There's been a lot of discussion covering some of these topics between my
writing this and posting it. Sorry if I'm out of sync. anywhere.
A few observations, not specifically related. I'm not trying to support or
attack anyone's ideas - I'm thinking out loud, trying to understand.
As I recall, 'Care and Feeding' gives a formula that relates Ia to (Vg +
Vs/us + Va/ua)^3/2. I don't remember if this is for dc values or incremental
changes. Eimac say that we can ignore the Va term. We know that keeing Vs
constant makes a linear amplifier, but Ia is then related to Vg^3/2. If this
is linear, then Ia related to Vs^3/2 could be linear too. I guess the
relative values of Vg and Vs/us matter. Any mathematician out there who can
expand (Vg + Vs/us)^3/2 ?
I put Ia and Vg values for 3 tubes into a spreadsheet and tried to fit the
3/2 law to them. I got mixed results, maybe because I lack the knowledge and
tools. For 4CX350, most of the curve came out around 1.35-1.4, for QY4-500
it was 1.65-1.7 and for 4CX800 it was a pretty solid square law. For the
latter two, this included significant positive grid voltage provided you
stayed clear of the low voltage end of the curves. For both of them, the
constant current curves deviate from straight and parallel at some Va value
whether Vg is positive or negative. It looks as if you could load these
tubes to be linear while running significant grid current. The driver would
have to cope, and the efficiency would be poor.
Close in IMD is caused by odd order distortion an an amplifier
characteristic. Even order (eg square law) produces IMD at harmonic
frequencies, which get filtered out.
The Ia/Vs measurements (Vg fixed at zero) at the beginning of G2DAF's
article show a straight line relationship over most of the graph, certainly
not 3/2 law.
Looking at the data published for the QBL5-3500 'DAF' amplifier, the
input/output graph shows excellent linearity over the measured power range -
I don't remember any sign of compression. The relationship between screen
voltage and rf voltage is anything but linear. I guess that the screen
current loads the doubler and the voltage sags. Again, I guess that it sags
so as to mirror any non-linearity that arises from positive grid voltage.
Whether this is by physics, design or fluke, I don't know. There might be
some form of self limiting, or negative feedback effect if the scrren
voltage sagging is due to loading on the screen voltage rectifier.
Steve
--
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
|