|[AMRadio] Re:Reactor Wanted|
quixote2 at ix.netcom.com
Mon Nov 12 14:00:06 EST 2007
>From: Brett Gazdzinski <brett.gazdzinski at verizon.net>
>Sent: Nov 12, 2007 12:14 PM
>To: Discussion of AM Radio in the Amateur Service <amradio at mailman.qth.net>
>Subject: Re: [AMRadio] Re:Reactor Wanted
>I also run 2000 volts on the plates of the modulator, 4x150a's, and usualy
>volts on the pair of 813's in the final.
>It seems to like it that way, i have done experiments as both supplies
>have variacs, things always seem to modulate better, and sound better
>with more mod voltage.
That approach flies in the face of decades of design convention; as a rule, AM
rigs either use the same supply for mods and final, or as the rig gets bigger,
the mod has it's own supply which is usually a lower voltage than the PA plate
The only reason I can see for improved performance using this approach would be
that when used conventionally the modulator operating parameters were incorrectly
chosen for low distortion AT THE POWER LEVEL REQUIRED. I've noted over the years
that most ham designers tend to be very optimistic when it comes to estimating the
power that they'll get out of a modulator, and tend to not recognize what they'll
lose in the transformer, especially those losses that come from a significant
impedance mismatch between the mod plates and the PA load impedances.
The usual result is that they can't get to 100% modulation, or if they're using a
pair of pentodes or tetrodes as modulators, they find it necessary to run the tubes
well into the nonlinear area of the curves caused by the action of the screen grid
in order to get enough power.
Jacking up the plate voltage to the modulator is ONE way to get the additional watts
needed without correcting the design defects that caused the problems in the first
place. In addition to giving more audio power, raising mod plate voltage will also
increase the amount of low distortion headroom the modulator has available.
Other than that, I can't see ANY reason, either in terms of physics or intuition,
for raising mod plate voltage above the PA causing an inherent increase in audio
On the other hand... if the modulator parameters have been properly chosen in the
design, AND losses in the mod transformer and due to the (almost inevitable) minor
impedance mismatch are taken into account (so that it's not necessary to drive the
mod tubes into nonlinearity) when the mod is run at the same B+ voltage as the PA
or less, jacking up modulator plate voltage shouldn't make ANY difference whatever
in resulting signal quality... except INCREASING the distortion, caused by the
change in mod tube plate to plate impedance creating a mismatch with the PA load!
What it boils down to is this...
It's GOOD to design a modulator that's capable of producing MORE than 50% of the PA's
input (i.e., a mod that has some headroom) no matter WHAT the plate voltage of the
modulator is relative to the PA. If it's got headroom, it's gonna distort less.
It's BAD to underestimate mod transformer losses (figure 20% or more of the audio you
put into it ain't gonna come back out). It's even WORSE to pull another common ham
stunt, driving your mod transformer core toward saturation (face it... mod transformers
are getting harder and harder to find, and they AIN'T CHEAP anymore. Using too small
a transformer for the design is VERY tempting sometimes).
Mr. T., W9LBB
More information about the AMRadio mailing list
This page last updated 21 Oct 2017.