[Amps] Fw: Good amp to buy /amp notes

[Jim Thomson]

From: Jim Thomson 
Sent: Friday, January 29, 2010 8:36 PM
To: amps at contesting.com 
Subject: Good amp to buy /amp notes

You may be surprised to know that the AL-82 uses the same capacitor vendor 
with the same plate spacings as the SB-220. It uses the same bandswitch 
construction as several Dentrons. Generic Modutec meters.
Carl
KM1H


## I bought a few new  Modutec meter's years ago.   They are pure junk imo.
As  for the pair of 813's... I got  1100 w out of several amps I built yrs ago  with pairs
of 813's.   I had loads of spares, and never used any of the spares. The 813  [or any 
other triode strapped tetrode, like a 4-1000, 4-400 etc] is an imd generator. 

##  I don't know why the ARRL  has this aversion to pulse tuning a GG  triode amp. I
changed the CW  dot length on my mk-v's.. from  stock of.. '10'.. down to 4-6.   This
reduces the duty cycle  [ while sending 60 wpm dots]  from 50%.. down to just  20-35%.
Then just use a pep meter, like the bird /cd.... or the array  solutions wattmeter, or scope.

## then when done, hit it  with a 3 x sec  dead cxr... just long enough to take some steady
state  grid /plate current meter readings. 

##  I have  3 x  L4B's... hooked up  nose to tail... and all tuned up on different bands.  The stock  L4B
will easily handle 2 kw  on bypass.  A rotary switch ensures  only one amp at a time is in the 'operate'
mode. [ key line form yaesu to rotary switch,  then 3 x key lines  from rotary switch to each amp, each amp
has it's op /stby switch  set to OP. ]   The modified L4B uses vac relay's  for input, output, and bias, same deal
handles 2 kw on bypass mode]

## for a good hb amp... the 3CX-3000A7  is superb.   The  225 watt  CCS  grid can be used as
a dummy load for ur  200 w xcvr !  IMD is superb.   50 watts of drive = 1.5 kw out   100 w of 
drive = 2.5 kw out, etc.  Since the grid is bonded to the cold chassis, actual CCS grid diss is a lot
more than 225 W. 

##   On brand new /rebuilt  3CX-3000A's ,  this is the procedure  I use  for max tube life. 
Thoriated tungsten Fil and blower ONLY  are  turned on.  I use a 750 VA  Sola  constant v  xfmr, that
in turns feeds a  constant, regulated  240 vac  to the variac.   Variac is set for  7.5 vac  [ measured at the
socket, via  100-200 uh chokes in each fil metering lead, + 200 ohm  3 watt mof , also in each metering lead
plus  .01 uf bypass caps. ]     Tube is  run  for a full  48 hrs [ hr meter on fil xfmr primary]  with NO  B+  and 
NO drive applied.   The getter is heat activated by the fil.    After  48 hrs, all residual gas is  purged. 

##  next, B+  and drive applied.. and tube put into normal operation.   AFTER  200 hrs have passed on the 
fil hr meter.... fil V  is REDUCED in small increments... until the pep meter JUST  drops a few watts  [while pulse tuning]  
[knee effect] .  Then fil V is increased by .1 V   At this point  fil V  is running at 15%  below nominal  7.5 vac  or 6.375 vac
As the tube ages,  fil V is increased to maintain full pep out.   I USED to use a separate fil step start set up.. with adjustable
0-15 sec agastat /P+B timer.   These days, I use the variac itself to do the controlled ramp up of fil V. [ then zero pri/sec  surge
like a step start does]    I also use the  variac to ramp down  the  fil V.. when done for the night.  A simple cam on the variac  shaft + microswitch /relay
ensures that IF the variac  was left cranked up the night b4.... that the fil V  won't be slammed on.  The variac HAS to
be reduced to almost zero, to engage the micro switch, which in turn activates a relay, which in turn, applies  
regulated 240 vac to the fil xfmr.  Now, you can increase the variac, slowly, to desired voltage [ as measured on
a  fluke 87 or HP DVM] 

##  run any thoriated tungsten tube in the above manner, and the tube will last forever.  Blower over run
is also used, and blower runs  for 5-10 mins  after the fil has been ramped down to zero.  

##  I USED  to have  rock solid  239.9 vac in winter.. and  rock solid 247.2vac  in summer.. for 10 yrs.
Now it's all over the map. As  I'm typing this, the laptop screen just got brighter, and so did  room lights. 
Now vary's  from  225v  to 243 v 

## the SOLA  cured all of these problems.. and will easily regulate to well under 1% .   IF you do use typ  step start  for
fils, it's a trick to get both the pri and sec surges the same value.  Even then, they are both too high imo.   The small variac
ramped up slowly, results in zero current surge, and no thermal shock.   I also ramp down slowly, when done for the night. 

##  I dispensed with the normal  electronic  grid  protection scheme.  Instead,  a simple  3agc fuse holder on rear apron,
with a super fast fuse installed, is simply  wired between chassis  and NEG terminal of grid meter. If excess grid current drawn,
grid fuse blows open.. and now NO path for dc grid current.  Amp shuts down asap.  Zero power  output on wattmeter's.
If you can't draw dc grid current,  amp can't be driven !    Also, the INPUT  swr  [ between  xcvr and amp]  will rise to infinity]
Xcvr  will shut down.  [ mine does anyway, since I use high swr protection on both the input and output of the amp... -7 vdc 
applied to xcvr  alc buss, also  key line is opened off, all  TX inhibit lead is disabled, the result is xcvr is putting out zero watts].

##  3agc grid fuse is simply replaced.. and you are good to go again.   This scheme  works  fb on the L4B amps  as well. 
  The super fast 3agc fuse just has to be sized correctly.   Some of the above may or may not be applicable to typ table top
amps.   One advantage of the HB amps is, you are free to size any component how you want, meter it how you want, cool it
how u want, etc.  You can upgrade any portion of it too.   I use ceramic vac cap  for both tune /load.  And vac relays  for input,
output, and also bias.   

## The tuned input consist of 2 x 4 x section broadcast variables  [16-540 pf  per section, all sections in parallel]  Both caps
are padded on 160m. Coil is 17 turns of  6 ga CU.. on a 1.5"  ID.  Coil tapped on each band.  A separate 2 wafer, 6 pos small
ceramic switch, changes coil taps  and also padding on 160m.   These days  I use a  3 x wafer switch, and 3rd wafer used to
switch the vac relay used for the plate choke relay.  Plate choke is just 180 uh  +  50 uh   Zero resonance's  from  1.8 to 30 mhz.

##  I used  6:1  Jackson bros calibrated ball drives for the tuned input.   Input swr is  dead flat across all band's.  Input tuning
is silky smooth,  and pre-set  work every time.  Pre-sets  for vac tune and load  loads work superb.  Typ vac cap requires
18-26  turns for min -max C.  Each revolution can be split into 100 x increments.. or  1/100th of a turn.  This puts  pre-sets
dead on.  Total time to change bands, including the tuned input is  45 seconds [ 160m... directly to 15m]  

sri for the diatribe.

Jim  VE7RF