Grid Driven 4-1000
Dennis Snyder
I read of others interested in a 4-1000A HF amp using untuned grid
drive. I have been using coffee stained 4-1000's (meaning well used
pulls) for years in grounded grid configuration, and the economy of grid
drive mentioned by others appeals to me as well.
Rick, K1BQT, commented on the merits of grid driven tetrodes in the
summer 1993 issue of Communications Quarterly. This was in conjunction
with the introduction of cheap ceramic tetrodes of Russian origin.
Svetlana, the US marketing firm, currently offers, free for the asking,
reprints of that article and one other from the Winter 1995 issue by
W6FR. These are general discussions, devoid of specific construction
details. The new Russians tubes are about one half the cost of similar
ceramic triodes and tetrodes offered by Eimac. However, like others,
I've got a few 4X1's, sockets, and chimneys waiting to be put to use.
There have been a number of articles on cathode driven 4X1000's in the
past:
5 Band Kilowatt Linear, W4OQ, Ham Radio January 1974
Ecology Linear, W7DI, Ham Radio March 1972
Separate KW Amps for the Contestman, K9LKA & W6SAI, CQ July 1969
The 4-1000A in Grounded Grid, K9KLA, QST July 1963
Build a KW Linear, W6DL, 73 Magazine
and probably many others. These I have saved over the years. These
sources provide a wealth of practicle data on the 4-1000A.
The original Eitel-McCullough data for the 4-1000a listed the following
typical operating specs:
CLASS C TELEGRAPHY
DC PLATE VOLTAGE 6000 VOLTS (6000 MAX)
DC SCREEN VOLTAGE 500 VOLTS (1000 MAX)
DC GRID VOLTAGE -200 VOLTS (-500 MAX)
DC PLATE CURRENT 681 mA (700 MAX mA)
DC SCREEN CURRENT 141 mA
DC GRID CURRENT 41 mA
SCREEN DISIPATION 71 WATTS (75 WATTS MAX)
GRID DISIPATION 6.1 WATTS (25 WATTS MAX)
PEAK GRID INPUT VOLTAGE 348 VOLTS
DRIVING POWER 14.3 WATTS
PLATE POWER INPUT 4086 WATTS
PLATE DISSIPATION 746 WATTS (1000 MAX)
PLATE POWER OUTPUT 3340 WATTS
OUTPUT CAPAITANCE 7.6 uuF
The 4X1000A is a hot tube! And they like voltage. I run mine at 5KV.
QST for May 1977 ran an article by W2AIH entitled "An Ultramodern Linear
Amplifier". This was a construction article describing a 4CX1500B using
an untuned input circuit terminating the rf drive signal in a 50 ohm
resister. The author stated that the 4CX1000A could be substituted for
the 4CX1500B with changes in grid and screen voltages. This points
to some flexibility in the circuit. These tubes are low-voltage high
current types designed for AB1 or AB2, so they are a breed apart from
the 4X1000a. However, the general configuration should offer a starting
point for an HF 4X1000a amplifier utilizing untuned grid drive.
Also in QST for May 1966, The Stanley Steamer, by W6MJG a vapor cooled
HF Amp using a 4CV1500B with untuned input circuitry terminated with 100
ohms. And "High Power Linear" by W6HHN again using a 4CX1500B with
untuned input circuitry and a novel grid bias arrangement. "A Low-Drive,
High Power All-Band Tetrode Linear Amplifier" by KN5S, CQ for July 1990
using untuned rf drive circuitry with a 4CX1000A. Finally, "Compact AB1
Kilowatt" by W6KEV in QST for November 1957 using a 4CX1000A with
untuned rf input using a 100 ohm resister.
And you'll want to read "Protecting Power Tetrodes" by KN5S in QST for
November 1989. Also "Understanding Tetrode Screen Current", by W6EMD in
QST for July 1961.
As time permits I'm going to start experimenting with untuned rf input
and the 4-1000a. If your're just starting out with a 4X1000a, know that
Eimac specifies .6 in of water back pressure for cooling. Read and
Heed! They will last almost forever in amateur service if you keep it
cool. This means you need a 5 or 6 inch blower three inches wide
running at 3000 rpm. Is it noisy? Yup. You'll know you have too much
air when the tube blows out of the socket.
Lastly, has anybody done this already???
73 de Dennis KY1S
Mark Mandelkern
Dennis -
Looks like an interesting project! I haven't used the 4-1000A, but had
one on the shelf for years and did think a bit about it. The main
difference, compared to the 4CX1000A, is the much higher grid
bias, and thus higher driving voltage. You will not be able to put 50
ohms directly across the grid. I think I remember that I once calculated
that 50 ohms at the input jack, then a 1:4 step-up ferrite xf, would
result in reasonable drive power requirements. Or instead, 200 ohms at
the grid might be even better. You'll have to work it out. Maybe
you'll need 1:9, I forget. The screen voltage used will greatly affect
the grid bias setting.
73, Mark KN5S
--
Mark Mandelkern
Las Cruces, New Mexico, USA
W8JI Tom
Hi All,
The correct AB1 formula for a simple shunt grid resistance is: R=
Eg^2 / 2Pdrv
A 50 ohm L pad can be calculated using ohm's law: .707bias / sqrt(p/50) =
Rgrid. The input series R would of course be: 50 - Rgrid = Rs
With 90 watts and 56 volts of bias, 30 ohms of grid-cathode shunt
resistance fed by a series 20 ohm resistor would provide both proper grid
resistance and exciter loading.
Years ago I ran a pair of grid driven 4-1000A tubes, but scrapped the idea
because of tuning and adjustment difficulties. AB1 grid driven amplifiers
are tough to tune correctly, even the slightest amount of grid current or
improper loading makes them poor IMD performers.
I eventually tried to run the 4-1000A's in the Collins circuit used in the
30L1, but was unsucessful due to less than sufficient catode to anode
shielding. The only circuit that worked well was conventional grounded
grid with screen and grid bias applied. There was some slight performance
improvement over no screen or grid bias in grounded grid.
As with any tetrode, screen regulation and grid current in both grids was
critical. As time passed I scrapped the 4-1000's in favor of a grounded
grid triode. IMD performance and ease of adjustment improved greatly with
this move.
Looking back on it all, there was almost no advantage for all the effort
in building the screen and bias supply in the cathode driven 4-1000
configuration, althought it was easy to use and relatively clean even when
slightly mistuned. The thing I wanted awas more power gain, but gain
barely increased over a conventional grounded grid circuit with no screen
or grid bias voltages.
The grid driven configuration gave greatly improved power gain, but was
pretty fussy to tune properly on SSB. The largest problem was creation of
IMD even when scope and meters appeared correct. I think the Collins
circuit (used in the 30S1 4CX1000 amp) would have been the best circuit
but the poor screen grid shielding in the 4-1000A's made regeneration a
problem.
There are pros and cons to every circuit, and I guess the best choice
depends on user goals. But the 4-1000A is a far cry from the 4CX1000 in
operation and stability.
73 Tom
Zack Lau
W8JI Tom wrote:
> Years ago I ran a pair of grid driven 4-1000A tubes, but scrapped the idea
> because of tuning and adjustment difficulties. AB1 grid driven amplifiers
> are tough to tune correctly, even the slightest amount of grid current or
> improper loading makes them poor IMD performers.
I wonder if this problem might be negated with a
simple microprocessor controlled tuning circuit?
A microprocessor might even be useful for grounded grid
amps--perhaps to automatically tune the input circuit with
relays. This could allow a higher Q PI input circuit without
having to worry about excessive mismatch to the driver.
> The grid driven configuration gave greatly improved power gain, but was
> pretty fussy to tune properly on SSB. The largest problem was creation of
> IMD even when scope and meters appeared correct. I think the Collins
> circuit (used in the 30S1 4CX1000 amp) would have been the best circuit
> but the poor screen grid shielding in the 4-1000A's made regeneration a
> problem.
The extra power gain is quite useful if you are
attempting to generate the cleanest signal that
technology allows.
Solid state class A driver amps can have good
high order IMD performance, but become impractical
much above the 10 watt level.
As Tom indicates, you want to use the best parts
for the job--which isn't always practical for those
on a budget. For instance, I've noticed that you
tend to get better IMD with higher voltage tubes.
What do you do if you only have a 2 kV supply?
Start over and get parts for a 3.5 kV amp?
Zack KH6CP/1 zl...@arrl.org