DECREASING OUTPUT

[KB5MD]

Why do some transmitters have a decreasing output the higher the frequency?  I have a 1949 model Mackay marine transmitter that is supposed to run 200 watts on 2-24 mhz, but
it will do 240watts on 160meters, 225 on 75 meters and 125watts on 40 meters.  Above 7 mhz, you can forget it!  All power levels listed are output watts.

[W7TFO]

Tube inefficiency & dielectric losses as frequency goes up.

73DG

[K1JJ]

Sounds like the rig may have a problem, since the commercial specs say 200 watts up to 24 MHz but you are seeing only 125 W on 7 MHz.  Alignment maybe?  Swap tubes?


In General:

Skin effect (IR losses) will rear its head on ALL RF surfaces as frequency goes higher. This includes RF connections, switches, tuning caps, coils, coupling caps, relays, bypass caps, etc., - anywhere RF flows.  

So coil stock that worked well on 160M may actually be very lossy on 10M unless the cross sectional wire diameter is increased substantially.

L/C  Q will also have a big effect on various bands.  Excessively high Q will eat up power needlessly.  Keep the linear amplifier final Q around 12-15 for a good compromise.

T

[KA2DZT]

Excessively high Q will eat up power??

[K1JJ]

Excessively high Q will eat up power??

Sure, in the form of heat from resonant circuit circulating I/R losses in an RF power amplifier tank.

As an extreme example, tune up a standard final amplifier with a Q of 12 at 1 KW and the tank coil runs cool.  Change the Q to 100 (less L and more C) and that same tank coil will have a meltdown due to high circulating RF currents within the LC resonant circuit.   IE, higher Q in a parallel resonant tank circuit results in higher circulating current.  It can be countered by using physically larger sized tank components.

T

[KA2DZT]

Excessively high Q will eat up power??

Sure, in the form of heat from resonant circuit circulating I/R losses in an RF power amplifier tank.

As an extreme example, tune up a standard final amplifier with a Q of 12 at 1 KW and the tank coil runs cool.  Change the Q to 100 (less L and more C) and that same tank coil will have a meltdown due to high circulating RF currents within the LC resonant circuit.   IE, higher Q in a parallel resonant tank circuit results in higher circulating current.  It can be countered by using physically larger sized tank components.

T

The higher Q will create higher peak voltages which should create higher output power or am I wrong.  You're right the I/R losses can be lessen by a larger tank coil.

[K1JJ]

In theory you are correct. By right, if the tank coil and surrounding associated RF components were of infinite size and had zero loss, (or were "super conductors") then the Q increase would have little effect in a power amplifier. Higher Q will even create the benefit of a slightly better linearity. However, in the real world, we usually work with smaller tank components that become lossy and get to a point of diminishing returns when the Q goes much above 25 or so, especially on the higher bands with skin effect.

It's all about real conductor size, skin effect and resultant I/R losses generating wasted heat.

T