Table of Contents
Advertisement
And since we set
T
0
~
2 T
A
v
p
B
3 75.
10-
3 -
,.
TAv
So for a particular bandwidth of analysis, the optimum Level Recorder PAPER SPEED is
given in Table 4.3. The DRIVE SHAFT SPEED equivalent to a particular PAPER SPEED is
also given in Table 4.3. If the frequency scale is to be expanded by using DRIVE SHAFT
II of the Level Recorder and a higher PAPER SPEED, it is the DRIVE SHAFT SPEED
which must be maintained as given in Table 4.3 since this will determine the Analyzer
sweep rate.
If a FREQUENCY SCALE other than
"x
1 Lin." is used, the Table 4.3 will not apply. In
such cases, it is necessary to recalculate PAPER SPEED starting from equation
7
and ap-
plying an appropriate factor of 1 0.
Bandwidth
Averaging Time
Paper Speed
Drive Shaft Speed
(Hz)
(s)
(mm/s)
(rpm)
3,16
10
0,001
0,012
10
3
0,01
0,12
31,6
1
0,1
1,2
100
0,3
1,0
12
316
0,1
10,0
120
1000
0,1
10,0
120
740198
Table 4.3. PAPER SPEED settings and AVERAGING TIMES for AC or
DC Recording of Random Signals
e
<± 1 0%, Linear Sweep
4.4. COMPLEX SINUSOIDAL SIGNAL ANALYSIS
48
For the purpose of this section, complex sinusoidal signals are considered as being those
composed of discrete sinusoidal components, e.g. rectangular waves. The analysis rate
is restricted by the consideration that enough time must be allowed for the filter in use
to respond fully to each signal component. A worst case can be defined in which the Le-
vel Recorder pen travels from zero deflection to full scale deflection (taken as being
50 dB) to give a level reading within 0,5 dB of the correct level and return to zero. A sui-
table relationship for the filter response time TF for B & K filter types has been found
from practical tests to be
2
B
(9)
A second parameter of interest is the averaging time, T
AV
.
If too short an averaging
time is allowed, the ripple on the resulting signal will be large, resulting in the Level Re-
corder pen tracing out a wide recording, making accurate reading of the analysis diffi-
cult. To restrict the ripple, the following minimum averaging timeT
AV
may be defined as
:
for
<
±0,25 dB ripple
(1 0)
In the case of Constant Bandwidth analysis and AC recording, the filter response time
dominates. For DC recording, the same situation occurs for 3,1 6Hz and 10Hz band-
width, but because of the minimum AVERAGING TIME setting of "0, 1 s", for 31,6 Hz
and wider bandwidths, the averaging time always dominates.
Advertisement
Table of Contents
