[ARC5] Modeling the PA stage

[Mike Hanz]

You must have the unfortunate Internet Explorer as your browser, unable 
to view any .gif files larger than 100kBytes or so.  Strongly suggest 
downloading Opera, which also lets you zoom in and out on images (great 
for ebay viewing), or any of the Mozilla browsers.  They're also less 
vulnerable to cyberattack at the moment.  I've reduced the resolution of 
the last two pages and reloaded them in .jpg format for those of you who 
are saddled with the sad Microsoft product, but it won't print out as 
crisply as the original .gif version.  (I left the first page as a .gif 
so you can see the difference.)  I also added a plot of reactance versus 
frequency on a DC-4 antenna from Sandretto for those who enjoy more 
graphic explanations.  You can see the switch from capacitive to 
inductive reactance a couple of times on that antenna...a DC-3 only had 
it once in that range, at about 5MHz.

This is an area of much compromise, Ian.  There are huge variations in 
antenna characteristics, depending on whether you are talking about a 
fighter plane or a bomber, a fixed (cockpit to tail) antenna or a 
trailing wire antenna, not to mention a spectrum that went from 200kHz 
to 18MHz or more, so one has to carefully caveat almost every darn word 
to make it understandable.  As Robert Downs observed, the command sets 
had a series capacitor you could hook into the antenna for certain 
conditions - we'll come back to that in a moment.  The ART-13 system has 
a *shunt* capacitor (three sections, actually - 
http://aafradio.org/flightdeck/CU-24-ART-13.jpg) that is used *only* 
when your particular situation approaches the high reactance or 
resistance values shown in the DC-4 plot.  It is avoided like the plague 
at any other time because it reduces the series capacitance of the 
aircraft antenna.  The ART-13 has no similar series cap, but the tank 
circuit has an unusually wide impedance matching capability compared to 
other aircraft transmitters of the period.

So, to answer your question about the reactances, you will note that 
they do in fact change signs on each side of resonance.  What was done 
in the command set tank design (and a lot of other aircraft transmitters 
of the period) was to "swamp the bugger out", essentially shifting the 
entire set of reactances toward the capacitive end of the scale.  Easier 
to deal with an antenna that is almost always capacitive than one that 
is a switch hitter... 

Sandretto refers to this in his discussion (essentially explaining the 
action associated with the command set series capacitor) by saying, 
"[This series capacitor] acts to cancel out the positive reactance of 
the antenna such as frequencies above 5,150kc on the DC-3 antenna" [and 
above 3.2MHz and 8.8MHz on the DC-4 antenna]. 

"Although the antenna reactance may not be positive, it is often 
necessary that [the series] condenser add negative reactance so that 
[the impedance presented to the PA plate/cathode driving point] may be 
made large enough to provide the necessary output tube load."  (Words 
within the [brackets] are mine to avoid all the subscript and figures 
necessary to describe the action.)

Have to remember that frequencies were assigned for missions in advance, 
so it wouldn't be up to the flight crew to have to make changes to all 
this stuff in flight.  It would all be carefully thought out and 
connected/tuned up on the ground for the particular aircraft and 
associated transmitters.  All the verbiage in the radio manuals is for 
emergencies, not normal flight crew duties.

73,
Mike

Ian Wilson wrote:

>Mike et al,
>
>Excellent information, thank you!
>
>The Aircraft Antenna Design article - only the first page loads when I try
>to look at it. Broken links maybe?
>
>I notice that for large planes (assume that ARC-5's, often being used aboard
>bombers, would be designed for this service), the quarter-wavelength
>frequency would be under 5MHz. Doesn't this mean that at 7MHz, a typical
>antenna would look like a few ohms in series with an inductive (not
>capacitive) component?
>
>Must go and dig out a few books, thank you all for the info.
>