[Top] [All Lists]

Re: [Amps] Why does power output increase?

Subject: Re: [Amps] Why does power output increase?
From: "jim.thom" <>
Date: Mon, 6 Jun 2022 10:24:06 -0700
List-post: <>
Date: Sun, 5 Jun 2022 13:55:56 +0300
From: Victor Rosenthal 4X6GP <>
Subject: Re: [Amps] Why does power output increase?

<This morning, for some reason, the phenomenon did not occur. I was
<getting close to 1400 watts the whole time. I always operate CW in the
<SSB position -- the warning in the manual against this is (I think) only
<relevant if the original slow t/r relay is in use, which hot-switches
<with semi-QSK. My amp has a reed/vacuum relay QSK circuit.

<Because I do not use SSB, I have adjusted the idle current lower than
<normal. It is 140ma for two tubes. This requires a bit more drive but
<produces more output without generating clicks. It didn't change more
<than 5 ma (a slight drop) during my test. The Zener that is in there is
<a 50-watt one.

<I agree that it is unlikely to have anything to do with the filament

<I've heard about using diodes instead of zeners, but the curve of
<current vs. forward drop of the diodes is sharper than that of zeners.
<The temperature characteristic also compares unfavorably. So I've stuck
<with zeners. I put in a 50 watt one because that is the one I had with
<the appropriate voltage.

<I still wonder if the filament voltage should be measured while
<transmitting or on standby, given that even during a transmission the
<plate current is only drawn at a maximum of 50% of the time, and the amp
<is probably in standby with zero plate current most of the time (my
<biasing circuit cuts off the tubes in standby).

<Victor, 4X6GP

##  How hot does the plate xfmr get, when running 1400 watts out on CW ??

##  For  CW use,  you can easily reduce the  idle current well below 140
20-30 ma works superb, and will not generate any key clix.  This will
require an increase in bias V.   Just add a few diodes, like 1N5408's in
series with your existing 50 watt zener.   That will reduce your idle power
a bunch, and take a huge load of the plate xfmr.

##  series diodes for bias works superb.  Just use real diodes, like
1N5408's, or 6A10's..or 10A10's, or PM-600's. You want lousy bias
regulation ?  On an  OEM SB-220, the V drop across the OEM grid chokes is
absurd.   Each choke has 25 ohms of DC resistance.  With 300 ma of dc grid
current, you end up with 150 ma of dc grid current, per tube.  .150 X 25
ohms = 3.75 vdc.  IE: bias starts at 5 vdc  (heath used a 5 vdc
zener)....then swings up to 8.75vdc.  Hence another reason for bonding the
grids to chassis.

##  You will get a miniscule increase in V drop across  diodes used for
bias, esp if plate and grid current combined  ( = cathode current) is  <
1200 ma.

##  If you wanted the bias to be 100%  rock solid, an electrolytic (like
5-10,000uf )  wired across the entire string of diodes will result in the
bias V not budging at all, with a full bore plate +, grid current load.
I use that method on the bigger  GG  directly heated metal tubes.

##  do you have  50 hz, or  60 hz power ?  If it's only 50 hz, the fan in
the TL-922 will run quite a bit slower, with much less RPM..and cooling.
 I use 10 ga wire, from a  100 amp sub panel... to each of my L4B amps,
short runs.   Long runs of  14 ga wire will cause issues.

##  I think your power shooting up 200 watts is a tube issue. (one or both

## You could leave the resistor in the primary of the fil xfmr.... as long
as you shunt it out on TX.

Jim  VE7RF
Amps mailing list

<Prev in Thread] Current Thread [Next in Thread>