail.  Now I didn't
put
> a choke on every single network cable.  That would be about 20 of them and
I
> don't have that many ferrite beads.

You may need to do that. OTOH, I suspect that it was only one of your
networking cables that was hosing up the hub. My guess is that one of
them is resonant and providing a fair bit of common-mode RF current
into the hub/switch causing it to fail and reset repeatedly, thus
taking all the ethernet NICs down.

> But you made an astute observation on overall network stability.  So I
tried
> surfing on the PSDR machine while in TX and it wouldn't surf, even though
> Vista does not show the computer disconnected from the network.  Indeed,
it
> is interfering with the switch, router or modem... or all of them.

Yup. The networking stack can't really tell that connectivity has gone
away so it relies on receiving the carrier/link-signal from the hub.
If the hub continues to provide link, the NIC in the computer will
think that the network is up even though no bits are flowing
successfully through the network. So the Windows networking stack just
assumes that carrier means connectivity. As you can see, that is not
necessarily a valid assumption.

> I lengthened  my RTTY test transmission and cranked it to 800 watts PEP
and
> after about 45 seconds, it does show limited connectivity in the system
> tray, although it doesn't show a complete disconnect.  I then comes back
> connected after the TX ends.

Yup.

> I checked another computer and it was able to surf, but at a snail's pace
if
> really at all. Then it comes back to full speed after the TX ends.

Yup.

> So suspecting that the RF was leaking into the switch from the PSDR
> computer's network cable, I grabbed one of the snap on chokes and wrapped
it
> around the PSDR computer's cable at the router.  (I had already tried the
> printer to no avail)  Bingo!  With that installed, I don't lose the HP
> printer and I can surf on the other machines while in TX.  So while I did
> already have a bead down on the network cable at the computer, that was
> clearly not enough.

No. The RF was not getting in from the radio computer. It was being
picked up by the cable in free space. That is what I suspected. I just
didn't know which cable it was. It could have been any of the Ethernet
cables going to the hub/switch. And it may be more than one as well.
All your Ethernet cables are different length so they will also be
resonant at different frequencies. Also different orientation will
make them behave differently. You may need to test each band to see
which cable is causing the problem and then slap on the common-mode
chokes.

> So I need to get the computer grounded to the RF ground and try adding and
> removing ferrite beads in strategic places.  Looks like I'll be calling
> Amidon on Monday and getting a bag of mix 31 beads!

That might be overkill but it won't hurt anything.

-- 
73 de Brian, WB6RQN/J79BPL