Date: Thu, 15 Jul 2010 19:44:21 -0700
From: "Bill, W6WRT" <dezrat1242@yahoo.com>
Subject: Re: [Amps] Size of Parasitic Chokes
The tricky part is using enough inductor to be effective without
having so much that the resistor overheats when operating on ten
meters. A delicate balancing act, believe me. You can simplify the
balancing act somewhat by using a high wattage resistor, say 25 watts
or more. Typically, hams use a resistor of only a few watts or so and
that is where it gets dicey. The resistor must be non-inductive, of
course.
## the cheap and easy way out is a 20-25 watt globar. On the 3CX-6000A7
amp we used a 50 ohm- 90 watt globar. [3/4" OD x 5" long] ..and a few wraps of
3/4" wide cu strap... [1"OD]. The C from anode to cathode is very low to
start with,
a lot lower than a 3CX-3000A7, so I doubt it even requires a suppressor at all.
The installed suppressor doesn't affect the eff at all. I had a bunch of these
90 w
globars on hand, so we used one.
## Just before Henry radio stopped building the 8-K ultra, they had dispensed
with the suppressor altogether. The original suppressor was 3 x 150 ohm , 1%
5-watt resistor's in parallel. They were not reliable on 10M. Next up was
their
infamous single turn ,chrome plated strap, 1" wide material, and super
thick...
and No resistor. Henry Radio then tells me the chrome did nothing, and then
went with a single turn of silver platedCU strap [ again , NO resistor].
The final version was no suppressor at all. Apparently the stability problem
was
traced to the 2 x piece plate choke they used [20 uh + 180 uh] It was
replaced with
a different version 2 x piece plate choke... and the stability problem
..vanished.
The 8k used a 3CX-3000A7 .
## IF the amp is NOT going to be used/designed for 10m, then you have less
problems with resistor's burning up. Rob won't have any problems with his
proposed 160-40m amp. I used the eimac plastic sockets and strapped to
chassis.
The best approach of all is to design the amp so no suppressor is
needed in the first place. This can be done by ensuring that the
cathode to ground input circuit has very low impedance at the VHF
parasitic frequency. If the impedance there is low enough, the tube
will not have enough gain to oscillate. Problem solved.
## Agreed. I noticed the Drake L4B uses C1 + C2 values on their
tuned inputs, that are double the value of those used in the SB-220.
The L4B has flat input swr across each band. It's clear that the heath
SB-2220/221 has a Q that's through the floor. Both the drake and heath amps
are 2 x 3-500Z types.
## On my HB 3x3 and 3x6 amps... we used a pair of broadcast variable
caps [ 17-540 pf per section, and all 4 x sections in parallel].. and a tapped
4 uh coil [ 17 x turns of 6-8 ga wire on a 1.5" ID]. Both caps padded on 160m
only...
[via a 2nd wafer on tuned input switch, which pads the C1 cap... and the 160m
position on the
1st wafer pads the C2 cap. 4 uh coil is not tapped on 160M, since it's all
being used, so that
free's up the 160m switch pos. ] I tried a lot of different loaded Q's on the
various bands.
Since both metal tubes are aprx 50 ohm inputs, we ran 50 ohm coax from the
C2 cap to
a bird wattmeter + dummy load, to test the tuned input. Another bird dummy
load on the input side.
With 200w drive.. [and caps tweaked for flat swr] the output wattmeter was 200w
on low
bands... and only 160W on 15M. [ both amps only cover 160-10m].
Just by INCREASING the coil by 1/4 turn, and DE-creasing the C1 + C2 caps a
tiny bit,
the power out, on output of the tuned input shot up to 190 W !
## so u don't want the loaded Q to be too high either!!
## I got the 2 x broadcast variable + tapped coil from Ron, VE7XR, who used
a
carbon copy of it on his 2 x 4-1000 amp. His tuned input version was
80-10m...
so neither of his broadcast caps were padded. [ no 160m] Ron got the idea
from
Orr's last book, where Orr used a pair of broadcast variables and a single
coil.
Orr depicts these tunable PI nets on his 160 + 80m monoband amps. His 160m
tuned input used permanently padded broadcast caps.
## Ron just built a tapped coil version, so it works on 80-10m. I took it one
step
further, and padded both caps for 160m. It was after a few go arounds with John
Lyles, that I added the 2nd bird line section on the output side. Since a
pair of
4-1000's and also a 3x3, 3x6 are aprx 50 ohms input Z... the length of coax
from
C2 cap of tuned input.. over to the socket[S] is not critical, and can be
viewed as a
50 ohm line extension. Lengths up to 18" have been used with great success.
## another trick to increase eff on higher bands is to install a 50-200 pf
doorknob
right at the socket. It's in parallel with the C2 cap of the tuned input
anyway. This
puts more of the 'C2' right at the socket itself, and increases tank eff.
## having slightly too high a loaded Q on the upper bands was an eye opener. It
was
the same calculated loaded Q as the low bands. Who would have thought to have
stuck
a 2nd wattmeter on the output of a tuned input? Yes, it is voodoo. It's all
voodoo,
then toss in stray C and stray L... and it's even more voodoo.
Jim VE7RF
73, Bill W6WRT
_______________________________________________
Amps mailing list
Amps@contesting.com
http://lists.contesting.com/mailman/listinfo/amps
|