249.6 and 238.3 pF for the nulled cable circuits.
sured using water as the reference material and
This overestimated capacitance for the distilled
Cwater, divided by the dielectric constant of the
water led to a high Cair, which in turn caused the
water, Kwater, at the measurement temperature.
calculated values for Kmethanol to be too low. The
Note also that the best accuracy would be ex-
conclusion is that the cable should be nulled to
pected when the cable has been nulled.
obtain best accuracy in the dielectric constant mea-
Using the above measurement sequence, a set
surements.
of 23 test probes was fabricated and used to mea-
sure the dielectric constant of methanol (K = 32.65).
Insulated probe considerations
A data set was taken for both probe circuits in
The measurement process described above may
boxes 2, 3, and 4, leaving box 3 unnulled. Probes
also be used for insulated tines. For example, the
M and U were insulated. Observations with each
probes may be calibrated and used in the same
probe were made in distilled water and in metha-
way as the uninsulated probes, or the water mea-
nol. The raw data and the capacitance and dielec-
surements may be used to determine a series ca-
tric constant calculations are shown in Appendix
pacitance that may then be used to correct the
B. A sample calculation is given below:
insulated probe readings so that they agree with
Box 2: Average capacitance for the distilled
the uninsulated probes. Both procedures are dis-
water (Cwater) is 249.6 pF; the dielectric constant
cussed here. It should be noted that for insulated
for water at 21.6C is 79.78:
tines the resolution of the measurement process is
reduced. Nevertheless, an insulated probe may
Cair = 249.6/79.78 = 3.1 pF .
be the only solution for a moisture content
measurement in a highly saline soil, and while
For probe A in methanol, C = 94.0 pF, so
the resolution of the measurement may be re-
duced, the accuracy may still be acceptable.
Kmethanol = 94.0/3.1 = 30.3 as measured by
To document the requirement to insulate at
probe A .
least one probe in order to make meaningful mea-
surements in a saline soil, a set of test probes was
The following is a statistical summary of the en-
fabricated using two types of insulating materi-
tire test sequence:
als: a clear polyolefin heat-shrink tubing and a
Box 2
two-part epoxy resin. The following is the perti-
average Cwater = 249.6 pF
nent electrical data for each material:
average Kwater = 80.5
Polyolefin
Dielectric constant at 1 MHz: 2.5
average for Cair = 3.1 pF
average Cmethanol = 94.3 pF
average Kmethanol = 30.4
Epoxy resin Dielectric constant at 100 kHz:
3.8 typical
Box 4
Dielectric constant at 1 kHz:
average Cwater = 238.3 pF
4.1 typical
average Kwater = 81.6
average for Cair = 2.9 pF
Dissipation factor at 100 kHz: 0.01
average Cmethanol = 86.3 pF
Dissipation factor at 1 kHz: 0.01
average Kmethanol = 29.5
Box 3 (unnulled)
Other insulating materials with higher and lower
average Cwater = 336.5 pF
dielectric constants and dissipation factors are
average Kwater = 81.7
commercially available. The above materials were
average for Cair = 4.1 pF
used because they were on hand, not because
average Cmethanol = 95.3 pF
they were the optimum materials for this particu-
average Kmethanol = 23.2
lar application.
The typical performance of probes insulated
Analysis of this data indicates that apparently
with the polyolefin and the epoxy are shown in
the cable correction for the distilled water mea-
Table 3, which shows the requirement for insulat-
surements for box 3 (unnulled) was too small,
ing the probe to make even modestly accurate
336.5 pF after cable correction, as compared with
measurements in saline soils. This requirement
8
Previous Page
