[Doc]

The vertical axis labels need to have 3 digits to the right of the decimal point instead of the current 2-digits.  As it is I have seen labels of (0.00, 0.00, 0.01, 0.01, 0.01, 0.01) which doesn’t correspond to any linear scale and I have no easy way to decypher what is displayed. The third label seems to correspond to 0.079 based on the value of the marker at the right edge of the display and the forth increment seems to be 0.0104 so perhaps the need is for 4 digits or consistent with what is shown elsewhere on the graph.

[Doc]

Another alternative would be to reuse the scroll bar that is implemented for the Home page for the Graphs page. It is not elegant but it is functional and it supports acceleration by mouse movement and incremental scrolling by clicking on the directional arrows. It would benefit from an ability to adjust how much time is displayed on the graph perhaps over a ratio equal to the scale difference between sequential time scales (Sec would be able to display a full minute of data when maximally compressed, while Minutes would display a full hour, Hours would display a full day, day might be able to display a full month, etc.  Where you are in the record is visually intuitive.

[Doc]

Another enhancement along this line would allow (on the Web App) for a double click on a particular Day (Week, Year, Hour, Minute) to switch to the next finer available resolution so that we can drill down where there is visible data to get a magnified view.  To do this on the mobile app is not so easy and until you commit to allowing the Web App and the Mobile App to diverge it is probably not doable but it is an alternate approach to focussing at a finer level of detail.

[Doc]

Hi all – Doc here. New user struggling to learn and discovering small things that should be fixed and generating questions along the way.

One thing I noticed and didn’t find it covered by a pre-existing post was how graphs scroll on an iPhone when Seconds are selected as the units of horizontal scale. The data scrolls in from the Right but the horizontal axis labels don’t scroll – so the data is always misaligned with the axis.  Seems to work much better at other scales (Min, Hr, Day, …).

The way the axis numbers scroll when it works seems to be that the algorithm rewrites the numbers to line up with the data rather than scrolling the whole frame like a picture which can be made correct by construction.  It is probably possible to write the frame as an image outside of the viewable space and allow the user to slide what they want horizontally into the viewable space. You just have to stop creating new image when the data source stalls.  Don’t care how you fix it but the current UI is really hard to use if you need data alignment to the second.  I know I can get what I want by manipulating the data after downloading it as a CSV file but that is neither timely nor trivial and not what I expect from a good UI.

The other observation is that the time tags on exported data is better than I can measure with a good watch, so congratulations to the master of time at Emporia.  Other unnamed sensor vendors don’t do as well.

I did find a comment about the transition to DST which is a common issue, but is there a way to capture data that is collected as we “fall back” to standard time in November before it is overwritten by the new data with the same time tag (like collecting data for 01:06:29 the first time it occurs but downloaded after the second occurance of the same time tag an hour later). I know that I can’t expect 1 Second resolution data to be available for more than 3 hrs but that is enough (unless it is overwritten after 60 min).

• This reply was modified 1 year, 4 months ago by Doc.

[Doc]

So long as you are combining circuits that are on the same phase you can run multiple branch feeder wires through a 50A CT in the same (correct) direction (you can easily fit 3 #12 THHN wires though a 50A CT though you do need to make sure that you are not exceeding the capacity of the CT), the sensed current is proportional to the sum of the branch currents and so long as the total is less than the saturation current of the CT you are golden. The article referenced above is not helpful in analyzing this case.

• This reply was modified 1 year, 4 months ago by Doc.

[Author]

Posts