Table of Contents
Advertisement
The residual resistance of 1 mR i s that of the
binding p o s t s themselves.
For low-resistance meas-
urements, short, heavy leads should be used a s con-
nections t o the unknown.
The zero r e s i s t a n c e of the
leads should be measured with the free ends connected
together, and subtracted from the bridge reading with
the unknown in place. The user should be particularly
careful when using banana-pin connections.
For b e s t
connection t o the bridge, tighten the binding post hard
enough t o notch the wire inserted in the hole.
Since there i s no internal Q adjustment on the R
bridge, reactance a f f e c t s only the ability t o get a good
sharp null. If the reactance is large enough t o prevent
a satisfactory balance, an external capacitor may be
used t o make a reactance balance (paragraph 3.5).
4.4
INDUCTANCE.
The accuracy of the L reading is
il%
if the
balance is made between 1 and 11 on the CGRL dial.
Below 1 on the d i a l the accuracy is
division. Thus
the over-all accuracy i s A% or
d pH,
whichever is
greater, since 1 p H is
'/:
d i a l division on the lowest
range.
T h e Q accuracy i s given in terms of D = 1/Q
and is
4%
or fl.001, whichever is greater, with a
CGRL reading of 1 or higher.
The residual (zero) inductance is l e s s than 0.2
pH, which i s l e s s than the accuracy of the bridge and
therefore negligible.
If external leads are used t o
connect t o the unknown, t h i s zero inductance i s in-
creased and should be subtracted from the bridge
reading.
The residual resistance of the bridge is 1 milli-
ohm, which c a u s e s a small D (1/Q) error.
T h i s error
is l e s s than 0.001 if L x i s more than 160 pH.
If long
leads are used t o connect t o the unknown, t h i s error
can become appreciably greater and require a correc-
tion. The D error is
R
R o
(the Q error i s Q ~ L
)
t-
wLx
wLx
where Ro is the total lead resistance.
The residual zero capacitance of 0.5 p F theoreti-
cally c a u s e s an error for inductors above 250 H.
However, t h i s small capacitance i s almost always
negligible compared with the capacitance of the wind-
ingof s u c h a large inductor. If the inductor is shielded,
a three-terminal measurement will reduce the effect of
stray capacitance to the shield (paragraph
3.8).
In
order to reduce the effect of the winding capacitance
it i s necessary t o reduce the measurement frequency.
The inductance error due to a shunt capacitance C o
i s w 2 ~ o ~ x 2 ,
and t h i s amount should be subtracted
from the bridge reading (paragraph 4.10).
T h e inductance accuracy is reduced slightly if
Q is l e s s than 0.1.
However, even with Orthonull,
balance t o 1% is impossible.
Errors at other frequen-
c i e s are d i s c u s s e d in paragraphs 4.11 and 4.12.
4.5 CAPACITANCE.
The accuracy of the C reading i s ?1% if the
balance is made between 1 and 11 on the CGRL dial.
Below 1 on the d i a l the accuracy i s
division. Thus
the over-all accuracy possible is +1% or *1 p F , which-
ever is greater, since 1 p F i s
%
a d i a l division on the
lowest range. The D accuracy is &5% or Kl.001, which-
ever i s greater, with a CGRL d i a l reading of 1 or
higher.
The residual ("zero")
capacitance of the bridge
terminals i s approximately
'/:
p F , which is l e s s than
the accuracy of the bridge and, therefore, negligible.
If external leads are used t o connect the unknown, this
zero capacitance i s increased and should be subtracted
from the bridge reading.
The residual resistance of the bridge i s 1 m a ,
which theoretically c a u s e s a D error of 0.006 when
C x
=
1000 p F .
In practice, capacitors of this s i z e
have such large D values that such an error i s negli-
gible. However, if leads are used t o connect large ca-
pacitors, t h i s D error can become important and a cor-
rection should be made. The D error is +w%
$
(where
R o is the lead resistance), and this amount should be
subtracted from the D reading.
The residual inductance c a u s e s negligible error
a t 1 kHz even if C x
=
1000 p F . However, connecting
leads could have enough inductance t o c a u s e a C
error when large capacitors are measured.
The error
is +wLoCx (when L o i s the lead inductance) and t h i s
amount should be subtracted from the C reading.
The capacitance accuracy is reduced on the C p
bridge when D becomes larger than 10. However, even
with the Orthonull balancing mechanism, balance t o
1% precision i s impossible; thus this error i s negli-
gible (paragraphs 4.8 and 5.2.4).
Errors for capacitance measurements a t other
frequencies are d i s c u s s e d in paragraphs 4.11 and 4.12.
4.6 EFFECTS OF CAPACITANCE TO GROUND.
The Type 1650 Bridge generally measures "un-
grounded" components, since neither UNKNOWN ter-
minal is connected directly t o the panel.
The panel
should be connected t o a good ground, especially i f
high-impedance components are t o be measured.
If
the panel is not grounded, stray capacitances from the
UNKNOWN terminals and panel to ground c a n produce
an effective capacitance across the UNKNOWN ter-
minals.
With the panel grounded, capacitances from
the UNKNOWN terminals t o ground have a much l e s s
ACCURACY
4-2
Advertisement
Table of Contents
