I am building an AM transmitter using a 829B final tube (with the two tetrodes
in parallel) operating in class-C (750V on plate, 200V on screen).

The modulation transformer has a single secondary winding.

To obtain both plate and screen modulation, I am planning to adopt the
well-known circuit whereby the tube screen is fed by the 750V plate supply
through a high-power series resistor.

I have two questions at that regard.

FIRST QUESTION

According to the ARRL Handbook (1968 edition), the screen resistor must be
connected to the "hot" end of the modulation transformer secondary. This seems
quite logical to me.

On the contrary, the EIMAC "Care and feed of power grid tubes" booklet states
that the screen resistor must be connected to the "cold" end of the secondary.
Their argument is that, when plate voltage increases (due to modulation), the
screen current decreases and so does the voltage across the screen resistor,
thus resulting in a higher screen voltage. In other words the screen is
subjected to an automatic modulation effect.

Does anyone have practical exeperience on the pros and cons of the two methods?

SECOND QUESTION

To protect the final tube, I am planning to have a fixed negative bias on the
control grid that sets the the tube close to interdiction (some 10 - 20 mA of
idling plate current) in absence of RF driving voltage on the grid.

In such resting condition the screen current should presumably be very low, with
a consequent low voltage drop across the screen resistor.

The screen voltage will set at some voltage that will surely be higher than the
normal operating voltage (200V). I am not sure whether this can cause problems
to the tube.

Would you have any suggestion to avoid such voltage growth effect, other than
the classical solution of adding an extra clamp tube that, in absence of RF
drive, draws current from the screen resistor?

Thanks in advance for answers.

Tony, I0JX

Rome, Italy

Antonio Vernucci wrote:
I am building an AM transmitter using a 829B final tube (with the two
tetrodes in parallel) operating in class-C (750V on plate, 200V on screen).

The modulation transformer has a single secondary winding.

To obtain both plate and screen modulation, I am planning to adopt the
well-known circuit whereby the tube screen is fed by the 750V plate
supply through a high-power series resistor.

I have two questions at that regard.

FIRST QUESTION

According to the ARRL Handbook (1968 edition), the screen resistor must
be connected to the "hot" end of the modulation transformer secondary.
This seems quite logical to me.

On the contrary, the EIMAC "Care and feed of power grid tubes" booklet
states that the screen resistor must be connected to the "cold" end of
the secondary. Their argument is that, when plate voltage increases (due
to modulation), the screen current decreases and so does the voltage
across the screen resistor, thus resulting in a higher screen voltage.
In other words the screen is subjected to an automatic modulation effect.

Does anyone have practical exeperience on the pros and cons of the two
methods?

SECOND QUESTION

To protect the final tube, I am planning to have a fixed negative bias
on the control grid that sets the the tube close to interdiction (some
10 - 20 mA of idling plate current) in absence of RF driving voltage on
the grid.

In such resting condition the screen current should presumably be very
low, with a consequent low voltage drop across the screen resistor.

The screen voltage will set at some voltage that will surely be higher
than the normal operating voltage (200V). I am not sure whether this can
cause problems to the tube.

Would you have any suggestion to avoid such voltage growth effect, other
than the classical solution of adding an extra clamp tube that, in
absence of RF drive, draws current from the screen resistor?

Thanks in advance for answers.

Tony, I0JX

Rome, Italy

There are two other ways of handling the screen voltage in a modulated
stage. The screen can be feed by a regulated voltage supply and a
separate winding on the modulation transformer is used to modulate the
screen. A variation on this is to put an audio choke in series with the
screen supply and to optionally couple the screen to the hot side of the
plate modulator winding via a capacitor.

There are two other ways of handling the screen voltage in a modulated stage.
The screen can be feed by a regulated voltage supply and a separate winding on
the modulation transformer is used to modulate the screen.

Unfortunately, as I had written, my transformer has a single secondary winding.

A variation on this is to put an audio choke in series with the screen supply
and to optionally couple the screen to the hot side of the plate modulator
winding via a capacitor.

Yes, I had seen that method too on the ARRL Handbook (as well as in the EIMAC
booklet), but I was trying to avoid the bulk of an extra choke.

Nevertheless, browsing the Internet, I found a dealer who sells 20H 50mA chokes
of reasonable size at a reasonable price. So, I am getting more and more
convinced that that is the way to go.

Still, I would be interested to hear from someone who has actually tried that
method.

Thanks & 73.

Tony I0JX

Antonio Vernucci wrote:
There are two other ways of handling the screen voltage in a modulated
stage. The screen can be feed by a regulated voltage supply and a
separate winding on the modulation transformer is used to modulate the
screen.

Unfortunately, as I had written, my transformer has a single secondary
winding.

A variation on this is to put an audio choke in series with the screen
supply and to optionally couple the screen to the hot side of the
plate modulator winding via a capacitor.

Yes, I had seen that method too on the ARRL Handbook (as well as in the
EIMAC booklet), but I was trying to avoid the bulk of an extra choke.

Nevertheless, browsing the Internet, I found a dealer who sells 20H 50mA
chokes of reasonable size at a reasonable price. So, I am getting more
and more convinced that that is the way to go.

Still, I would be interested to hear from someone who has actually tried
that method.

Thanks & 73.

Tony I0JX
You can also just ignore the issue and only apply modulation to the
plate of the pentode/beam power tube. In theory you won't reach 100%
modulation, however given that the screen input is a tiny fraction of
the plate input you'll probably achieve somewhere between 95-99%
modulation. You don't want to go over 100%, and unless you are using a
speech compressor you won't ever hit 100% modulation peaks when properly
set up anyway.

Kenneth Scharf wrote:
Antonio Vernucci wrote:
There are two other ways of handling the screen voltage in a
modulated stage. The screen can be feed by a regulated voltage supply
and a separate winding on the modulation transformer is used to
modulate the screen.

Unfortunately, as I had written, my transformer has a single secondary
winding.

A variation on this is to put an audio choke in series with the
screen supply and to optionally couple the screen to the hot side of
the plate modulator winding via a capacitor.

Yes, I had seen that method too on the ARRL Handbook (as well as in
the EIMAC booklet), but I was trying to avoid the bulk of an extra choke.

Nevertheless, browsing the Internet, I found a dealer who sells 20H
50mA chokes of reasonable size at a reasonable price. So, I am getting
more and more convinced that that is the way to go.

Still, I would be interested to hear from someone who has actually
tried that method.

Thanks & 73.

Tony I0JX
You can also just ignore the issue and only apply modulation to the
plate of the pentode/beam power tube. In theory you won't reach 100%
modulation, however given that the screen input is a tiny fraction of
the plate input you'll probably achieve somewhere between 95-99%
modulation. You don't want to go over 100%, and unless you are using a
speech compressor you won't ever hit 100% modulation peaks when properly
set up anyway.
Another idea would be to use a second low power modulation transformer
with a separate amplifier feed off the same mike amp and with it's own
level adjustment to only modulate the screen. You'd need two level
adjustments in addition to the main mike gain control to set the proper
levels for the high power plate modulator and the low power screen
modulator. The screen modulator would only need to be a few watts,
maybe single class A 6L6 or 6V6. I realize this is getting more
complicated than just the choke in the screen circuit, but that's the
beauty of home brew, trying things out just for the heck of it.

Another idea would be to use a second low power modulation transformer with a
separate amplifier feed off the same mike amp and with it's own level
adjustment to only modulate the screen. You'd need two level adjustments in
addition to the main mike gain control to set the proper levels for the high
power plate modulator and the low power screen modulator. The screen
modulator would only need to be a few watts, maybe single class A 6L6 or 6V6.
I realize this is getting more complicated than just the choke in the screen
circuit, but that's the beauty of home brew, trying things out just for the
heck of it.

I sometimes tend to be perfectionist, and I would then not like to start a
project pursuing a compromise solution from the very beginning. So, I would like
to have the screen modulated too, not just the plate.

With regard to the separate audio amplifier for the screen, it should work but I
am not sure whether some phase shift could occur between the two amplified audio
signals (apart from the 180-degree phase ambiguity that can be easily resolved
crossing the wires of one of the two modulation transformers).

I have come to the conclusion that using a filter choke on the screen probably
is the simplest way to go, and the safest for the final tube too.

73

Tony I0JX

In article ,
"Antonio Vernucci" wrote:

According to the ARRL Handbook (1968 edition), the screen resistor must be
connected to the "hot" end of the modulation transformer secondary. This
seems
quite logical to me.

On the contrary, the EIMAC "Care and feed of power grid tubes" booklet states
that the screen resistor must be connected to the "cold" end of the
secondary.
Their argument is that, when plate voltage increases (due to modulation), the
screen current decreases and so does the voltage across the screen resistor,
thus resulting in a higher screen voltage.

Tony-

What if you connected a capacitor across the screen resistor connected
to the hot end of the transformer? This might tend to keep the DC
voltage drop constant, while coupling more audio to the screen grid.

Fred
K4DII

Antonio Vernucci wrote:
I sometimes tend to be perfectionist, and I would then not like to start a
project pursuing a compromise solution from the very beginning. So, I would like
to have the screen modulated too, not just the plate.

I think that if you do this, you have higher audio distortion even though you
have a slightly higher modulation level.

With regard to the separate audio amplifier for the screen, it should work but I
am not sure whether some phase shift could occur between the two amplified audio
signals (apart from the 180-degree phase ambiguity that can be easily resolved
crossing the wires of one of the two modulation transformers).

It depends mostly on how well matched the modulation transformers are. Most
of the group delay will be the result of the transformers.

Better to use a transformer with two secondaries than two individual amps.

I have come to the conclusion that using a filter choke on the screen probably
is the simplest way to go, and the safest for the final tube too.

Is the best possible fidelity important to you? Is having a tiny little bit
more modulation important to you? I think in the manuals to the Bauer 707
broadcast transmitters there is a paragraph or two about the compromise.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

Tony-

What if you connected a capacitor across the screen resistor connected
to the hot end of the transformer? This might tend to keep the DC
voltage drop constant, while coupling more audio to the screen grid.

Fred
K4DII

Frad,

I am afraid I am unabke to follow your reasoning.

I cannot see how adding a capacitor, even a BIG one, can help keeping constant
the voltage drop across the screen resistor.

When you switch back from TX to RX, the screen current will go to nearly zero
and the voltage drop across the resistor will then tend to do the same. A
capacitor in parallel with the resistor would not succeed to keep the voltage
drop constant, as it would get discharged in a very short time.

73

Tony, I0JX