Tektronix 7D15 Instruction Manual page 29

standard deviation. A probability distribution graph for the
previous example, where the timeinterval is 11 ns, is shown
in Figure 2-8. Compare this graph with the probability
distribution graphs for 10.1 ns and 15 ns. The probability
range for a time interval of 10.1 ns is narrower than for a
time interval of 11 ns or 15 ns. Readings in the shaded area
of the graph represent the range of answers that may be
given 50% of the time.
Another variable that can change the shape of the
distribution curve is the number of averages taken. The
graphs shown in Figure 2-9 represent the probability curve
of an 11 ns time interval that is averaged 10, 100, and 1000
times. The graphs show that the probability of obtaining an
answer near 11 ns increases with the number of averages
taken.
It should be noted that the previous examples assume a
uniform random distribution of time coincidence. If the
input time interval and clock is synchronized an erroneous
answer may be given; see Figure 2-10. The answer does not
vary, but is wrong. Anything short of pure synchronization
is usually acceptable.
If synchronization is suspected, a check can be made by
comparing the repetition rate of the time interval to be
measured with the 7D15 clock rate. This can be done by
triggering the oscilloscope with the 7D15 PSEUDO GATE
and observing the CLOCK OUT signal. Since all the 7D15
Clock positions are synchronized with each other, for the
purpose of display, a lower clock rate position can be used.
Synchronization is indicated by a display with little or no
drift.
The amount of acceptable drift can be determinedfirst,
by calculating the time needed to make a time interval
average measurement (Teas) by the following:
+ _ Numberof averages
meas Repetition rate of measured time interval
Second, observe the waveform and measure the time of one
cycle of drift. Correct for the time interval actually used.
a
Generally, synchronization will not occur if this figure is
less than Treas.
Example: A time interval with a repetition rate of
100 kHz is being measured and averaged 1000 times, using
a clock of 10 ns.
1000
Tmeas = GookHz ~ [2m
REV. A, SEPT. 1977
Operating Instructions—7D 15
TIME INTERVAL = 10.1 ns
a" BLOCK RATE 10 t@ "i
AVERAGES = 1000
1.25
1.00
75
25
»|_lllli,
9.9 10.0 10.1 10.2 10.3 10.4 10.5 10.6 10.7 10.8 10.9 11.0
RANGE OF ANSWERS (ns)
055 TIME INTERVAL = 11 ns
.050 CLOCK RATE= 10 ns
045 AVERAGES= 1000
040
035
.030
025
.020
-015
010
005
-000
10.4 10.5 10.6 10.7 10.8 10.9 11.0 11.1 11.2 11.3 11.4 11.5
RANGE OF ANSWERS (ns)
-0350
-0325
-0300 CLOCK RATE = 10 ns
TIME INTERVAL = 15 ns
-0225
-0200
-0175
-0150
-0125
-0100
-0075
-0050
0025
-0000
144 14.5 146 14.7 14.8 14.9 15.0 15.1 15.2 15.3 15.4 15.5
RANGE OF ANSWERS (ns)
THE ABOVE EXAMPLES ASSUME A UNIFORMLY RANDOMDISTRI—
BUTION OF TIMING COINCIDENCE. qaaDt6
Fig. 2-8. Probability versus time interval.
2-9