Table of Contents
Advertisement
pected with careful adjustment i s plotted against Q in
Figure 4-4. In the f a c e of the fact that for low Q v a l u e s
the eventual lack of accuracy i s justified. For example,
if Q
=
0.03, a 5% change in inductance i s a change of
only 45 parts per million in impedance.
As far a s the user is concerned, the balancing
procedure with Orthonull i s essentially the same a s
without
it.
However, s e v e r a l suggestions for its u s e
a r e given in paragraph 2.6.
4.9 D AND Q RANGES VS FREQUENCY.
T h e D and Q readings and ranges are functions
of frequency.
Also, in order t o avoid errors in the C
and L readings, the D or Q of the unknown i s further
limited.
T h e resulting allowable D and Q ranges are
given in terms of frequency and D or
Q
of the unknown
a t the measurement frequency in Figure 4-5.
The numbers on the various limits refer t o the
explanations below:
1. End of DQ rheostat range.
2. F i r s t division on Low D (0.001) and High Q
(1000) s c a l e s (no C or L error).
3.
Limited by D of standard capacitor (no C or
L error).
4. 20-Hz limit b e c a u s e of meter response.
5. 20 kHz, a nominal limit (range narrow above
20 kHz).
6.
C or L error due t o capacitance a c r o s s stand-
ard C T and RT.
7. C or L error due t o inductance in DQ potenti-
ometer and phase of CGRL potentiometer.
8. End of the low D and high Q s c a l e s .
Use
the low Q s c a l e t o extend the low D range, and the
high D s c a l e t o extend the high Q range.
9.
Limit of 1% C and
L
accuracy, even with
Orthonull (refer t o paragraph 5.2.4).
10. C and L error may be 2% above t h i s line
owing t o inductance in the DQ potentiometer.
Note that in the overlap area either the C s or
the C bridge may be used.
Below 100 Hz is a n area
P
not covered by either bridge, requiring a n external
adjustment (refer t o paragraph 4.11).
4.10 CORRECTIONS FOR RESIDUALS.
At high frequencies, the errors resulting from the
residual bridge impedances and from the connecting
lead impedances become more important, often requir-
ing corrections. The formulas for the correction terms
are given in Table 4-1.
T h e s e correction terms are
first-order terms only.
4.11 OPERATION BELOW 1 kHz.
The wide overlap of ranges (Figure 4-5) permits
D and Q coverage down t o 1 0 0 Hz without external
adjustment. Below 1 kHz, more of the low D and high
Q range can be used than is calibrated. In this region
the low Q s c a l e c a n be used t o indicate D directly and
the high D s c a l e used t o indicate Q directly with a
maximum additional error of 2%.
Below 100 Hz there i s a D and Q range not
covered by the internal DQ adjustment.
An external
rheostat or decade box may b e used t o extend the
range of any of the D or Q s c a l e s .
However, t o avoid
error, the low D and high Q ranges should not be ex-
tended beyond a value of 1 a t frequency of measure-
ment (Figure 4-5).
The low D and low Q s c a l e s are directly pro-
portional t o frequency.
Therefore, the total D or Q
value is the sum of the d i a l reading plus the wRC
product due t o the external resistor. That is:
low D
=
(low D d i a l reading
+
0.628R x f ( k n , kHz)
low Q
=
(low Q d i a l reading
+
0.628R x f ( k n , kHz)
The low Q circuit h a s a fixed 32-ohm resistor in
s e r i e s with the potentiometer, but that is included in
the d i a l calibration.
The high D and high Q s c a l e s are inversely
proportional t o frequency, and the effects of the inter-
n a l and external r e s i s t o r s are therefore not additive.
The DQ rheostat should be s e t t o a minimum (high Q
=
a o r high D
=
50), and the whole adjustment will be
on the external r e s i s t a n c e and will be:
1.592
high Q
=
-
fR
1.592
high D
=
f (R + 0.032)
where f is in kHz and R is in k n .
4.12 OPERATION ABOVE 20 kHz.
Although the specifications for the 1650 certify
performance up t o only 20 kHz for a c measurements,
the bridge c a n be used with accuracy only somewhat
reduced u p t o 100 kHz. At frequencies above 20 kHz,
limits other than those shown in Figure 4-5 restrict
-
the accuracy attainable with the bridge.
T h e s e limits
can be stated a s a percent error, which should be
added t o the b a s i c one-percent accuracy given in the
instrument specifications.
The added error introduced
above 20 kHz is always negative, and the net effect of
the two errors will probably be negative.
Table 4-2
shows C
-
L, accuracy a t CGRL d i a l settings between
0.4 and
2
T h e average of the net accuracy limits i s -0.5%
a t 5 0 kHz, -1.25% a t 100 kHz. If this amount is added
t o the measured value, the accuracy c a n be stated sym-
metrically a s f 1.5% a t 5 0 kHz and +2.25% a t 100 kHz.
ACCURACY
4-5
Advertisement
Table of Contents
