In my National NCL-2000, I can run 200W carrier and 800W PEP, but
I do not, as 8122's are costly and the NCL's blower is really too
small. I usually do my testing at 100W carrier and 400W PEP. In
that amp, tuned up properly for each of two carrier level cases
(approximately):
loaded for 800W PEP;
200W carrier = 600W dissipation,
800W input, 25% eff.
loaded for 400W PEP;
100W carrier = 300W dissipation,
400W input, 25% eff.
The loading is adjusted such that in each case, the amp can't put
out more than about 20% over the desired PEP limit without
clipping. This gives a much lighter load on the tubes when the
unit is adjusted to run at a lower power.
If the loading was not adjusted lighter for the lower power
level, the amp would not reach its peak efficiency at the PEP
level and dissipate substantially more than the necessary power.
Loading lightly also requires a close eye on the screen current.
Impending clipping is accompanied in that unit by a sharp rise in
screen current on peaks.
Testing was done with a sine wave in order to determine these
values. There is no claim to how a voice signal will behave, but
the claim is made for the relationship between 'adjustment of
loading' and 'maximum power output utilized' having a direct
relationship on overall efficiency during high levels of modulation.
Since the blower in the amp is rather small, and mounted to suck
air in from the bottom, I built a "wind chest" out of a 12" BUD
rack with no lid, on which the amp sits. A dual 4x4"
squirrel-cage blower from a mainframe rack is mounted in the wind
chest and sucks air in the front and blows it up through the amp
and keeps things so cool that you can touch the tubes after a
full power test and not get burned. The power supply regulation
on the NCL2000 is such that in AM use, 800W PEP is about all it
will do cleanly. In SSB mode with low to moderate speech
compression it will do more like 1200W PEP due to a lower overall
power requirement.
