[Amps] TL-922 80m load issue

[Jim Thomson]

Date: Wed, 14 Sep 2016 23:20:48 -0400
From: K2CB <k2cb at comcast.net>
To: Amps at contesting.com
Subject: [Amps] TL-922 80m load issue
Helping a friend with a TL-922. 

On 80m, it is suffering from not enough loading capacitance. The peak at low drive levels is at 1.  If driven to 1kw the peak it not attainable, as it appears to be past the lower limit. 

I suspected that maybe the padding capacitor for 80m, or its associated band switch wiring failed  Also the second portion of the load variable cap which is switched in for 80m.   All tested good, including the band switch. 

The issue seemed to begin when both tubes were replaced with new RFP 3-500ZG tubes. At that time the 4.7k resistors and 470uh inductors on the grids were replaced. The original external wire wound style inductors were NLA, so standard sealed resistor-like style 470uh inductors were used. 

On the upside, the amp now does 1000w with only 50w drive now.  So with that  in mind, I wonder if the lower drive requirement is related to the "lack of sufficient loading" ?  

I recall reading in the past that directly grounding the grids requires increased drive. Thus my suspicion that the new inductors may be the culprit.  

Could the lower drive condition be related to the lack of sufficient loading capacitance?

Eric
K2CB 

##  The RFP  3-500ZG  tubes are  MU= 200.  The eimac  3-500Z is  MU=130
The   RFP version  will draw LESS  idle current for a given  amount of bias V. 

## Directly grounding the grids  results in LESS drive required.   With all 6 x grid pins 
bonded directly to the chassis  with wide copper strap,  you will see a  drive reduction
of  20-25 watts.   Get rid of the caps from each grid to chassis..and ditto with the  choke
from grid to chassis...you dont need em.   The amp will be far more stable with the 6 x 
grid pins bonded to chassis.    Directly grounding the grids has been done at least 2 dozen times
that I know of..all on TL-922s..and all report the same results, 20-25 watts less drive required. 
( ditto with drake L4B, SB-220, henry anps etc, etc)

##  The original idea with the grid caps + chokes was a NFB idea. Its fubar at best..plus you then
get a dc Vdrop  across each choke = unwanted, varying bias..on top of the normal regulated bias.   

##  since the xcvr now has to put out  20-25 watts less power, the IMD on the xcvr will improve
by quite a bit, so the drive signal to the amp is now cleaner, resulting in a cleaner IMD  from the
output of the amp.  ( esp on run of the mill  100 w xcvrs). 

##  IF  you have an LCR meter, easy matter to measure the capacitance of the 80m padder.
Ok, what about 160m ??  Even if he doesnt use that band,  you might want to check  160m..
into a dummy load.    The problem is either the padder, bandswitch..or both. 

Jim   VE7RF