Scott Dorsey wrote:
wrote:
Scott Dorsey wrote:

How high speed do you want and why don't you want to use a gas tube?
With a vacuum tube, you pretty much have to run class C in order to get
complete cutoff.

I actually want to implement class E amplification with a valve.

Of what? If you want to put it on 160M you might manage it, but you are
going to have very poor efficiency.

Figure you want a switching rate less than a tenth of a cycle, if all you
want to do is just reproduce the RF signal. If you want to use the PWM
device to do the modulation as well, you will need to go much faster.

You may want to look at some of the Brown-Boveri AM broadcast transmitters,
which use switching output stages, driven by PWM oscillators that handle
both the modulation and the carrier. The output integrator is a nightmare,
even at 200M.

They are using big switching FETs. You don't really need the sharpest
square wave coming out of the output stage, though, because it's just
producing stuff that the harmonic filter is going to eat up anyway.

There are some very high speed gas tubes out there, like krytrons, but most
of them aren't very high power. The real issue, though, is what kind of
frequency you really want and if you want to do the modulation or just the
carrier with the switcher.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

Just the carrier. Was wanting to see if I could stretch a tube the
normally puts out 1 watt class A into 5 watts or more in class E. So
tell me, why the difficulty using a tube rather than a Mosfet? Don't
tubes and Fets behave similarly from an electrical point of view?

The Eternal Squire

wrote:
es, tres precis."

Just the carrier. Was wanting to see if I could stretch a tube the
normally puts out 1 watt class A into 5 watts or more in class E. So
tell me, why the difficulty using a tube rather than a Mosfet? Don't
tubes and Fets behave similarly from an electrical point of view?

The Eternal Squire


Are you sure you don't mean "Class D" ?
ISTR that Class E is a PWM switched signal fed into a low
pass filter.

Class D uses a technique that aligns the current and voltage
zero crossings with the switching, or something like that.

-Chuck

wrote:

Just the carrier. Was wanting to see if I could stretch a tube the
normally puts out 1 watt class A into 5 watts or more in class E. So
tell me, why the difficulty using a tube rather than a Mosfet? Don't
tubes and Fets behave similarly from an electrical point of view?

Similarly but not identically, and the top of the curve is what's important
in a switching application where you're trying to saturate the device. You
want something with as abrupt a transition in the curve as you can get, and
something with a huge amount of gain. But you also want something that
switches really, really fast. Tubes in general tend to be slower, lower
gain, and more linear.

Honestly, if you just want the carrier, though, it doesn't buy you anything
over normal Class C operation.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

Similarly but not identically, and the top of the curve is what's important
in a switching application where you're trying to saturate the device. You
want something with as abrupt a transition in the curve as you can get, and
something with a huge amount of gain.

You mean, the device needs to be nonlinear?

But you also want something that
switches really, really fast.

Like the 12AF4 or one of the other tubes used in VHF tuners?

Honestly, if you just want the carrier, though, it doesn't buy you anything
over normal Class C operation.

Then why are people putting so much store by class E operation for CW
amplification?

I guess back to the drawing board. What I was thinking was to use a
genny as VFO to
inject a carrier into a buffer amp driving a CW amplifier.

wrote:
Similarly but not identically, and the top of the curve is what's important
in a switching application where you're trying to saturate the device. You
want something with as abrupt a transition in the curve as you can get, and
something with a huge amount of gain.

You mean, the device needs to be nonlinear?

It sure helps if it is. You're really fighting things otherwise.

But you also want something that
switches really, really fast.

Like the 12AF4 or one of the other tubes used in VHF tuners?

Hmm... maybe. A sharp-cutoff pentode for VHF service might do the trick.

Honestly, if you just want the carrier, though, it doesn't buy you anything
over normal Class C operation.

Then why are people putting so much store by class E operation for CW
amplification?

Damned if I know. Class C already gets you extremely good efficiency,
and it's not like with the legal limit being what it is that you're going
to save much on your power bill. If you're running 250KW, of course, every
little bit counts, and the ability to do AM modulation at the same time
makes it very valuable for high power broadcast applications.

I guess back to the drawing board. What I was thinking was to use a
genny as VFO to
inject a carrier into a buffer amp driving a CW amplifier.

What is a "genny?"
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

Scott Dorsey wrote:
wrote:
Similarly but not identically, and the top of the curve is what's important
in a switching application where you're trying to saturate the device. You
want something with as abrupt a transition in the curve as you can get, and
something with a huge amount of gain.

You mean, the device needs to be nonlinear?

It sure helps if it is. You're really fighting things otherwise.

But you also want something that
switches really, really fast.

Like the 12AF4 or one of the other tubes used in VHF tuners?

Hmm... maybe. A sharp-cutoff pentode for VHF service might do the trick.

Honestly, if you just want the carrier, though, it doesn't buy you anything
over normal Class C operation.

Then why are people putting so much store by class E operation for CW
amplification?

Damned if I know. Class C already gets you extremely good efficiency,
and it's not like with the legal limit being what it is that you're going
to save much on your power bill. If you're running 250KW, of course, every
little bit counts, and the ability to do AM modulation at the same time
makes it very valuable for high power broadcast applications.

I guess back to the drawing board. What I was thinking was to use a
genny as VFO to
inject a carrier into a buffer amp driving a CW amplifier.

What is a "genny?"
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

'Genny' means "regenerative receiver circuit", typically a direct
conversion regenerative mixer/oscillator followed by an audio amplifer.
Sometimes preceded by an RF amplifier to help prevent the carrier from
radiating back into the antenna.

There's plenty of them out there but only recently did I see a few that
used tubes wired in space charge mode for 12V operation. One of the
designs slapped on a crystal oscillator as QRP transmitter to create a
full transciever, but why not simply use the regenerative oscillator as
a VFO into the transmitter?

The Eternal Squire

wrote:
I guess back to the drawing board. What I was thinking was to use a
genny as VFO to
inject a carrier into a buffer amp driving a CW amplifier.

What is a "genny?"

'Genny' means "regenerative receiver circuit", typically a direct
conversion regenerative mixer/oscillator followed by an audio amplifer.
Sometimes preceded by an RF amplifier to help prevent the carrier from
radiating back into the antenna.

Where's the PWM coming from? You could have a stock sine wave oscillator
driving a PWM modulator built as a relaxation oscillator, but somewhere you
need the modulation.

(Of course, doing this in the analogue domain kills most of the advantage,
but it still might e worth trying).

There's plenty of them out there but only recently did I see a few that
used tubes wired in space charge mode for 12V operation. One of the
designs slapped on a crystal oscillator as QRP transmitter to create a
full transciever, but why not simply use the regenerative oscillator as
a VFO into the transmitter?

Because the regenerative oscillator isn't exactly stable, to begin with.
Also, if you're running a switching output stage, you need a PWM signal
whose duty cycle in some way corresponds with the waveform you want to
produce, and that takes a bit more than just a simple free-running oscillator
(which is what a regenerative amplifier is if you crank that regen control
up too much).
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."