[Amps] new 2200m/630m amplifier - distorted waveform output

[Jim Thomson]

Date: Sat, 2 Jun 2018 14:31:47 -0500
From: "Jim W7RY" <jimw7ry at gmail.com>
To: "Roger Graves" <ve7vv at shaw.ca>, "Manfred Mornhinweg"
<manfred at ludens.cl>
Cc: <amps at contesting.com>
Subject: Re: [Amps] new 2200m/630m amplifier - distorted waveform
output

"the heatsink is cooler running at 200W output than at 125W output"

<This is typical on both VHF and HF amplifiers. Trying to turn down a 
commercial VHF or UHF repeater amplifier (or mobile radio pressed into 
repeater service) down to keep it cooler doesn't usually work. It just gets 
hotter at low power. Karl, AK2O, has had success reducing the DC voltage 
from nominal 13.8 volts down to 10 volts. The PA is much more efficient. A 
100 watt PA can easily be run at 90 watts without over heating when run at 
10 volts.

Same for an HF solid state PA. Much less efficient at lower power that it 
was designed for.


Thanks
73
Jim W7RY

##  agreed,   I get similar results on all  4 of my  HF  xcvrs.    50%  eff  when run at 200 watts out.
But only  35%  eff, when run at 100 watts out.   PO drops in half, but current only drops a little bit.  
Diss with  100 watts  out and  35 % eff  =  185 watts.
Diss  with 200 watts  out and  50%  eff = 200 watts.  
So it does indeed run a bit.... cooler, when run at  half power, but only by a miniscule amount.
Your Z changed when running lower power, so eff goes to hell..fast.   In other cases, like what W7RY
noted above, diss gets worse. 

##  The fix of course is to reduce the voltage.  Notice on all the old  tube amps that came with a high +  low
B+  position, that  the higher V position was always  41% higher.   Double the power, the ratio is  2.
Sq  root of the ratio    2,  is of course   1.414 

## If ur supply voltage is  say 13.8 vdc, then  use     13.8  X  .707 =  9.76 vdc   Use a similar calc for 
any other supply vdc.     Eff will improve, BUT the imd wont improve  with the lower voltage

##  On another note,  Roger said he used a T type  LPF.   A  L-C-L..with the C being shunt, will  work
as  a LPF..barely.  I have given up on  L-C-L LP matching  networks.    A  PI, with the normal C-L-C..and both 
caps being in shunt, works good.

##  still on topic,  somebody e mailed me months ago about designing a GG  triode amp for use on 475 khz.   Everything
in the final design is  quadruple what it is on 1900  khz !   Doable on 475 khz..with a lot of effort on my part.
I would not consider a  GG triode or any other tube config...on 137 khz. 

Jim  VE7RF