Multifunction DAQ

cancel
Showing results for 
Search instead for 
Did you mean: 

System clock used to update

The PCI-1588 won't drive your windows system clock.  It has it's own NI-Sync GetTime function that returns the current time.  You can also program it to timestamp your sample clock or generate a synchronized clock.  All of this may be over kill if what Jeff says will work.

 

For more information on NI-Sync check out the user manual and the help.

 

Here are some examples.

  • http://decibel.ni.com/content/docs/DOC-3421
  • http://decibel.ni.com/content/docs/DOC-3362

 

-Josh

0 Kudos
Message 21 of 31
(1,771 Views)
How is the clock updated in a PC?  I mean are interrupts being missed by running a busy Labview program such as mine?  Maybe I can change the BIOS control over APM and change the APM settings in windows so that windows controls the power usage.
0 Kudos
Message 22 of 31
(1,755 Views)

Most computer clocks are updated via the internet, and are synchronized automatically about once a week- this is true for Windows 2000 and later.  I'm not sure how your system is updating without being connected to the internet, but you may find this link helpful.

aNItaB

Applications Engineer

National Instruments

0 Kudos
Message 23 of 31
(1,735 Views)
My computer is not connected to the internet so what clock in the PC is Labview using to update the time?
0 Kudos
Message 24 of 31
(1,727 Views)

Jeff,

 

How is the system clock updated by the PC?  Does it resync to the Real Time Clock regularly when the computer is on or is it based on periodic interrupts?  I am running a Sony Vaio on Windows ME.  Where can I get documentation on how it is updated because I think my updates are taking longer to be triggered as the system clock that times it is slowing down?

0 Kudos
Message 25 of 31
(1,717 Views)

BME

 

The system clock is updated only when Bill Gates gets around to it, or not updated at all if the OS is too busy.  It has no connection to any external clock by itself. System clock is a very low priority in non NT based OS's. Essentially it is there as a convienience for the user.


"Should be" isn't "Is" -Jay
0 Kudos
Message 26 of 31
(1,709 Views)

Well, is this a sympton of a clock issue...the fact that the AO updates take longer and longer as the clock slows down more and more by a few seconds over the 60 minute run?  I think my DAQ card is sending interrupts to the CPU during continuous data acquisition and when it services the interrupts it misses some clock updates.  So the the error accumulates which manifests itself as accumulating lag in the AO updates.

0 Kudos
Message 27 of 31
(1,703 Views)

Sorry but just to confirm, do you need to use the computer's time in this application? Would just letting the DAQ run with it's oscillator and relative time be sufficient?

 

Thanks,

0 Kudos
Message 28 of 31
(1,699 Views)
Since I cannot change anything I am stuck with using the computer's time.
0 Kudos
Message 29 of 31
(1,696 Views)

BME genuis wrote:
Since I cannot change anything I am stuck with using the computer's time.

BME-

If this is the case- then this is the case.  Your only recourse is to live with the effects of reliance on the system clock and publish this as either a known issue or a feature in the read-me file so the user understands what is occuring.  You can get some relief by slowing down any loops that you have so that your code never overburdens the OS. (add "wait for next ms tick" in loops to slow them down)  The OS can use the time between iterations to do its housekeeping chores and this can reduce the clock lag.


"Should be" isn't "Is" -Jay
0 Kudos
Message 30 of 31
(1,691 Views)