Table of Contents
Advertisement
INTRODUCTION
of 0.105% at full scale and 0.15% at one-tenth of full
scale. Therefore, the final balance should be made with
as many digits on the counter as possible.
1.5.2
TEMPERATURE COEFFICIENT.
The over-all
temperature coefficient of the instrument is less than
30
ppm/°c.
This means that there may be a 0.03%
change in reading for a temperature change of
lQoe
(50 OF). This change is usually negligible compared with
the change in the unknown component for a similar tem-
perature change.
For the most accurate measurements,
the bridge 'and components to be measured should be
stabilized at a temperature near 23
°e
(73 of).
1.5.3
ADDITIONAL ERRORS FOR HIGH D CAPACI-
TORS/ LOW
Q
INDUCTORS/ AND HIGH
Q
RESISTORS.
The DQ dial adjustments used for phase balance on the
e
and L bridges are wire-wound rheostats.
When lossy
(high D or low Q) components are measured, the limited
resolution of these adjustments prohibits balance of the
e
or L adjustment to its full resolution.
A term of
0.5%D 2 is added to the specifications to allow for this
effect, but somewhat better accuracy is possible with
extreme care.
Precision components generally have a
low enough D or a high enough
Q
to make this term neg-
ligible (see Figure 1-5).
kc
Figure 1-6; Resistance and conduct-
ance errors vs frequency.
FREQUENCY
J
ff
~'I
/
J
V
1;/
VI/
I
j;
/
TOTAL ERROR
a·'
~
- '
~
r
IbOOI(~r
~OTAL
ERloR Q·O
~
VI
.1
0
1
1%/
V
/
1
12
-1 /
0002(~)
1/
1/
I
Ike
10ke
100
o
0.0
100
10
The Q adjustment for the R s and Gp bridges con-
sists of two decades of mica capacitors and a variable
capacitor with infinite resolution.
Losses in the mica
capacitors appear as an R or G error when Q is rela-
tively large, and the added error term of O.l%Q is there-
fore
necessa~y
(see Figure 1-6).
100
1.5.4 FREQUENCY ERRORS.
The main cause of addi-
tional error on the
e
and L bridges at higher frequencies
is the inductance of the pridge wiring in series with the
standard capacitor, effectively increasing its value .
This error is proportional to f2, and is accounted for in
the added error term 0.001% (fl1 kc)2. This term, which
amounts
to
a 0.1% error at 10 kc and a 0.4% error at 20
kc, is large enough to account for other smaller sources
of error (see Figure 1-5).
F or high D (low Q) measurements at high frequen-
cies, there is an added error term due to the inductance
of the DQ rheostats.
This term is O.lD (f/1 kc).
(See
Figure 2-7.)
The series rheostat (e s and L p bridges)
is phase-compensated to a large degree, but neverthe-
100ke
10ke
FREOUENCY
Ike
Figure 1-5. Capacitance and induct-
ance errors vs frequency.
/
VJ
/
VA
W
TOTAL ERROR 0·1
"-..
~
V
/
(/II
I
~
///
(j
l.---
V
~
~
'0
-
0'0.2
. /
~
/
I
.1
TOTAL ERROR 0 . 6 /
I
1{~0IC~.)2
I
I
o
0.0
100
10
5
Advertisement
Table of Contents
