John E. Coleman (ARS WA5BXO)
wa5bxo2005 at pctechref.com
Fri Oct 21 20:08:05 EDT 2005
I said I though it was magnetic (not capacitive) coupling.
And it is not necessarily grid coil to plate coil, but that is a
place to start. Shield the input circuit and the exciter as well. Get
yourself some wood of fiberboard and make a temp mount for the final coil
and link and move it any where but where it is to see what the effect is.
Document the reading before and after.
Here is a check you can make feed the RF from the exciter into the
output link (you won't get much load on the exciter but you can pull a good
arc from the output tank) then look at the input link with a scope. The
idea is that no RF should pass from output to input or from input to output
while the supply to the plates is disconnected. If it does then and it is
different from one direction to the other then I would bet it is some type
of inductive feed back that you are experiencing. This can also be
neutralized with a third link but it is very critical to position. It is
generally easier to just eliminate the magnetic feed back rather than
neutralize it. You might need to put a dummy load on each of the coil links
while doing the testing and measuring; it makes the tuning less critical.
The thing about magnetic feed back is that you might get it all
working good on a dummy load but when the antenna starts working it will
radiate signal into all parts of the rig and if the grid circuit or the
exciters circuits are exposed, or if the RF comes in from the AC line it can
get back into one of the input circuits and screw with your mind. If you're
absolutely sure the exciter is shielded OK then it might be easier to move
the grid coil around and shield it. Temporary shielding can be made from
card board and tinfoil to test the effects. If the exciter is picking up
the radiated RF from the antenna then you will need to do better shielding
and bypassing on the exciter or move the antenna farther away.
From: amradio-bounces at mailman.qth.net
[mailto:amradio-bounces at mailman.qth.net] On Behalf Of W5OMR/Geoff
Sent: Friday, October 21, 2005 12:03 PM
To: Discussion of AM Radio
Subject: [AMRadio] neutrailzing
Seems I have issues, still.
Personal issues notwithstanding, I'm talking about the transmitter ;-)
John/WA5BXO was here before and his diagnosis is that the final tank
coil is too close to the grid tank coil, and there's some capacitive
interaction between the two.
After reading Don/K4KYV's method of using a sensitive wattmeter on the
output of the final, I thought "hmmm... this 20MHz scope will take 40vpk
- surely I can drop the output of the Viking II (used as an RF only
exciter) to get a reading, and then start from there", and that's what I
did. I disconnected the plate B+ supply to the final, disconnected the
AC input voltage to the final power supply (don't want that 'snake'
whipping about and striking me, with a couple of thosands volts of DC in
it's fangs!) applied grid drive to the final, and reduced the drive from
the Viking II exciter as much as possible. Getting a reading wasn't a
problem. Neutralizing worked 'ok' also, except.... the stray
capacitance my hand created when near the final, would change the scope
I said "Self (said I), what you need is a long, non-inductive
screwdriver to adjust that final with." Failing to actually posses one
of these critters, I decided to -make- one. Glanced around the shack
for something that would be suitable, and found one of those fiberglass
rods, like what you find at Home Depot for driveway/curb marking.
Checked it real quick, but flopping on the exciter again, watched the
scope, and stuck the fiberglass rod in the area of the final tank.
Absolutley -no- change in scope readings. PERFECT! Now, to find a way
to make a 'slot' screwdriver on the other end.
Enter into the picture the World's Famous, Harbour Freight $9.95 dremel
tool kit knockoff. I slid a grinding stone in the end of that little
dude, and grabbed the end of my fiberglass rod, and proceed to make me
an insulated, l-o-n-g screwdriver, for nuetrailzing my final. I decided
that 36" was acutally too long, so I cut the thing in half. A Foot and
a half long is far enough away from the final, as to not affect the
tuning of it.
THAT worked beautifully. Got the output from the tank with now
'up-to-normal' exciter drive to show much less RF leakage through the
final and on to the scope than when I started (well, it's down as far
as I can get it, anyhow) but, when all voltages are applied again, the
grid meter doesn't 'peak' where the plate meter 'dips' (resonance) The
maximum output power and minimum plate current used to be around 100mA
apart, now they're around 20mA apart and I can't get it any closer.
Here's a picture of the topside of the RF deck.. the Tank coil is obvious.
And, the bottom, where the grid coil sits
It's a steel Chassis. I wouldn't think that there'd be interaction from
the grid to the final, but apparently there's enough stray capacitance
to make things kinda haywire.
Perhaps, moving the grid wires where the 'cross' back underneat the
chassis, and only allow them to come up at the grid pins of the 250TH's?
Thanks in advance.
73 = Best Regards,
AMRadio mailing list
Post: mailto:AMRadio at mailman.qth.net
AMfone Website: http://www.amfone.net
AM List Admin: Brian Sherrod/w5ami
More information about the AMRadio mailing list
This page last updated 22 Jan 2018.