10
How oscilloscopes work
(2): circuitry
Figure 10.1 shows the block diagram of a typical dual trace, high-
performance oscilloscope. Two identical input channels A and B
are switched alternately to a common amplifier, which drives a
delay line. This is shown diagrammatically as composed of
discrete inductors and capacitors, although in a modern instru-
ment it would usually consist of a length of delay cable. This is
similar to coaxial cable, except that it has a centre conductor
wound in the form of a spiral and hence provides much greater
delay per unit length. As the drive to the trigger circuit is picked
off before the delay line, the delay introduced by the latter
permits the whole of the leading edge from which the scan was
Y
go,n
sh~ft control
TT
final Y
l
tr,gger
rt
c,rcu,t
~ [
1
A A
and B and
chonnel
ext
~ shop,ng I _ O+
om p network I"o~ ~
_. I.y-y- - J'l.
free running 9
]
t~mebase
t r, ggered
9
},
preomp
eHt -1
power /
supplyj
I molns
J~.
input
L
-I '~ J
";" ~"
r
_ 9 ~ e.e~pon ded
sh=ft normol
Figure 10.1 Block diagram of dual trace, mains-operated oscilloscope (courtesy
Enertec Instrumentation Ltd)
How oscilloscopes work (2): circuitry 189
triggered to be observed. This assumes of course that the risetime
of the leading edge and the 'wake-up time' of the trigger circuit
are together less than the delay introduced by the delay line,
which is generally tens of nanoseconds.
The final Y amplifier produces the push-pull voltages that
drive the Y plates, and in a higher-performance instrument the
peak-to-peak output swings required might be little more than a
few tens of volts or less, especially if using a tube with a high
p.d.a, ratio and a scan-expansion lens. The X amplifier has to
provide several times as much voltage swing as the Y amplifier, as
the X-plate sensitivity is less than that of the Y plates. Fortunately,
a substantially smaller bandwidth suffices for the X amplifier,
easing the circuit design problems: the c.r.t, designer takes
advantage of this to maximize the Y-plate sensitivity at the
expense of the X-plate sensitivity.
The X deflection amplifier is driven with a sawtooth waveform
produced by a 'sweep' or 'timebase' generator, which itself is
triggered by a pulse from the trigger circuit. The trigger circuit
produces a pulse each time the Y input voltage crosses a given
threshold voltage, which is usually adjustable by the front-panel
trigger level control. Thus the sweep always starts at the same
point on the waveform, the sweep generator thereafter being
insensitive to further trigger pulses until it has completed both
the trace and the following (blanked) 'retrace' or 'flyback'.
Circuit elements
Traditionally, oscilloscope designers made use mainly of discrete
components, especially in critical stages such as the Y amplifier
output stage driving the c.r.t, deflector plates. However, integrated
circuits are being used to an increasing degree, especially in high-
performance oscilloscopes, and this trend will doubtless continue
and accelerate. Few if any integrated circuits are produced by the
major semiconductor manufacturers specifically for oscilloscopes
in the way that i.c.s are mass produced specially for TV sets. The
largest oscilloscope manufacturers have their own in-house i.c.
facilities, often producing i.c.s in hybrid form, since in scope
applications one is always seeking to wring the last ounce of
performance out of every circuit. The same consideration is likely
190 Oscilloscopes
to ensure that certain sections of oscilloscopes will continue to be
designed using mainly discrete components.
Figure 10.2 shows two of the basic circuit 'building blocks' used
in oscilloscopes. The long-tailed pair is widely used in both forms
shown, the second being especially common in analogue inte-
grated circuits. It provides balanced push-pull outputs, even if
only one input terminal is driven; i.e. it converts from unbalanced
R L R L
~JTR 3
__ __j,_
(a)
k IRL
TR2
T R1
w
(b)
Figure 10.2 Basic" circuit 'building blocks' commonly used in oscilloscopes: (a)
long-tailed pairs, (b) casc()de circuit
How oscilloscopes work (2): circuitry I91
to balanced signals. This is an important function, as oscilloscope
inputs are usually 'single-ended' or unbalanced, whereas a push-
pull or balanced drive is almost invariably applied to the Y (and X)
plates. The reason for this is simple. If balanced drive is used, only
half the peak-to-peak voltage swing is required at each plate
compared to the swing required for the case where only one plate
is driven, the other remaining at a constant potential. Thus with
balanced drive the supply voltage to the transistors driving the
plates can be halved. With only half the voltage across each
transistor, the current through it can be doubled without
increasing its heat dissipation, which is important in the output
stage of a deflection amplifier, as these transistors are invariably
run very near the maximum permitted dissipation. With half the
supply voltage and twice the current, the load resistor
R E
will only
be one-quarter of what it would have been for single-ended
deflection, resulting in a fourfold increase in bandwidth.
The cascode circuit- Figure 10.2(b) -can be seen to consist
of a common-emitter stage with a common-base stage as its
collector load. This arrangement has two advantages. First, the
maximum voltage that can be applied to TR2's collector is equal
to the collector-base breakdown voltage Vcb, which for high-
frequency transistors is often substantially higher than the
common-emitter breakdown voltage Vce, enabling a larger
output voltage swing to be obtained from the stage. Second,
there is inevitably, owing to the construction of a transistor, a
capacitance of a few picofarads between its collector and base
terminals, denoted Ccb. In the cascode circuit, the input capaci-
tance at the base of TR1 is approximately
Ccb ~ + Cbe ~
(where Cbe~
is the base-emitter capacitance of TR1), since the input imped-
ance at the emitter of grounded-base stage TR2 is very low and
there is therefore negligible signal voltage at TR1 collector. If a
simple common-emitter stage were used in place of the cascode
stage, the input capacitance would appear much larger, as the
end of Ccb connected to the output would be changing in the
opposite sense to the input voltage, by an amount greater than
the input voltage swing. In fact, if the stage gain is A, the input
capacitance would be approximately Cbe + (A + 1 )Ccb, the well-
known Miller effect. If A is large it would prove difficult to drive
192 Oscilloscopes
the stage satisfactorily, a problem that is avoided by the cascode
circuit.
Y deflection amplifier
Oscilloscope designers frequently make use of the advantages of
both the long-tailed pair and the cascode, as shown in Figure
10.3. Here, the total output capacitance Ct shunting
R E
is equal to
Ccb2 plus the load capacitance, several picofarads if this is a
deflection plate of a cathode ray tube. If both transistors have
high cut-off frequencies, the-3dB bandwidth (70.7 per cent
response) of the stage is given
byf_3d B --
1/2~RLCt,
showing that
for maximum bandwidth both RL and Ct should be as small as
possible. There is little the oscilloscope designer can do about the
plate capacitance of the c.r.t., other than find another tube with
the same sensitivity and lower plate capacitance if possible, but
4- .4.
R L
Y
plotes
RC
TR2 TR2'
Rl
TR
R3
Figure 10.3 Basic deflection-amplifier circuit
c.r,t
..................Content has been hidden....................
You can't read the all page of ebook, please click here login for view all page.