[BC] Leonard Kahn

[Robert Meuser]

This discussion has so many partial truths, rumors, selective recall, 
etc that I feel compelled to comment.

I partially agree with Rich's comments below as well as those of Tom O. 
But the entire story is much more complex. I have a fairly long history 
with many of Mr. Kahn's developments. When I was in college one of the 
stations where I worked part time was considering Mr. Kahn's original 
compatible SSB system that he attempted to sell to the world. His 
original background was with the RCA shortwave overseas stations where 
he worked with a number of very well known inventor/engineers of the 
day. I believe this was the primary influence on much of his engineering 
vision. He has always been a very driven person which is both a benefit 
and a deficit.

I worked with almost all of his broadcast products beginning with the 
ubiquitous Symmetrapeak and including the voice line, lines plus, STR 
77, STR 84, nonsymmetramod, IFO, and the few modified radios he built.I 
never touched the flatterer because I thought it was a bad concept and 
when I told him that he became very annoyed. I dealt with him on various 
levels for over 15 years. At one point I worked for a group where we put 
a number of his stereo systems on 50 KW stations. We did all the install 
ourselves so we knew the hardware well. We also had a continuing dialog 
with Mr.Kahn regarding various issues and requests for additional 
devices. He always was, as Rich indicates below, a complete gentleman. 
Every NAB he bought us a steak dinner (along with some other engineers 
in his favor) and sent small but practical gifts around Christmas time. 
The problem was even among friends he had a very thin skin. He thrived 
around people like Rich Wood who never asked hard questions but only had 
adulation for his works. When confronted with reality, things could 
change quickly. Being part of a group where we had unrestricted access 
to his equipment, we found faults that others who just got the install 
missed. Despite the heavy metal chassis and the dual power supplies, the 
actual implementation of both stereo exciters was very poor and 
difficult to adjust and difficult to maintain stability. There were some 
very basic reasons for this problem and some pretty simple cures. He 
never wanted to discuss any of that. Eventually we approached a very 
well regarded broadcast equipment manufacturer who was a supporter of 
the Kahn system and who agreed to build exciters if the proper 
agreements were in place. Leonard went nuclear when we approached him 
with this idea and immediately launched a vendetta against the 
manufacturer who tried to help him.  I had similar experiences 
previously dealing with his telephone line extenders (which were quite 
clever) where he would never accept any constructive criticism regarding 
any of his designs. This is where the two sides are clearly revealed as 
this was a case of friends trying to help friends and not somebody 
trying to steal patents.

Correcting the mis recollections of AM stereo, let's start in 1958. That 
was the year Mr. Kahn was pushing compatible single sideband. He had it 
running on a number of major stations, most notably WBZ and KDKA where 
the idea was to use opposite sidebands and eliminate mutual splatter. 
The system had two problems with the first being the way FCC rules were 
written at the time. Proof of performance rules at the time effectively 
required a proof of performance measurements to be made with a device 
that contained an envelope detector. This could have been worked around 
in several ways common to present day consultant. One of which being to 
get a more precise definition regarding the measurements or another to 
push for type acceptance of a modulation monitor employing synchronous 
detection. Leonard being pretty much of a loaner did neither. Rather he 
developed a system that in theory would compensate for the quadrature 
distortion by adding out out phase second harmonic distortion on the 
desired sideband. This worked well in the lab and on many mod monitors 
but not with the variability of consumer radios. The other problem was 
that due to the approach, bandwidth was not reduced as much as it was 
shifted. The desired sideband became wider. This is why his SSB system 
was explicitly prohibited in ITU technical rules.
The second problem was that for the system to really work the 
transmitters needed to be DC coupled in the audio path which was not 
possible in transmitters of the day. The mechanisms developed to 
accomplish this system included EER which is probably his most brilliant 
contribution to RF transmission and is how every present day digital 
broadcast system modulates medium and shortwave transmitters. It is also 
used in some cell phone and other digital transmission systems. In some 
quarters it is also known as the Kahn method.

The compatible single sideband system did not catch hold for both the 
reasons stated above and from those I know who used it, it's highly 
unstable nature which required a lot of complex maintenance (spectrum 
analyzers were not that common then).  More modern technology probably 
would have solved the stability problem. At the time the FCC was in the 
process of evaluating stereo transmission systems. Originally systems 
for both AM and FM were under consideration. There were two systems for 
AM. One was RCA which was an FM L-R subcarrier under an L+R AM signal. 
The other was one that Leonard going back to his RCA shortwave roots was 
an ISB adaptation of his compatible single sideband system. The FCC 
eventually decided that only FM being an underdog should get a stereo 
system.

In the early 70s Kahn started pushing to get AM stereo back in the mix 
and as I recall he had one or two experimental authorizations on air to 
re-prove the concept. Eventually the idea began to grow. As was already 
mentioned there were ultimately 5 systems under consideration. From the 
early 60s attempt, both the Kahn and RCA systems which was now the Belar 
system. Magnavox, Harris and Motorola. Belar wisely dropped out fairly 
early. It's closest cousin, Magnavox was a phase modulation L-R under AM 
rather than FM under AM. Although the FCC originally made it the 
standard, there were too many things that were poorly conveyed and 
caused the system to be misunderstood. In retrospect based from hands on 
experience in both installing and measuring that system, the FCC made 
the right call. It took very talented engineers at Continental to revive 
the system and in essence attempt to cleanse it of of the corporate 
bureaucratic missteps committed by Magnavox.

Although I did not feel that way at the time, Harris stepped in it with 
the way they presented their system. They had enough corporate clout and 
legal power to get a few minor changes through the FCC to make a pure 
quadrature stereo system work. It is not well known but it could easily 
do pure ISB stereo as well. Harris got hung up in the same issues that 
Kan had in the 50s, namely receiver distortion specifications. In their 
case if they could have codified proof of performance at the terminals 
of their synchronous detection stereo mod monitor, all would be fine. 
Instead they pursued various means of meeting proof specifications via 
an envelope detector. This was eventually their demise as Kahn called 
them on it in front of the FCC. Any consultant worth his salt today can 
talk the FCC out of much more than what Harris needed to do.

Motorola is an interesting story. This is a case where unlike two other 
large companies (Harris and Magnavox) that were too big to have a real 
clue, Motorola  seemed to have it's legal and engineering teams in sync. 
The system basically was a hybrid between Harris and Kahn. Like Harris 
they used quadrature amplitude modulation but to work around envelope 
detector issues, they employed a pre distortion scheme somewhat similar 
to Kahn. Since Motorola and Kahn ended up being the final two systems 
much more of a microscope was applied to the engineering details. 
Motorola had a lower level of out of band emissions than Kahn with 
Motorola the primary distortion was odd order while with Kahn it was 
even order. Both systems suffered from the same problems found in a 
typical FM receiver in that there is almost never the bandwidth possible 
to recover the sidebands necessary for perfect reception. At the 
transmitter end of the equation, Motorola had a straight path for L+R to 
get into the TX without overshoot while Kahn had to deal with Hilbert 
transform issues which causes unlimited overshoot of square waves. The 
solution was to have breakout point in the STR84 exciter and re-clip L+R 
and L-R which destroyed the generation of real stereo. The STR 77 did 
not have this problem but it created an approximation of stereo and not 
quite the real thing.

Breaking down 4 of the original stereo systems:

Magnavox was a simple to implement system and inexpensive to put in 
radios. It worked very well and sound very good. No pre-distortion was 
required. When the Continental version was installed, it was a great 
system. The system was truly a system as there were receiver chips that 
faithfully matched the transmission.

Harris had some end to end solutions but since the system changed so 
much it was hard to keep up.

Motorola also had a true end to end solution and all three of these 
manufacturers either produced chips or had the capability of doing so.

The Kahn system was never a system. There was no known means of building 
a receiver that truly matched the transmitter. Every Kahn Install I have 
ever done relied on a spectrum analyzer to measure separation while 
Harris, Magnavox and Motorola could be done from a mod monitor as is 
done with FM.

People often speak highly of the Sony receivers. The Sony chip in those 
radios did not exactly match any of the systems. If it matched close to 
any it would have been Harris and Magnavox. It is basically a quadrature 
detector for all but Kahn where they throw in a 90 degree phase shift 
network which is why they had to add the switch to go between Kahn and 
everyone else. Those radios were nowhere close to magical but they were 
a primitive start.

Those who speak of the legal issues between Motorola and Kahn often over 
state what was the real underlying issue. Norm Parker of Motorola and 
Leonard Kahn hated each others guts. It was an old feud that went back 
to NTSC committee days. They would do anything to try to screw the 
other. I could write about how colorful Norm was but I will leave to 
those who knew him.

Leonard's main problems were that he was to a degree a snake oil 
salesman in that he over sold his inventions and over stated the faults 
of his adversaries. As brilliant a conceptual engineer he was, his 
execution in actually manufacturing products was poor and he would not 
accept advice from the outside. I think his legacy lies in phase 
scramblers long found in many well regarded audio processors and 
especially EER or as some call it the Kahn method which is used in many 
digital transmission schemes.

I base this on knowing the man since 1973 and my use of many of his 
products and my independent installations of Magnavox (Continental) 
systems, Harris systems, Harris systems I converted to CQUAM, Several 
Kahn STR 77s, more STR 84s and more Motorola than I can count. That 
along with personal contact with the primary parties involved with all 
of the above systems, all being highly experienced and well respected 
broadcast engineers in the very truest sense of the word.

On 6/22/12 4:15 PM, Rich Wood wrote:
>
> I first met Leonard when we put the Kahn stereo system back on the
> air. Mexico allowed it before the FCC did. It's sad to hear constant
> criticism about his "personality." If Leonard liked and respected
> you, you couldn't ask for a kinder, more loyal friend. Kahn
> Communications is a small company without the political and financial
> clout of Motorola. Money talks much louder than technical brilliance.
>
> Rich