chart scale error

[Torten]

I am coding a logging system in labview 7.1

On the front panel I have several charts. If I try to change the x-scale by typing new end of scale value while running - labview crashes (LV7.1 has encountered an error, please tell Microsoft...)

If I change the x-scale when the program is not running - labview sometimes crashes.

The graphs are indicators OWGraph, Wave Height, Temperatures, Flow and in the lower loop, pressure difference.

The data comes from on demand aquiring from a pci6120 card which runs in one loop - the displaying is handled in another loop (on top in posted code image) except from one chart which needs quicker updates.

[LuI]

Torten,

this should not happen!
But I had seen some misbehaviour too. What usually helps is to drop a fresh chart onto the FP and configure it similar to the one causing the problems. Finally disconnect the problematic one from the BD and connect the new one instead. After having re-linked all Refs and locals, you just delete the old one and move the new one onto thats position.
Sounds as a whole lot of work, but usually this can be done in just a few minutes. AND it usually helps.

You know, LabVIEW has to maintain a lot of lists with objects and properties and data and more. Sometimes there is a little error there that might cause such kind of problems.

Greetings from Germany!
--
Uwe

[Torten]

I found the sinner, it cant handle the offset. I use an offset of the number of seconds since 1904... to display the current time on the x-axis. If I plot two or more plots in the chart and use property node-> x-axis.offset it crashes every time I try to manually change the x-axis. I used most of the day figuring this out and the rest of the day to find a workaround. Since I dont use the date - I ended up calculating the smallest number of seconds to put into offset so that the time got right. Because if you wire a constant (any size) into the property node it is no problem - if you wire a time stamp converted to double - it crashes. When I use a probe on the value, it displays it in scientific mode (3,2 e+9) I wouldnt think this is any different than a constant written 3200000000 - but it is the only explanation I have right now - and I am not using one more second to find out - cause my workaround works...

In the code attached - the timestamp is input to a chart with a single plot, that however causes no problems...