UM-0085-B09
DT80 Range User Manual
Page 326
RG
Measurement Timing
When an
FW
channel is evaluated, the measurement process is as follows:
1.
The pluck circuit begins charging. The
MD
timer starts here also.
2.
After about 100ms, the pluck circuit releases its energy in the form of a narrow high voltage pulse. Inside the
sensor this causes the wire to start vibrating. The resonant frequency will typically be in the range 500-5000Hz.
3.
Once the pluck is complete (about 0.2ms), the DT80 disconnects the pluck circuit and begins listening to the
sensor. Inside the sensor the wire's vibrations are sensed and a corresponding electrical signal is generated.
4.
Over the next 100-200ms the DT80's phase locked loop (
PLL
) locks on to the fundamental frequency and filters
out noise and harmonics. During this time the amplitude of the signal gradually decays as wire’s vibrations
decay.
5.
Once the
MD
timer expires the DT80 will begin measuring the frequency of the filtered signal.
6.
Once the required sample period (gate time) has elapsed, the DT80 reports the measured frequency value.
It is important that the measurement delay is set such that the incoming signal is stable and of adequate amplitude for
the duration of the measurement period. The default value is suitable for most gauges, but in some cases it may need
adjustment, as discussed in
Measurement Delay and Sample Period (P326)
.
Troubleshooting
Listening to the Gauge
The DT80G/85G GeoLoggers provide a headphone output socket which can be useful in diagnosing problems, given
that the sensor’s vibration falls within the audible frequency range (500-5000Hz).
To check a gauge, connect headphones or speakers, then enter the following:
P21=1
this keeps the DT80's analog section powered after the end of the measurement
P62=1
this maintains the multiplexer settings after the end of the measurement
1FW
this samples the VWSG connected to, in this case, channel 1 + and – terminals.
Each time you send the
FW
command you should hear a clear "ping" sound which decays over a period of a few
seconds. If not:
•
If there is no sound or only random noise, double check that you entered the correct channel number. Check all
connections. Check the resistance of the gauge by connecting it to the * and # terminals then entering
1*R
several times. These 2-wire resistance measurements should return stable values. If not then a cable or gauge
fault is indicated.
•
If a ping can be heard but it is faint or buried in random noise, then the cable is too long or is "leaky", or the
gauge sensitivity is too low.
•
If the ping is not clean and pure, then the gauge is possibly faulty. The gauge may have been mechanically
damaged during installation.
•
If you can hear a low frequency hum, then noise pick is a problem. If the gauge is placed near a transformer,
electric motor, high current power cables, etc, then relocate or reorient the gauge for minimum pickup. Ensure
that the cable is shielded to prevent capacitive pickup. (Connect the shield to
DGND
or the DT80 chassis earth
point).
Be sure to reset the
P21
and
P62
settings when finished.
Measurement Delay and Sample Period
If a strong and clear signal is heard, but the frequency measurements are unstable (variations of 10-20Hz or more) then
there may be strong harmonics present in the gauge's vibration. Because harmonics usually decay faster than the
fundamental, it will often help to increase the
MD
setting. This will mean that the actual frequency measurement phase
starts later, at which point the amplitude of the harmonics should be less. If the measurement delay is increased too
much, however, the overall signal amplitude may decay below the noise level.
If the signal is clear but decays rapidly then the default
MD
setting may in fact be too long – by the time the measurement
completes the signal has decayed to nothing.
Some trial and error may be required to find optimal settings. The recommended procedure is as follows. For each step
perform several measurements in order to gauge the stability of the readings.
1.
Start with minimum values for measurement delay and sample period, e.g.
1FW(MD150,30)
2.
Increase the
MD
setting in, say, 20ms steps until stable readings are obtained, then go one step further.
3.
Further improvement can usually be obtained by progressively increasing the sample period so that the
frequency is measured over a longer time interval. If this is increased too far however then the signal amplitude
will descend into the nose and readings will get rapidly worse.
