Aloha to all on the reflector interested in the discussion
of the conjugate match, amps, and antenna tuners.
And to Walt, to whom I send a copy of this message, and
to Dr. Jerry, whose thoughtful comments continue to
stimulate this discussion!
I recv'd the following from Walt Maxwell tonight, it required
a couple of days for him to respond. He is responding directly
to Dr. Jerry, K0CQ
"----- Original Message -----
From: "Walter Maxwell" <w2du@iag.net>
To: "Jim Reid" <kh7m@hsa-kauai.net>
Sent: Saturday, January 27, 2001 5:49 PM
Subject: Re: R sub p and R sub s
----- Original Message -----
From: Jim Reid <kh7m@hsa-kauai.net>
To: Walter Maxwell <w2du@iag.net>
Sent: Thursday, January 25, 2001 12:50 AM
Subject: R sub p and R sub s
> Aloha Walt,
>
> Have been in a discussion on one of the Internet reflectors
> about the subject of the conjugate match and active amplifiers.
> Felt I was defending you position, but have run into the following:
>
> "These two explanations are antagonistic. Maxwell ascribes a value, Rp as
> a characteristic of the tube when the plate resistance varies over the
> cycle from far lower than his Rp to full open (unless class A, then the
> plate resistance of the tube only rises quite high, but not open
> circuit). Bruene fought with Eimac all his working life about that Rp
> being a none existent quantity.
Walt Maxwell:
Jim, perhaps it's due to the manner of my wording, but for whatever reason,
some of the things I'm getting as feedback are not what I intended to
convey. The standard definition of Rp is the non-dissipative resistance
resulting from dividing the change in plate voltage by the change in plate
current resulting from the change in plate voltage. I don't remember which
side Bruene and Eimac were on, but it is certainly a non-existent quantity
that defies measurement. I don't remember ascribing it as Jerry says, but
Bruene insists that Rp, which he calls 'Rs', is the source resistance of the
tube. I disagree, because the change in plate voltage resulting from a
change in grid voltage derives the source of power and the change in plate
voltage that effects the Rp is detrimental to the power source. Its action
is the same as negative feedback.
>
> I said, while its not a quantity you can measure with a bridge, its a
> handy artifice for designing the output network impedance
> transformation. Rp has no physical realization in the active device,
> whether gridded vacuum tube, bipolar transistor or FET. All these
> operate as time variable resistors connecting portions of the supply
> voltage to the output network. The output network exists to provide a
> suitable load line for that variable resistor. That load line IS
> constant over the cycle. The active device output resistance is NOT.
Walt Maxwell:
I agree exactly. The reason the load line is constant over the cycle is
because it is fixed by the adjustment of the pi-network tank, and the energy
storage effect of the tank is what keeps it constant.
> Maxwell wants to say that the active device is matched so the reflected
> wave isn't reflected again. I don't see it that way because the output
> resistance of the device varies wildly over the RF cycle. What I see is
> that the tuner or Pi network is adjusted to accept the feed line input
> impedance, whatever magnitude and phase angle, and to transform it to
> the proper load line for the active device. In that process the network
> has to absorb stray reactances of the active device and its connections.
>
> 73, Jerry, K0CQ"
Walt Maxwell:
I don't remember ever saying that the active device is matched so the
reflected wave isn't reflected again. If the active device is matched to its
load it means that the reflected wave would be absorbed in the internal
resistance of the active device. This not what happens.
If there is no matching device between the source and the mismatched load
the resulting reflected wave becomes incident on the source. In this case
the source simply reduces its delivery of power by the amount of the power
in the reflected wave, and the power in the reflected wave then adds to that
of the reduced power now being delivered by the source. The result is that
the same amount of power enters the line as that which entered before any
reflected wave returned to the source.
However, I agree with the last sentence in Jerry's paragraph above, because
the matching device re-reflects the reflected waves, preventing them from
reaching the source.
>
> The above was in response to the following:
>
> "Dr. Jerry wrote, in part:
>
> > There is no conjugate match to the output of any active
> > source.....
Walt Maxwell:
This is true unless the source is a Class A amplifier in which the plate
current is continuous.
> > All those that demand conjugate matching through to the
> > active device are wrong.
Walt Maxwell:
True in cases of Class B and C RF amplifiers.
> > Its given the more maddening appearance of there being
> > a conjugate match because of the way RF solid state devices
> > are measured. Commonly they are installed in a test jig, and
> > the universal matching network is adjusted for maximum power
> > output (a side effect on conjugate matching of no time varying
> > components) and then the test jig is split and the Z shown to
> > the active device is measured and its conjugate declared the
> > output Z of the device. It is NOT. The impedance measured is
> > the load the allows the active device to produce maximum
> > power (within distortion and efficiency and gain limits). For circuit
> > designers, it is convenient to say the output impedance of
> > the device is r + jx and the load then must r - jx to be a
> > conjugate match but that's an improper statement of fact.
> > Its plain wrong, but it makes circuit designers produce output
> > networks that supply the appropriate load for the device.
> >
Walt Maxwell:
I don't know how solid-state devices are tested. However, one must specify
whether he's talking about the output impedance of the active device or the
output impedance of the output network that couples the device to its load.
If the device is delivering all of its available power to its load the
output network is conjugately matched to the load, but the device itself is
not.
> > During any one cycle of RF the output Z of the active device
> > varies from open to nearly shorted and that's nothing that
> > can be matched by a constant impedance. Any time the
> > active device is a constant impedance, its not producing any
> > output, its output Z MUST VARY to produce output.
> >
> > But if the load Z is set for the available voltage swing to produce
> > the desired output power, the power supply is saved and the
> > power output is controlled.
>
> Now, if only Bruene had been able to state what Dr. Jerry has
> given us in the above so clearly....... Instead, Bruene set about
> to annihilate Walt Maxwell's work back in 1991, in his QST article:
>
> "The RF Power Amplifier and the Conjugate Match",
> W. Bruene, QST, November, 1991.
>
> So, Dr. Jerry has expressed his and Bruene's view. Now here
> is the explanation by Walt Maxwell:
>
Walt Maxwell:
>From the sentence above am I to understand that Jerry's view is that of
Bruene's? I should hope not!
> " I realized that much more measured data was needed to prove
> that output resistance of the pi-network of the amplifier is exactly
> equal to the load impedance after the network is adjusted to
> deliver all the available power. Using a standard IEEE procedure
> I took many more measurements that indeed prove conclusively
> that the network output resistance and the load resistance are
> equal when the network is adjusted as stated above"
>
> Sounds very similar to what Dr. Jerry describes above. I believe
> the problem between these men is one of the use of words and
> language to describe what they are doing/talking about. However,
> Maxwell has attempted to again describe his non-dissipative
> resistance in his claim of an actual conjugate match:
> So, are these two differing "explanations" antagonistic or
> complementary? To me, both make sense.
>
> 73, Jim, KH7M"
>
> I hope I quoted the correct material from your writings, Walt!
>
Walt Maxwell:
Your quotations were fine, but if you quote my RF amp article to those on
the Reflector it would help if you also quoted the portions concerning the
storage effect of the tank circuit. This is the crux of the problem, because
that storage effect is what permits a conjugate match at the output while no
conjugate match can ever exist at the input of the network when the source
enters in pulses.
I hope my comments are helpful, and if they disagree with what has been said
earlier I would like for you to relay them to those who disagree.
Jim, could you tell me which Reflector this discussion is in? I'd like to
read the mail.
73, Walt"
Aloha Walt, and thank you for your response. The reflector
is the Ten Tec reflector at contesting.com. You can see the
address above. We are all looking forward to your coming
new edition of Reflections, 2nd Ed.!!
73, Jim, KH7M
On the Garden Island of Kauai
--
FAQ on WWW: http://www.contesting.com/FAQ/tentec
Submissions: tentec@contesting.com
Administrative requests: tentec-REQUEST@contesting.com
Problems: owner-tentec@contesting.com
|