> The x-scale is always 01/01/1904 even after I selected the 'absolute
> time' option from waveform chart properties. If I add 2004 as offset
> of x-scale as suggested by the previous post, the waveform of the
> signal cannot be displayed, the system think it is still running in
> 1904.Could you give me some further help?
>
Your description doesn't give enough information to pinpoint the
problem, so I'll list off a few things that could be going on.
The graphs and charts accept many datatypes. If you are using the
waveform, display it on a regular waveform indicator and see what the
timestamp and deltaT are. If these are correct, then the graph should
be able to display things with no further changes except that you may
need to turn on X Scale Autoscaling so that it will zoom and shift to
show the plotted data. Charts will shift, but by default will not zoom
in, and it might make sense there sometimes too depending on the delta T
and amount of data plotted.
If the timestamp on the waveform is incorrect, then it is much easier to
construct the properly marked waveform than make the graph add to it to
label the waveform's time differently.
If you are not using the waveform, then you may be using a cluster of
X0, dX, and Y array wired to the graph. This is very similar to the
waveform, and again, it is better to add offsets to get the X0 to
display correctly on a numeric. You can extract this number or display
the cluster in an indicator, then format the numeric to display absolute
time.
If you are simply wiring the Y array, then LV will construct the X
values based upon the number of points beginning at 0 and incrementing
by one by default. When displayed as absolute time, this will default
to midnight Jan 1 2004 plus an offset based upon your timezone. In this
case, since there is no X information being piped into the graph, it is
better corrected by going to the graph properties for the X Scale and
setting the X Offset to the seconds between now and 1904. This is a
constant you can compute when your program starts up or you can use a
calculator, whichever is easier. You will also likely want to set the X
Multiplier to the deltaT between successive points. 0.001 for 1KHz, 1
for 1Hz, etc.
Finally, if these suggestions don't make sense or don't help, you might
want to explain more about what doesn't work.
Greg McKaskle
