Hello All,
I am developping a continuous acquisition software.
I use a 6062E PCMCIA card plugged into a DELL Laptop, Inspiron 5160, Pentium 4, 2.8GHz, 512MB , Windows XP, with Labview 7.0.
Number of channels acquired = 3, Scan Rate = 160KHZ (Internal Scan Clock).
Briefly, the block diagram is composed of 3 loops synchronized by Notifiers: Loop 1 contains a state machine which takes into account the main switches actions and intitializes the acquisition, Loop 2 realizes the acquisition and Loop 3 is dedicated to do the display.
When I run the software, I am able to do the acquisition and display the data with no problems; however I noticed that after a short period, the laptop clock starts to shift and I wanted to know if I could prevent this to happen; indeed, I need to retrieve the laptop clock time (As accurate as possible) to create the file names.
----------------------------------------------
Here are some observations after having recorded 37 files, 640MB per file.
NB: The names of the files are decomposed this way: MM-DD-YYY_HH-MM-SS
FILE NB: START: END TIME: DIFF:
1 1-11-2005_21-38-0 9.45pm 7 mins
2 1-11-2005_21-45-2 9.52pm 7 mins
--
--
5 1-11-2005_22-5-33 10.32pm 27 mins
--
--
14 1-11-2005_23-27-27 12.11am 44 mins
--
--
23 1-12-2005_1-6-31 1.50am 44 mins
--
--
32 1-12-2005_2-45-33 3.29am 44 mins
--
I noticed that most of the time, the difference between the Start time indicated in the file name and the End Time indicated in the Explorer was 7 mins, except for the files numbered 5, 14, 23 and 32 (Notice that always 9 files separate the first number from the second).
7 mins is actually wrong as well: I did a simulation using an external timer, and I saw that to create a file without using the display, it actually takes 11 mins and 7 secs. No difference when using the display option.
--------------------------------
Thanks for your feedback.
Nicolas