Re the Tek 585 and 585A.

The vertical sections of these scopes are more-or-less "one of a kind"
and they represent an end point in design in several areas.  They are
the last high-performance all-tube scopes that Tek made, and one has
to consider both the CRT and the vertical amplifier separately.

So far as the CRT goes, it is typical of what was used in the first 15
years after WWII, except for the distributed vertical deflection.
Essentially, this means a Wehnelt gun, a collimating focus structure,
deflection plates, and post-deflection acceleration (PDA).  The
distributed vertical deflection is laid out such that the vertical
wave propagates down the deflection plate structure at the same speed
the electron stream passes through it.  It is a lumped-constant delay
line.  There are coils in the CRT between the plate segments, and the
plate segments are constant capacitances.  This type of delay line has
a very definite cutoff frequency.  

A problem with post-deflection acceleration is that the PDA field is
distorted by the deflection plates---higher voltage centered between
the four plates than at the plates, which makes it a converging lens.
This decreases deflection sentsitivity.  The reason for using PDA is
plain vanilla quantum fizzicks---more electron volts in the beam where
it hits the phosphor, more phosphor activity, which improves the
writing rate.  In this answer, I'll ignore the effective potential of
the phosphor screen, although controlling that is significant when
designing CRT's.  

There are several gimmicks that can be used to control the lens shape
of the PDA field.  HP, Tek, and Dumont all used a flat Faraday shield
screen in front of the deflection plates through the early sixties.
This reverses the lens from converging to diverging---at the price of
widening the spot, problems with the shield wires intercepting the
beam at places, and deflection linearity.  The 547 was the first Tek
design to use such a Faraday shield, and the effect on deflection
sensitivity was dramatic---I've forgotten the numbers, but it is
something like 10 volts/CM instead of 30 in the vertical plane.  The
585's never had such a tube.  Peter Keller's book on CRT's does talk
about other methods for improving sensitivity and so forth in the
post-deflection electric fields.  I don't know of a "comprehensive"
text on all this----most of what was published in the forties ends
with the type of CRT used in the 530/540 scopes, and I felt Keller was
a bit obscure for the casual reader who isn't steeped in New Clear
Fizzicks concepts.  Modern methods are to slow the signal down with
digital methodologies (the early sixties sampling scopes are
essentially digital, although they don't convert the signal to digital
data).  I'll ignore the Tek 519 here---that was pretty much a
laboratory exercise with no vertical amplifier, and most certainly not
a general purpose bench scope.  But it did use CRT technology similar
to the 585.  

The vertical amplifier in the 585 is a fairly conventional distributed
amplifier---principals of operation are not much different than the
original 514 and 545's.  However, the amplifier is double-terminated,
not just source-terminated as in the 545.  This cuts deflection
sensitivity in half.  When looking at tubes, one has to consider the
transit time through a vacuum tubes, and it is mighty long compared
with transit times through the base region of modern semiconductors.
You have effects that are somewhat analogous to transistor stored
charge and so forth, but you have to pay attention to things like
virtual cathode points in the tube, as well as the transit time across
distances on the order of a millimeter between grid and plate.  I'm
oversimplifying a bit here (go read Spangenberg and some of the other
later texts) but it boils out that it is difficult to get a tube
structure small enough, and with enough electron density to give you a
gain-bandwidth product much more than 200 Mhz (I think this is the
number for the 955 acorn---a product of 1933 technology).  Nuvistors
and frame grid toobs of the sixties did a little better, but they
don't compare with a 600 Mhz. transistor Ft (those little dimey jobs
you can buy anywhere) by a long shot.   The crossover in
gain-bandwidth performance moved to transistors with etched "mesa" and
similar technologies in the early sixties.  Once the manufacturers
learned how to get yield out of angstrom-sized structures (10E-10
meter, a handy size nobody talks about any more), fundamental physics
said that they were going to give peformance---if you could live
within the voltage and power constraints that come with teeny tiny and
semiconductor lattices.  

Historically, the 585 was announced as a "100 Megacycle" scope in 1959,
using an 80-P80 plugin.  This is a cathode follower
gets-hot-in-your-hand job that isn't what most people consider
"general purpose."  As time went by (and a lot of time went by) Tek
managed to come up with a general purpose plug-in (the 82) and probe
that just plain weren't going to deliver enough bandwith (about 150
Mhz needed) at the vertical amplifier connector to give "a hunnert
motorcyles" (as the argot went at Tek when this got discussed) at the
CRT.  At the same time, CRT yield, production tweaking, etc. pointed
to backing off a bit from the original target.  By 1963, Tek's razzle
dazzle backed off to 85 Mhz for the design.  Most 585's will beat that
handily.  The general philosophy behind backing off was more to
control customer expectations.  The 547 was presented entirely
differently---50 Mhz at the probe tip, not 75 Mhz with a specialty
plug-in, degrading the more general-purpose you make it.  

The real differences between the original 585 and the 585A are not in
the vertical section, other than making the delay line termination
tweakable.  The big changes were the addition of a tunnel diode
trigger setup that would work out to some reasonable frequency, and
incorporation of a bunch of modification packages for the 585 that
were suggested/recommended/mandatory for using the 82 and keeping the
scope running for a while.  In short, all the backing off from "100
Megacycles" was more like backing off from test stand horsepower to
horsepower delivered to the tires.  

Henry van Cleef