There is a plate resistance in tubes, but it is not used in output
matching calculations for RF amplifiers. It is the dynamic plate
resistance, the resistance that the plate circuit offers to a small
increment of plate voltage (to quote Frederick Terman in Radio
Engineering). It is the reciprocal of the slope of the plate
current/voltage characteristic curve. It is NOT the ratio of plate
voltage to plate current at one point, however. According to Terman,
it depends on the dimensions and position of the cathode, grid and
plate, in a triode. Tubes with high amplification factor have a high
plate resistance, as the plate voltage has little effect on the
electrostatic field near the cathode. The plate resistance of small
beam power tubes, pentodes and tetrodes is very high, where the plate
current is substantially independent of plate voltage. Thousands of
Ohms to Megohms are possible values of Rp. When building a broadband
video amplifier (such as what USED to be in a television set), the
value of Rp works in parallel with the capacitance of the plate to
cathode, and the load resistance, to determine the upper frequency
limit of the stage.
Langford-Smith thoroughly covers Rp in circuit calculations for radio
receiver design and audio work, in Radiotron Designers Handbook.
When using large triodes with handles, the plate resistance can be
quite low, as seen from the slope of the plate voltage
characteristics. Recently I have found that even large tetrodes have
measurably low plate resistance.
For instance, a Thomson TH555A operating at about 20 Amperes of plate
current, in AB1, has about 630 Ohms for Rp. This is in shunt with the
load in calculating the output impedance of the amplifier. In some
particle accelerator applications it becomes important to know what
the particle beam sees when looking back into the amplifier. While
this Rp is not really a dissipative resistance in the tube, it is a
real parameter that can be used in the design of the amplifier. I am
speaking, of course, of class A continuous operation, for the Rp to
be a single valid number. Yes, I suppose that partial conduction
would cause a discontinuous variation in Rp.
One would not design the RF amplifier to match Rp to the load, as we
already do this with the standard RF amplifier formulations - as many
here have constantly reiterated.
> >Notice that at no time was the actual output impedance or load
> >resistance of the tube used!
>
>Huh? It is a fact that maximum power transfer occurs when a device is
>operating into a load of the same real impedance that that device has.
>Period. Now, we really don't mess with the output impedance of the tube
>because the tube is not linearly biased. It acts more as a switch and
>therefore that impedance varies over the drive cycle. This is the reason
>also that the tank circuit needs to have a fairly decent Q so that it can
>constantly deliver energy to the load over the drive cycle of the tube.
>I wouldn't want to stake my life on it, but that specified load impedance
>is probably pretty close to the average output impedance of the tube over
>the drive cycle. That's the way the laws of physics require it to be in
>order to maximize power transfer.
>
> >The fact that the tube is not "matched" to the load also explains why
> >most of the reflected
At 11:46 PM -0400 10/1/99, Amps Digest wrote:
To: <amps@contesting.com>
>Date: Fri, 1 Oct 1999 16:26:07 +0200
>From: "Maurizio Panicara" <i4jmy@iol.it>
>Subject: R: [AMPS] Conjugate matching and amplifiers
>
>I don't really agree with this statement at high power levels.
>
>An high level amplifier is not designed ONLY at a stated output power, but
>it's designed to be at the highest gain possible and for the best possible
>efficiency at such output. (Imagine a 2 MW Amp that wouldn't be designed
>under this concept....)
>
Sometimes the 2MW amplifier is designed for high linearity or
stability, and then efficiency and gain go down the drain. Our class
A 100 kw amplifier is about 20% efficient, but only operates at 5%
duty factor, so the average dissipated plate power is "reasonable"
(less than 30 kW).
John
K5PRO
--
FAQ on WWW: http://www.contesting.com/ampsfaq.html
Submissions: amps@contesting.com
Administrative requests: amps-REQUEST@contesting.com
Problems: owner-amps@contesting.com
Search: http://www.contesting.com/km9p/search.htm
|