[RFI] Excellent RFI notes by ON4WW

[Hare, Ed W1RFI]

> One thought, however.  I STRONGLY object to use of the word "overload"
> to describe every case of RFI to equipment, because it's
> counterproductive. That is, it doesn't get interference fixed, because
> it doesn't get to the fundamental circuit mechanisms that couple RF into
> the circuitry that generates (or detects) the trash.

In an engineering environment, you are completely correct, Jim. However, in terms of discussing interference from licensed transmitters, the use of that simple term is an important way of differentiating interference caused by spurious emissions from a transmitter, which is the responsibility of the transmitter operator to correct and interference caused by the fundamental signal of a licensed transmitter, which is not the responsibility of the operator to correct.   This is especially important when an Amateur is dealing with interference a neighbor's equipment, but it also applies to the important first step in troubleshooting many RFI problems -- answering the question about whether the interference is caused by a transmitter defect of some sort or by insufficiencies in the affected equipment.  Do not discount that important distinction. 
 
> In fact, most RFI detection is NOT caused by overload of a gain stage,
> but rather by simple detection in the non-linearity at the turn-on
> corner of a semiconductor junction somewhere in the signal chain!  It
> takes a LOT more signal to overload that gain stage.

That is overload, Jim, as those stages are not often intentionally operated in the non-linear region near the cutoff of the active device being affected.  "Overload" is a very generic term, not to be confused with "saturation," which is what I believe you are discussing.  The use of that term as an important delineator of responsibility in the case of interference involving a licensed service is useful and should be continued.

While I agree about "the pin-1" problem being a major contributor to common-mode-related interference problems, the term that applies well to some audio connectors is a complete misnomer with respect to most devices.  Unfortunately, in many cases, it is simply not possible or practical to correct the internal design issues in consumer equipment, so knowing the basic cause is not always necessary or helpful.  Those internal grounding problems in some equipment can even extend to the way that PC grounding traces are run on a board, so simply concentrating on the common-mode nature of the pickup on most wires is more useful, IMHO.

73,
Ed Hare, W1RFI
ARRL Laboratory Manager
225 Main St
Newington, CT 06111
Tel: 860-594-0318
Email: W1RFI at arrl.org

________________________________

From: Jim Brown [mailto:jim at audiosystemsgroup.com]
Sent: Mon 12/27/2010 7:53 PM
To: rfi at contesting.com
Subject: Re: [RFI] Excellent RFI notes by ON4WW



On 12/27/2010 4:14 PM, Jimk8mr at aol.com wrote:
> The transmitter section was being grossly overloaded
> with  HF RF from 20 meters, causing it to put out RF garbage at who knows
> where!

Great story!  Many thanks.

One thought, however.  I STRONGLY object to use of the word "overload"
to describe every case of RFI to equipment, because it's
counterproductive. That is, it doesn't get interference fixed, because
it doesn't get to the fundamental circuit mechanisms that couple RF into
the circuitry that generates (or detects) the trash.

In fact, most RFI detection is NOT caused by overload of a gain stage,
but rather by simple detection in the non-linearity at the turn-on
corner of a semiconductor junction somewhere in the signal chain!  It
takes a LOT more signal to overload that gain stage.

So back to those coupling mechanisms, which are 1) Pin 1 Problems in the
equipment (that is, mis-termination of cable shields); 2) Poor shielding
of equipment, and poor circuit board layout within equipment;  3) poor
shielding of interconnect wiring;  4) Use of untwisted wiring for signal
conductors 5) excessive bandwidth (or inadequate bandwidth limiting) of
inputs and outputs.

My guess is that the dominant coupling was common mode, exciting #1. If
that's true, it could probably be killed with a multi-turn ferrite choke
on one or more of the leads going to that XM rcvr.

BTW -- your technique of reducing TX power to find out how much
attenuation it takes to get rid of the interference is a VERY good one
-- it tells you that you need at least 16dB of suppression at the victim
device to kill the interference.

73, Jim K9YC
_______________________________________________
RFI mailing list
RFI at contesting.com
http://lists.contesting.com/mailman/listinfo/rfi <http://lists.contesting.com/mailman/listinfo/rfi>