Letters

Re: Leonard Kahn's AM Powerside

Bob Gonsett's CGC #318 newsletter asked the following question:

THE KAHN POWERSIDE - HOW USEFUL OF A DEVICE?

Do you have any firsthand experience with the Kahn Powerside for AM
broadcasting - the device that concentrates a station's modulation on
one sideband or the other?

Leonard Kahn discloses where the device seems to work well, but are
there instances where a Powerside has little or no (or even a negative)
effect?  Would a Powerside help a simple non-directional AM station
without a first adjacent channel interference problem?

Send a note.  Tell us your thoughts.

Replies:

In order to understand the potential of this device let's review a few
facts.

1. The Powerside does not increase the peak envelope power produced by
an AM transmitter. In fact it may somewhat reduce the PEP due to slew
rate limitations and on board peak limiting within the device.

2. Part 73 of the Commission's Rules clearly defines the emission mask
that AM stations must adhere to (i.e. NRSC ) so that parameter is fixed
also. It is possible for an engineer installing a POWERSIDE unit to
mis-align the device thereby causing the facility to be in violation of
the Commission's mask. It is dubious as to the potential benefit that
would result from such an alignment.

Since the Powerside is essentially a re-worked Kahn AM Stereo exciter I
do not see any significant benefit for a station that does not suffer
from a first adjacent channel issue. This device simply shifts the
spectral energy distribution from an AM transmitter to favor either the
upper or lower sideband.

Jim Stanley W6GH / AFA3CY 
Regional Engineering Coordinator 
Cumulus Broadcasting Inc. 
jim.stanley@cumulusb.com 

Back in 85-86 we tested the Powerside, along with "The Flatterer", on
WNBC. We, and WQXR, tested just about everything Kahn came up with. The
Powerside did what he said, namely shifted the energy about the carrier
frequency in a controlled, asymmetrical manner. For us it was pretty
useless, and precluded the use of AM stereo, which we were running at
the time. The tests were ended after maybe a week. As a non-DA I
didn't see much to gain from it, since we were a clear channel with no
big first adjacencies causing trouble (at least not until you got into
Ohio...)

WABC ran it for a long time, but I don't know what their experience was.
Today, with digital radios, some of the benefit would be lost. Like
you, I would be VERY interested in knowing how well it works for folks
with DA's and bad, high distortion nulls. (If I were still with
Chancellor, I'd love to try it on KFYI.) I suspect it might be very
useful, on a station by station basis, for such cases.

The Flatterer, on the other hand, was somewhat useful. That was a
modified Kahn stereo exciter which allowed the two sidebands to be
individually "EQ'd" to correct for non-flat antenna VSWR, etc. (Sort of
like stagger tuning a TV xmtr for flattest passband response thru the
load.) It did work, with no discernable audio side effects (much to my
amazement), and allowed us to partially compensate for the large antenna
upper sideband impedance swing we had from the 660/880 mother of all
diplexers. (We shared the tower with WCBS. They still do.) This was
compatible with Kahn stereo, so we continued to use it until the day
Emmis took over and put in that ghastly C-Quam. It is not C-Quam
compatible, so it remains only an historical footnote.

PS: Remember the Non-Symmetra-Mod?? Whew, almost blew the modulators
out of their sockets in our 317-C2 with all the DC offset shift!....

Gary Blau KOOL/KISO
Phoenix Chancellor Media 
gblau@radiomasters.com 

I was just forwarded your recent "Communicator", requesting info on the
Kahn PowerSide. This comes at an interesting time, because we are now
testing it on a two-tower 1 kW station. The station (KRRS, Santa Rosa)
has a PSA power of 500 watts, and a nighttime power of 33 watts. The
1000-watt signal doesn't do what we would like it to, and the lower
powers are correspondingly worse.  The transmitter plant is in good
shape, with a new BE AM-1A "Power Miser" transmitter, and stock Optimod
9100 processing.

My reactions, so far:

1) It's a lot of money. $16,000. 2) We were promised great things
that didn't materialize. Later blamed on inaccurate maps. (!!) 3) Its
main claimed strength, the ability to produce a louder signal off to the
side of the main carrier, is lost on the center-tuned digital radio; now
becoming quite popular in cars, boomboxes and Walkpersons. 4) It DOES
work, sort of. Listeners in the fringe were able to tell fairly
reliably when the system was switched-in or out. Hard to say if it
works better than state-of-the-art audio-only processing, though. We
intend to try that, next. Coverage is increased at the expense of audio
quality.  I suspect there is some pretty ruthless clipping going on. 5)
 It seems to add its own artifacts; a kind of fluttering in areas that
had no problem before. 6) All "normal" test procedures are out the
window; there's not much in the way of (understandable) documentation. 
Mod monitor is useless, because it attempts to decode the "stereo"
signal......

Randy Wells CE: KRRS, KSRO, KXFX, KFGY, KMGG 
RadioRan@aol.com 

Regarding your Power Side query vis a vis small non directional station.
The answer is "it depends."

The old conventional wisdom about AM radio states that it doesn't suffer
from multipath effects. If AM stereo taught us nothing else, it taught
us that this conventional wisdom is false. AM multipath behaves
differently because of the longer wavelength of the frequency: instead
of 'picket fencing,' we get a slow rolling effect.

AM consists of a carrier and two sidebands in a very specific phase
relationship between them. If you disrupt the phase relationship
between upper and lower sidebands, or the phase relationship of the
sidebands to the carrier, you have distorted the signal, and it won't
demodulate properly in an envelope detector. AM may be a bit more
forgiving than FM in this regard, but it is not immune to this.
Reradiation (multipath) causes signals to arrive out of time, and thus
causes phase cancellation anomalies that damage the signal.

With that little bit of background, power side's strength lies in its
ability to alter the signal such that one sideband is no longer of
concern to the receiver. This has the potential of reducing multipath
effects by about one half.

Incidentally, it is my experience that most broadcast engineers have a
high understanding of transmission, but little understanding of
reception technology. This causes much confusion and misunderstanding
of what this system does.

I think the bottom line is that Power Side can be a benefit to any
station who's signal is being "damaged" by re radiation (AKA multipath).
In AM radio, sources of this are high rise buildings, power lines,
certain natural phenomenon like canyons and hills, and also,
interestingly, tunnels. It has been my observation that the system
noticeably reduces the 'fading under bridges' that plagues AM radio.

Of course another source of phase distortion is directional antenna
nulls, and probably the most striking changes can be noticed there.

Another large benefit can occur with manually tuned receivers. The user
will tune for the best signal, and the carrier will fall into the slope
of the IF filter causing a drop in the AGC voltage. This causes a sharp
increase in the loudness of the signal. Digital tuners won't normally
see this benefit; however, in areas impacted by re radiation, the
recovered audio will be higher, which improves the noise floor even in
center-tuned radios.

What are the down sides? The thing that I think needs to be stressed is
that Power Side cannot create signal where there isn't any to begin
with. I think most of the negative experience with the system has come
from stations who were looking for a miracle cure for their inherent
inability to cover outside their service area. However, there is one
small station that reportedly now gets into the downtown area.
Apparently the problem wasn't so much signal strength as it was
reradiation from high rise buildings.

As to negative side effects, on either side of a directional null, the
character of the audio may be altered somewhat. It will tend to sound
different on either side of the null, and the distortion may be a bit
higher on one side of the null. Generally, the benefit to receivability
in the null area far outweighs the slight change in audio
characteristics. Sometimes experimenting with which sideband is
suppressed can optimize this situation (you only have to throw a switch
in the unit to change which sideband is emphasized and which is
suppressed). I just want to emphasize that this only applies to the
areas adjacent to a null, not to the main lobe of the signal.

Sometimes there will be areas where the signal may worsen. Changing
which sideband is suppressed may cure this.

Another negative is that the system can be difficult to interface with
processing. If you process aggressively, the audio phase shifters in
the unit will introduce tilt. There are two ways around this. The
Power Side contains adjustable hard and soft clippers. You may opt to
defeat the clipping in your processing, and use the clippers in the
Power Side for peak control. This is the most common solution. The
best approach is to use a true matrix stereo processor, and process at
the insert point of the system, which is downstream of the phase
shifting. Here the signal is sum and difference, which is why matrix
processing is needed (never put anything here that disrupts the
amplitude or phase between the sum and difference). The only processor
that does this 'off the shelf' is the AM Analog Optimod (stereo
configuration). The CRL stereo unit can be made to do this, but it
requires modification.

You can see at this point that there isn't a simple answer. The
problems the system solves are complex and difficult to analyze. But
there are stations that simply would not be without it. Most of these
are on the east coast (probably because that is where Kahn is located). 
Mr. Kahn is normally quite willing to refer people to users, and users
are usually quite willing to share their stories. One hint: ask to
speak to the general manager. Engineers usually are not as willing to
endorse things they don't understand, but the managers know what it
brings to the bottom line, which ultimately is what matters.

Chris Hays Chief Engineer
KRLA(AM)/KLSX(FM), Los Angeles
chays@compuserve.com 


There is a zip file on Barry Mishkind's Old Radio web page which
"includes a number of messages shared" on the PowerSide subject,
according to Barry. The URL is:

http://www.oldradio.com/current/powrside.zip


I have an audio interview with Leonard Kahn in which he discusses
PowerSide and other issues. If anyone's interested, please contact me.

Mark Durenberger
durenberger@teleportmn.com