I think that best way so to deploy my article and thoughts this is by adding first
Using Trend-plot description of our User Guide ( By Fluke ).
TrendPlot provides a visual representation of the measured signal over time. About three-quarters of the Meter’s display is used to plot the maximum and minimum readings vertically, while the horizontal axis represents time.
The vertical and horizontal axis’ are not calibrated, and only represent relative time and amplitude depending on the input signal. Now, as vertical separators they are
Five vertical lines of four dots its one.
Horizontal Spacing between its separator to the next this is 50 pixels and 200 pixels in total (as horizontal length).
First separator this should be named as Zero point.
Second and third separators they should be named as compressed storage area.
Fourth and Fifth separators they should be named as uncompressed storage area.
Now according to Fluke, Compressed graph (at Second and third separators) this gets actually compressed by retaining only positive and negative spikes, this translates that vast majority of average values they get discarded.
Trend-plot sampling : This is at interval of 1:1 plot sample per second and this is as such just for the first run and before
a first time compression occur.
Trend-plot concept : By software implementation of sample points collection and review of plot marks.
Trend-plot Mathematical concept : Trend-plot this is an visual representation of Min/Max recorded values, handheld multimeter they do that too by the use of small quantity of memory and they can record a)Minimum b)Max and c) Actual measurement and some more advanced DMM they include and d) Average measurement too.
What is the most well implemented Trend-plot at a graphical multimeter? Well GDS-310 Oscilloscope and multimeter this has the most successfully implemented Trend-plot because this is totally uncompressed, but this is possible because the user he can select a fixed duration of time from seconds to minutes or hours.
Practically there is no uncompressed Trend-plot, what there is actually this is two types of Trend-plot, GDS-310 this has uniformly compression at entire screen but when recording time exceeds User choice of time scale, all first recorded graph values they get erased and what you have this is all the action recorded within this specific selected time window.
Such Trend-plot implementation works best so to review for example a thermostat operation in a refrigerator, in a period of three hours it will be nicely recorded the timing in which a thermostat activates and deactivates by us measuring temperature due a K Type thermocouple.
Best scenarios so to use Trend-plot at Fluke 8845A DMM in your benefit? Trend-plot this is a useful secondary in value tool (when this gets compared to detailed data logging tool) when we need to inspect stability over time of measured value.
For example I did inject 10mA of current at 400mA range of my 8846A and with active Trend-plot I did discover that 8846A
resistor current shunt this was needed approximately 30 minutes so temperature coefficient to fully stabilize.
This translates that when you need high accuracy at Low current 400mA range, you will have to leave the DUT working for 30 minutes prior measuring anything. Trend-plot with fixed sampling at one time per second this is not the best tool so to record inrush current.
Therefore best use of Trend-plot at Fluke 8845A, this is for a first evaluation of slight changes over time, best example here, this is battery charging or discharging curve.
For example a friend gave you a used NiMH 1.5V AA, battery condition unknown, you do charge this battery and then discharge it by monitoring voltage at Trend-plot.
By you constantly monitoring voltage and time duration, you may come in to a first conclusion of this battery health status.
Therefore 8845A Trend-plot this is mostly a
electric power monitor, for a quick visual analysis / DUT behavior basic review.
Best scenarios so to use Trend-plot at Fluke 8846A DMM in your benefit? Fluke 8846A has a higher quantity of measuring range regarding DC Low current, among with the ability to measure temperature exclusively by the use of Pt100 & Pt1000.
Evaluation of DUT power consumption when an electronic circuit this is at Sleep-Mode this is possible.
Additionally you may use Trend-plot for temperature measurements at applications that a hour of monitoring this is enough, Trend-plot graph plots compression as is implemented at 8845A/8846A this becomes less useful as data logger review tool, at higher logging time duration.
How ITTSB Blog evaluated Trend-plot 8845A/8846A behavior and capabilities ? Well I was had to monitor Trend-plot activity for one hour, and to take more than 30 pictures with my digital camera.
Very soon I did realize that plot compression this stealing away 50% of Dot matrix display total length and the other 50% this now it does
Live plot at 100 available pixels, now when these 100 available pixels get full of data then Trend-plot this updates and compress again, so these 100 available pixels to get free again.
I did manually a measurement of pixel counts between these vertical separators, this is how I got awareness of 50 pixels spacing.
Such a 50 pixels spacing this translates that Min/Max Plot this it does require a minimum of two pixels (in series), first pixel will plot a Max value and it neighbor pixel will record a Min value, therefore we have actually 25 recordings in to a 50 pixels spacing.
I did also timing recording of first four Trend-plot record / compress cycles. Execution times was (in minutes):
1) 00:01:33
2) 00:03:03
3) 00:06:15
4) 00:12:30
This translates that while the clock counting in seconds, plot sampling interval this get double (slower) at every new cycle of operation.
Therefore you can not use recorded graphical representation so to define exact time which an event was occurred, except if you personally record on paper (take a note) of the time that this event it just appear = Got plotted on your screen.
Yes now it makes sense of why Fluke talks about of not calibrated Trend-plot, with so many undocumented variables, a visual representation of the measured signal over time this is all that you get.
Pictures below they represent:
a) Trend Plot fully expanded and before a next time compression occur.
b) Trend Plot compression just occurred and this now starts a new cycle.
c) This is my souvenir at the duration of 60 minutes.