[RFI] Why we badly need "party line" materials for all hams

dgsvetan at rockwellcollins.com dgsvetan at rockwellcollins.com
Tue Jul 20 14:51:01 EDT 2004


I had concerns about using S-meters back when the Cedar Rapids BPL fiasco
began.  Just prior to that, I had begun the arduous task of calibrating my
4 or 5 HF receivers at home (cross-reference chart:  Input in dBm vs.
S-meter readings for levels of S-1, S-5, S-9, and S-9+20).  When it became
obvious that I could grow a very long beard by the time the cals got done
because each radio varies by band in its S-meter characteristics, I
reconsidered and agreed with Ed that anything is better than nothing.

If you go look at the Cedar Rapids test report, you will see that we took
the data with a calibrated Agilent spectrum analyzer.  W0SR had just his
IC-765, but when the '765 said "S-9" for BPL, there was PLENTY of BPL to be
found.  I also had e-mail communication with Greg Lapin, N9GL, who heads up
the League's ARIA project, over the subject.  In the end, we all agreed
that while calibrated S-meters would be nice, they are not required to make
a point.  (ARIA formal reporting does rely upon calibrated receivers, but
that is another issue.)

On the other hand, let me hasten to add that if anyone else is going to go
through all of the time and trouble to perform actual testing at a ham's
QTH (I am NOT referring to performing a Part 15 conformity test of a
working BPL system) to evaluate the effects of a BPL system upon the
ability of the ham station to communicate, then (IMHO) you really need to
have a calibrated receiving device.  For shear ease of use and speed in
data taking, a spectrum analyzer (or EMC receiver) with the ability to
record plots into a file is the way to go.  If there is no other resource
except a communications receiver with an S-meter, then I would strongly
suggest a calibration of that receiver to a known calibrated signal
generator.  Be very careful with anything you plan to use as a means of
demonstrating interference; while it is obvious that all signals received
by a given antenna are relative to each other, being able to define
specific levels is critical when arguing a point in which receiver
performance comes into play.  Because S/N for a given receiver is related
to a specific input level to that receiver, it does become necessary to
relate the signals measured to the receiver performance.  One way in which
I did that in the Cedar Rapids report was to superimpose a line at -107 dBm
on the plots to show where 1 microvolt falls in a 50 ohm system.

I suggest you obtain and read through a copy of the Cedar Rapids report.
(It was filed as an appendix to my NPRM reply comments, as well as by
others.  The pdf file is provided by FCC for downloading.)  Note the

1.  ALL measurements at W0SR were made using his ham band antennas only.

2.  Antenna gains and antenna factors are not discussed in the report
because they are not relevant.  Each antenna "saw" all signals within its
bandwidth equally, be they communication signals or BPL interference.
Cable losses were discussed because that is good engineering practice in
EMC test reports, but those wash out, too.  All signals at a given
frequency are equally attenuated by the cable loss characteristic.

3.  Most measurements at W0SR were made with a 3 kHz IF filter in the
spectrum analyzer so that the instrument would "see" the spectrum as the
IC-765 saw it.  However, since the BPL carriers were spaced roughly 1.1 kHz
apart, a 3 kHz filter tends to integrate them into one "hump" on the
screen, making ID of the BPL signal difficult.  We used 1 kHz resolution to
show the carriers in a couple of the bands.

4.  Appendix 2 of the report is a table that gives typical receiver
performance for a number of receivers dating from the Collins S-LIne to the
FT-1000.  Luckily, input signal tests to achieve an S-9 reading is a
standard test in ARRL lab reports.  Those reports clearly show the big
variances.  Heck, I've got both a 75S-1 and a 75S-3, and there are a lot of
meter reading differences between those two sets because Collins added
another IF gain stage to the 75S-3.

Tom, I have to say that S-meter readings are not useless.  However, all too
many hams have no idea whatever of just what the meter on their rig is
actually telling them.  However, few would argue that S-9 interference will
make a long day out of trying to copy S-7 signals.  If someone wants or
needs to get data taken and nothing else is around, use the S-meter and go
with it.  Any data beats no data, but a simple S-meter reading doesn't mean
much without supporting information.  The rig can always be calibrated at a
later date if the need arises.  It is important to note
band/frequency/antenna used.  If different antennas are used, separate
readings should be taken for each one.  It is also important to record the
antenna's physical characteristics:  type (dipole, yagi, etc.), band or
bands it is designed for, type of mounting (fixed, rotating), and height
above ground.

73, Dale Svetanoff
EMC Engineer
Author of Cedar Rapids BPL Steering Committee report

                      "Tom Rauch"                                                                                                      
                      <w8ji at contesting.        To:       "Hare,Ed, W1RFI" <w1rfi at arrl.org>, <rfi at contesting.com>                       
                      com>                     cc:                                                                                     
                      Sent by:                 Subject:  Re: [RFI] Why we badly need "party line" materials for all hams               
                      rfi-bounces at conte                                                                                                
                      07/19/2004 05:25                                                                                                 
                      Please respond to                                                                                                
                      Tom Rauch                                                                                                        

Few hams have the resources to collect their data in any
other terms, Tom.   I am not sure what to do about that. >>>

Almost any ham has the ability to provide useful data. All
he has to do is:

1.) Use a step attenuator to calibrate the dB change in his
particular S meter.

2.) Drive out away from poles, and record noise levels or
weakest readable signal levels

3.) Go into the BPL area, and note the change and provide it
in dB for a given receiver bandwidth.

S-units are useless.

<<One one hand, gathering some information about the level
of pre- and post-BPL noise is important, and if all we can
get is S meter readings, then that is better than no data at
all.  Although the S meter has not been adopted world-wide
as a standard, IARU Region 1 has adopted an S meter standard
that uses 50 uV for S9, 6 dB per S unit, so those readings
to have some standards-based traceability.>>>>

Because the IARU in region 1 says "all S-meters equal X at
S-9, and are X dB per S unit it suddenly means S meters are
useful? I don't think so. It's pretty well established most
S meters are way out of whack for anything meaningful.

Not only that, no one has any idea what the antenna loss is
in the mobile.

The problems would largely disappear and the data would be
useful if people just followed a three simple steps. They
can record  the data in useless S-units, but give the
results in meaningful terms  like "the BPL system increased
background noise 22dB at 9 meters from the line using a
typical mobile station with XYZ for equipment and antennas".

Presenting data in S-units using a system with unknown
antenna efficiency, unknown absolute levels,  and unknown
change per S unit makes us look pretty foolish.

Why don't we tell everyone how to do a calibration chart, or
find people to offer those services for free for anyone
wanting to present data?

73 Tom

RFI mailing list
RFI at contesting.com

More information about the RFI mailing list