Someone indicated removing the small mica's from the control grids to
ground in the SB-220 would be harmful for stability.
Actually the opposite is true.
The series caps increase the anode to cathode feedthrough
capacitance, even slightly decreasing grid shielding near 180 MHz
where the 3-500Z's try to oscillate.
While I'm sure the intentions are meant to be helpful, the use of an
external GDO coupled to the grid tells the user NOTHING about the
grid impedance. A GDO is simply not an impedance measuring device.
The grid effectively "floats" at some high frequency (about 180 MHz)
because of the grid to cathode capacitance. That capacitance forms a
very complex tuned circuit who's inductance is primarily the grid
lead to chassis inductance, counting the long leads inside the tube.
The circuit looks like this:
grid----int grid lead---grid supt ring--pins and leads to chassis---
^ L ^
4 pF 5 pF
---------------------------------------------------------------------
Changing the value of series C at the chassis has very little effect
on VHF signals, unless the value of C approaches the shunt C of the
grid.
That is down near 10 pF. But at that point the PA would oscillate
like crazy below the series tuned frequency because the grid would
be poorly grounded.
You can see the effects of changing the grid connections very plainly
on a network analyzer by measuring the anode to cathode transmission
loss through the tube.
The most significant change of adding series grid caps of the values
used in the 220 to a 3-500Z PA is the S21 loss decreases, meaning the
tube has MORE undesired feedback and less stability. The largest
increase in instability occurs at the lowest frequency, but the caps
also make the PA less stable at VHF because they add series impedance
to the grid path above 120 MHz, and the tube tries to oscillate above
that frequency.
That circuit was a mistake, added because of pressure from a
person who incorrectly thought the 30S1 grid circuit could be added
to class AB2 triodes. The 220 produces better IMD performance, has
flatter gain, is more stable, and is less likely to damage your
exciter during an arc IF the grids are grounded.
The correct grounding procedure is to install a ground lug at each
grid pin, and use zero lead length to connect the chassis to the
socket lug.
If you do a stability test, by removing turns from the suppressor
until the PA oscillates, you will find the suppressor needs about
HALF the inductance with the grids grounded through lugs as it does
when the caps are installed.
If you measure transmission loss through the tube you'll see the
transmission loss increases by about three dB at VHF with the caps
gone. The caps do put one slightly steeper null at about 110 MHz, but
that 2-3 dB increase in loss isn't worth the 20 dB decrease at HF, or
the 3 dB decrease at 180 MHz. The entire tube has more undesired
feedback across the spectrum, in some cases many dB worth, when the
caps are added EXCEPT at one narrow point where feedback isn't a
problem anyway.
Be careful following back yard mechanic suggestions, they may be well
intended but without proper test equipment they can point you in
the wrong direction. A GDO tells almost NOTHING about complex circuit
parameters, it's only reliable when "dipping" a single coil and cap
in an completely isolated environment. A GDO can't even reliably be
used to test an antenna trap UNLESS the trap is removed from the
antenna and isolated. It's impossible to use a GDO to measure the
internal IMPEDANCES of a tube, it can't tell the difference between a
low impedance series resonance and an undesired parallel resonance, a
resonance is a resonance so far as the GDO (and the careless
user) cares.
73, Tom W8JI
--
FAQ on WWW: http://www.contesting.com/ampfaq.html
Submissions: amps@contesting.com
Administrative requests: amps-REQUEST@contesting.com
Problems: owner-amps@contesting.com
Search: http://www.contesting.com/km9p/search.htm
|