Application Note

2 Fluke Corporation Power quality recording and analysis: techniques and applications
Techniques for tracking
trends
Trending tracks power quality
parameters over hours or days.
Power loggers measure param-
eters like voltage, current, or
power and log them over time.
Trend recording is good for
tracking normal power, subtle
changes, and exceptional condi-
tions but may have a limited
ability to catch fast events.
However, instrument makers
have come up with some creative
ways of showing faster events
while allowing recording lengths
of weeks or even months.
Fixed interval logging
This is the simplest form of
digital recording. To set it up, you
choose a time period, or interval,
between readings—usually in
seconds or minutes. The instru-
ment calculates an average of the
rms values during each interval
and stores it in memory. This
technique is useful for tracking
changes longer than the logging
interval. Unfortunately, a very
short measurement interval will
catch fast events, but will also
use memory quickly. So even
though fixed interval logging is
easy to set up, it can’t capture
fast events over hours or days.
Min/Max/Avg logging
This technique is similar to fixed
interval logging since it uses
a preset interval. But instead
of taking just one reading per
interval, the instrument takes
many high-speed measurements
over each interval. Processors
within the instrument crunch
through the measurements and
log three numbers for each
interval: a minimum, a maximum
and an average. The min and
max indicate the worst-case,
short-duration events, in some
instruments as short as a single
power cycle. The average tracks
the overall trend. Graphs from
these instruments will often plot
min, average and max on the
same graph.
Automatic time compression,
TrendPlot
TrendPlot is a logging technique
in some Fluke instruments. It is
a form of min/max/avg recording
in which the timescale auto-
matically compresses whenever
the trend approaches the end
of memory. When the recorder
starts to run out of memory,
signal processors quickly go to
work. They combine adjacent
intervals into a new min, max,
and average. You still get to
see the worst-case measure-
ments and the overall trend. And
because you choose when to stop
the measurement, you automati-
cally get the best time resolution
with the available memory.
3
2
4
10 min.
10 min.
20 min.
20 min.
30 min.
30 min.
40 min.
40 min.
3
2
4
X
X
X
X
Actual variation over time
Plot using fixed 10 minute interval, extremes are lost
Figure 1. Fixed interval voltage trend.
Figure 2. Min/Max/Avg trendplot.