Date: Mon, 12 Dec 2016 10:01:33 -0500
From: "Michelle" <michelle_melt@windstream.net>
To: <amps@contesting.com>
Subject: [Amps] Drake L4B to 6 Meters Help
Hi Group,
I have converted a Drake L4B to 6 meters and I seem to be stuck at 750
watts output, I have tried a PI input (no change) , different tank coils (less
output), The current setup listed below I have checked both input and output
with a antenna analyzer at 50.125 and they both seem to be right on spec.
Thanks in advance for your comments.
Plate volt 2300 under load
plate current .650
Grid 220
Drive 100 watts
Output 750 watts
Input circuit T match (works as it should, flat match)
Plate cap 3-30 pf vacuum variable
Load cap 25-350 pf (original Drake plate cap) 1/3 meshed at max output.
Tank circuit PI-L, L1 3 1/2 turns 1/4" copper tubing 1 1/2" id, L2 4 turns 16
ga 1" id.
RF choke 42 turns of 22 ga on 1" porcelain form
Bifilar choke Stock Drake 20 turns each leg of 10 ga on 5/8 ferrite rod
All grid pins directly grounded to chassis
Parasitic suppressors are 47 ohm 2 watt with 1 turn 16 ga
T/R relay changed to vacuum relays
What am I missing here? Has to be something really simple I'm overlooking,
Thanks again for your comments,
73s
Michelle
WB8QGR
## I was going to suggest using a PI for the input instead of a T...but you
tried that already.
A buddy tried the T input on his 2 x GS35B 6m GG triode amp..and couldnt get
the input swr
flat. Swapped to a Pi, and all is well. Less loss with a PI input, since
only one coil. Dont go cheap
on the coil gauge, since skin effect on 6m is pronounced. Use something like
10-12 gauge..or better
yet, cu strap, wound flat. Like .25 to .375 inch wide cu strap on a real
small ID. Then u end up with
no stray C between adjacent turns. Strap coils are like 2 x knife edges facing
each other.
## A PI net for the tuned input is the same as a low pass filter. You want
harmonics to be suppressed, right
at the input to the cathode. And I dont mean harmonics from the xcvr.
## Your loaded tank Q with a PI-L on the output will be sky high. I would
dump the PI-L in favour of a
modified PI net. The pair of 3-500Zs already has loads of stray C between
anode to grounded grids.
That stray C is in parallel with the 3-30 pf tune cap. Insert a small value
of uh, like .2 uh, in the form of
a strap coil, .5 to .75 inch wide cu strap, and small ID, just BEFORE the
main PI net.
## The tube C plus the small coil forms a step down L network. This will
transform the plate load Z
from 2081 ohms ( 2300 vdc and 650 ma) down to far less, like 500- 1000 ohms.
Then the PI net
can easily be built to work with the lower plate load Z. Ok, now you can
achieve a lot lower loaded
tank Q. Circuit is then called a L-PI. Harmonic suppression is excellent.
EFF will improve a bunch.
It can also be used with a PI-L.....in which case it becomes a L-PI-L.
## .25 inch OD tubing has the same aprx circumference as .375 inch wide
strap. Strap conducts RF on
BOTH sides. Tubing only conducts RF on the outside of the tubing. .5 inch
wide strap = .318 inch OD tubing.
Again, strap, wound flat, has no stray C between adjacent turns. Easy to wind
flat strap coils vs tubing coils,
+ overall diam is reduced greatly.
## what are you using for plate block caps ? The drake OEM should more than
suffice.
Watch out with the wattmeter setup. A buddy with a hb 8877 on 6m had lousy
eff. He tried rewinding the
tank coil several times, and every other trick in the book, like rebuilding
the tuned input......all with very very
slight incremental increases in eff. Turns out the damned bird Slug he was
using....was out to lunch.
After a calibrated slug was installed, measured eff was way up. Problem
solved. The basic amp was not
the problem, the slug was.
Jim VE7RF
_______________________________________________
Amps mailing list
Amps@contesting.com
http://lists.contesting.com/mailman/listinfo/amps
|