[Boatanchors] Short Wave Broadcast Folks:

[Richard Knoppow]

----- Original Message ----- 
From: "David Stinson" <arc5 at ix.netcom.com>
To: <boatanchors at mailman.qth.net>
Cc: <boatanchors at theporch.com>
Sent: Wednesday, April 16, 2014 9:08 AM
Subject: Re: [Boatanchors] Short Wave Broadcast Folks:


> What's the "60 Hz to HF" power conversion efficiency?
> In other words- How many KiloWatt-hours does it take to 
> run a 100KW transmitter for an hour?
> Anyone know?
>
    This is in the specs of all commercial transmitters. 
The overall efficiency is probably the most important factor 
in operating cost. The increase in efficiency from the first 
broadcast transmitters has been enormous.  The first 
transmitters were made up of a series of linear amplifiers 
running about Class-B with a low level exciter usually grid 
modulated. The original RCA Model B, which was employed at 
many early 50KW stations drew an average line demand of 
something like 250,000 watts, perhaps more. I don't know if 
this included cooling costs. These were water cooled 
requiring a system of pumps and radiators.  The next 
improvement was the use of plate modulation with Class-C RF 
amplifiers and push-pull Class-B modulators. The improvement 
in efficiency was considerable. An RCA 50KW transmitter drew 
an average of about 150KW. Average because the line demand 
varies with the modulation.  In 1935 Western Electric and 
Bell Labs announced a new "high efficiency" transmitter 
based on the inventions of William Dougherty. This is a low 
level modulated linear amplifier but uses a clever circuit 
to vary its operating condition with the amount of 
modulation. It uses two tubes connected such that one tube 
operates alone for the carrier at about saturated Class-B 
and the other tube is idle. For downward modulation the 
first tube works as a linear amplifer. For upward modulation 
the second tube begins to contribute.  However, Dougherty 
used an interesing circuit to change the effective load on 
both tubes so that at the peak of modulation the "carrier" 
tube is supplying double the power it does under carrier 
conditions and the second or "peak" tube also supplies about 
double carrier power resulting of a peak of four times 
carrier, which is what is needed. Now, the problem with a 
Class-B linear is that it must be able to handle the full 
power at 100% modulation. That results in very low 
efficiency for the carrier only, no more than about 33 
percent. On an overall or "all day" basis with normal 
broadcast modulation the efficiency is not much more than 
this. The Dougherty amp however, maintains about the same 
efficiency regardless of the amount of modulation so has an 
overall efficiency of about 66%. This is comparable to a 
plate modulated transmitter but has the advantage of not 
needing a modulation transformer. The early WE transmitters 
were hard to tune and could have fairly high distortion but 
the circuit was modified later both by Dougherty and by 
James Weldon, the founder of Continental Electronics. 
Continental is still building Dougherty type transmitters 
but they have much better operating characteristics. 
Another common high-efficiency transmitter is the pulse 
modulated type originated at Gates Radio, now part of 
Harris.  This is a so-called Class-D transmitter. It is 
essentially plate modulated with all RF stages running well 
into Class-D. The plate supply to the final amplifier is 
pulse modulated in such a way that the pulses are integrated 
in the final tank and produce AM there. The Harris 
transmitters have very good efficiency, approaching Class-C 
and very low distortion.  Another saving in both types is 
cooling. Since the plate (or collector in solid state amps) 
is dissipating less energy the cooling can be less and the 
cost of cooling is reduced.  The line demand of a modern 
Dougherty or pulse type transmitter is perhaps no more than 
25% greater than its output. Both types have been adapted to 
HF use but the pulse type requires very simple tuning, on 
the order of a Class-C telegraph transmitter so they are 
very suitable for frequency agile transmitters.
     Some time around the 1960s RCA decided to use a very 
old idea called Cheirex (I've misspelled this I am sure) 
modulation. It sounds like a good idea at first but doesn't 
work very well. Cheirex put two phase modulated transmitting 
chains together such that the two sides were each phase 
modulated in opposite directions and the output combined at 
the final tank. The fixed phase was set to produce the 
carrier and the phase varied with the modulation.  The 
problem is that the system is not linear although at first 
blush it would seem to be. Its also not very efficient since 
both sides are runnig full power all the time.  RCA called 
this the Ampliphase transmitter and it was supposed to be a 
high-efficiency and high fidelity unit. It turned out to be 
neither.  RCA jumped through hoops to correct the distortion 
inherent in the system. The 50KW Ampliphase had something 
like 30 tubes in it compared to a Continental Dougherty type 
with 9 tubes and better performance. Although some short 
wave Ampliphase transmitters were built they were not at all 
satisfactory since to maintain any kind of phase 
relationship through them required a complex tuning 
procedure.  MW broadcast stations unfortunate enough to have 
bought them found they often sounded very bad on the air and 
were less efficient that either the Continental or Gates 
type or, for that matter a conventional plate modulated 
transmitter.
    The Top-40 and later programming spelled the end of a 
lot of plate modulated transmitters because of the enormous 
amount of processing these stations used. The put the 
modulation transformers under a sort of strain never 
envisioned by the designers and many failed.  The Dougherty 
and pulse type transmitters would run at 100% modulation all 
day without problems so a lot of still servicable plate 
modulated transmitters were replaced.
    Enough of this.


--
Richard Knoppow
Los Angeles
WB6KBL
dickburk at ix.netcom.com