UM-0085-B09
DT80 Range User Manual
Page 372
RG
To summarise: the DT80 has certain temperature-stable characteristics, whose absolute values are measured and
recorded during manufacture. This process corrects for any differences between individual units. Using these
characterised values, further measurements are taken during operation to correct for differences between the current
temperature and the temperature at which the factory characterisation was performed.
Analog Warm Up Time
By default (
P21=0
), the DT80's analog section will be powered up 50ms prior to execution of a schedule containing
analog channels, and then powered down once the schedule is complete. This minimises overall power consumption.
The 50ms delay allows some time for the analog power supply to stabilise. However, the analog circuits also exhibit a
"warm up" characteristic. This means that readings may vary slightly during the first few minutes after the analog section
is powered up, until the components' temperatures stabilise. This is not normally a concern because:
•
the automatic calibration process ensures that temperature related drift is compensated for
•
in most systems all analog measurements in a schedule are completed within a second or so of each other so
the differences in readings due to the analog warm up characteristic are negligible.
If power consumption is not a concern then the analog section can be kept powered all the time by setting
P21=1
. This
will generally reduce the number of calibration cycles which occur due to analog warm up. Calibration cycles will, of
course, still occur if the ambient temperature changes.
Grounds, Ground Loops and Isolation
Experience has shown that ground loops (sometimes called "earth loops") are the most common cause of measurement
difficulties. Excessive electrical noise, unexpected offset voltages and erratic behaviour can all be caused by one or more
ground loops in a measurement system.
Grounds are Not Always Ground
Electrical grounds in a measuring system can be an elusive cause of errors.
In the real world, points in a system that one could reasonably consider at ground potential are often at different and
fluctuating AC or DC potentials. This is mainly due to earthed neutral returns in power systems, cathodic corrosion
protection systems, thermocouple effects in metal structures, lightning strikes and solar storms. Whatever the cause, the
result can be loss of measurement integrity.
Ground Loops
If grounds of different potential are connected by cabling used in the measuring system, ground currents flows — this is
the infamous
ground loop
. The magnitude of the currents can be from milliAmperes to tens of Amperes, and in the case
of a lightning strike, can be thousands of Amperes. Frequently, voltage drops along cables (caused by these current
flows) are superimposed on the desired signal voltage.
A ground loop can arise when a measurement system has more than one path to ground. As
Figure 158 shows, this can be caused by
•
connecting a sensor to a ground point that has a different potential to the ground of another sensor — a
sensor-to-sensor
ground loop is likely to flow through the return wires of the two sensors
•
connecting the DT80 to a ground point that has a different potential to the ground of one or more of the sensors
or instruments connected to the DT80 inputs — a
sensor-to-equipment
ground loop
•
connecting the DT80 to a ground point that has a different potential to the ground of the host computer — an
equipment-to-computer
ground loop.
In these situations, conduction paths can occur from one ground point to another through the sensor and/or equipment
and/or computer, making measurement errors inevitable (particularly if sensor wires are part of the conduction path).
Figure 171: Some of the possible ground-loops in a measurement system
