From: "YZ6X Sasha Kraljevich" <sasha@panet.bits.net>
To: "AMPS" <amps@contesting.com>
Subject: [AMPS] Shorted Pi-Networks Turns revisited
Date sent: Wed, 13 Sep 2000 13:54:18 +0200
>
> A while ago, there was a question involving subject of band switches
> able to short out coil turns, so no leads remain unshorted and therefore
> avoid Tesla coil effect.
The effect is not a Tesla-coil effect
per se.
The Tesla coil depends on a step-up transformer using magnetic
coupling.
The effect in a PA is one of back-to-back L networks where the
high series L in the tank and the low shunt C at the center makes
a back to back L network at some high frequency. You could also
picture this as a "T" network formed by the inductance in the tank
and the stray C at the middle of the tank.
It is the same effect in a choke, where we call it "series
resonance".
Now, someone (I can't recall who was it)
> suggested that it is more likely that if you have non-shorting band
> switch, there is greater possibility of arcs within Pi-L then with Pi tank
> I'm not sure to follow this ( I believe it is just the opposite
> situation) and
> therefore I would like to hear/learn about this more.
> I did some math for a plate impedance of 2000 ohms and Z load 50 ohms for
> Pi and Pi-L tank (example from ARRL handbook). Since there is step-down
> impedance between L1 and L2 in Pi-L, voltages across L1 and L2 must be
> less than voltage across L in Pi. Inductance of L1 in Pi-L is about twice
> the L in Pi - therefore less volts per turn then in L in Pi, which led me
> to conclusion that non-shorting band switch combined with Pi tank is more
> prone to arcs then when combined with Pi-L.
I'm not entirely sure what you mean above.
A Pi-L network is nothing more than a Pi designed to match an
impedance higher than the load impedance. That Pi section is
followed by a step-down L. The back-to-back capacitors at the
junction of the pi and L are integrated into one parallel component.
The inductance in the Pi-L is nowhere near twice the inductance in
a conventional Pi with normal selections of intermediate Z in the pi
L and with reasonable Q's, but the inductance IS larger in the Pi-L.
That means any unwanted resonance would move lower in
frequency by a small percentage and might slightly increase
voltage.
It makes little difference overall which network you use, unless the
unwanted resonance falls on or near a band you are operating. A pi
L, all things equal, moves the resonance down a bit and probably
would increase the voltage slightly. In both cases it would be a
good idea to use a "pick-up and hold" type switch.
By the way, it might be best to not call this a "shorting switch". A
shorting switch is a switch that connects to the next position
BEFORE releasing the last connection. A shorting switch is
generally a bad switch to use, because it has lower voltage
breakdown from contact to contact than a non-shorting switch of
the same wafer type.
The correct term is "progressively shorting switch" or better yet
"pick-up-and-hold" switch. The best wafer is a "pick-up-and-hold"
switch wafer with "non-shorting contacts".
Switch manufacturers consider a "make before break" contact a
"shorting contact" and a "drop before make" contact a "non-
shorting" contact.
You want a drop-before-make style contact, because otherwise the
rotor sticks out more towards the next open contacts. In the
JV9000 series switch, a shorting contact is changed to a non-
shorting by adding a ceramic riser between the fixed contacts. This
insulating riser slightly reduces the likelihood of an arc from
stationary contact to adjacent stationary contact.
In other switches, they extend the contact area on the rotor placing
it closer to the next open stationary contact. This causes a large
decrease in voltage rating.
73, Tom W8JI
w8ji@contesting.com
--
FAQ on WWW: http://www.contesting.com/FAQ/amps
Submissions: amps@contesting.com
Administrative requests: amps-REQUEST@contesting.com
Problems: owner-amps@contesting.com
|