2-3
IM 704510-01E
Explanation of Functions
3
2
1
4
5
6
7
8
9
10
11
12
13
14
15
The binarized signal is selected as a measurement signal according to the measurement
function by the signal multiplexer 1 and 2. In addition, the fractional pulse generator 1
and 2 generate fractional parts of the measurement signal with respect to the
measurement clock as fractional pulses. The fractional pulses are passed to the time-to-
voltage converter (T/V converter). The time of the fractional pulse is measured in 25-ps
resolution by the T/V converter, then the fractional pulse time is sent to the data control
gate array (G/A) where measurement data is created.
As shown in the diagram, two identical circuits are provided after the signal multiplexer.
The data is generated alternately by the two circuit systems. When making dual
measurements, each system performs measurements independently. The acquisition
controller controls the overall acquisition process. Acquisition controller 1 also controls
the external arming signal or inhibit signal.
In the time stamp mode and inter-symbol interference analysis mode, both the measured
values and time stamp data (elapsed time) are acquired in the acquisition memory. In
the hardware histogram mode, only the frequencies of occurrence of each measured
value are acquired in the acquisition memory.
The retrieved data is read by the CPU where it is used as statistical calculation data or
displayed on the LCD. In addition, the measured results can be printed on the built-in
printer or saved to a floppy disk or PC card (optional).
Either the signal from the internal crystal oscillator (compensated against temperature
drift) or an external input reference signal (signal from the REFERENCE IN terminal) can
be used as the reference clock. The measurement clock uses the frequency multiples of
this reference clock. In either case, the signal passed through a 10-MHz bandpass filter
is output externally as a 10-MHz signal (10MHz OUT). The gate output (GATE OUT)
terminal outputs binary signals indicating the measurement interval (time over which the
signal is being acquired).
The TA720 can be controlled using a PC via the GP-IB or Ethernet (optional) interface.
Measurement Principle (Pulse Width Measurement Example)
Time shorter than the period of the reference clock is called fractional time. In general, since the
signal being measured and the measurement clock are not synchronized, fractional time exists at
both the beginning and the end of measurements. This TA720 generates a “fractional pulse”
which is a pulse signal amounting to the sum of the fractional time and one cycle of the reference
clock. If the pulse width of the signal being measured, the period of the measurement clock, and
the times of the fractional pulses are expressed as T, t
0
, T
a
, and T
b
, respectively, pulse width T
can be broken into two terms: integer multiple of the measurement clock, N•t
0
, and the time of the
fractional pulses, T
a
, T
b
(see the equation below).
T = N·t
0
+ (T
a
–T
b
)
The TA720 converts the time (T
a
, T
b
) of the fractional pulse that it generated at the
beginning and end of the measurement to voltage values. The voltage values are then
converted to digital values using a 7-bit A/D converter, measuring the fractional pulse
time at 25-ps time resolution per LSB. Pulse width T is derived by substituting the
measured fractional pulse time into T
a
and T
b
in the above equation.
T
1
2
t0
N
Ta
Tb
V =k·T
V =k·T
a
a
b b
T =N·t + (T –T )
b
a
0
Signal to be
measured
Measurement clock
Fractional pulse
Time-voltage
conversion
A/D conversion
A/D conversion
k: Coefficient used in the
A/D conversion
2.2 System Configuration, Block Diagram, and Principles of Pulse Width Measurement
