the SPS cycletime

    Hello,

    1. would somebody tell me the the cycle time of SPS ?
    2. I wrote some lines code to record the sps cycle time, and I learn it is about 12ms. But I have a question I can not understand my codes: when I monitoring the value of $timer[2], why is it not always 12, sometimes it is 0(the initial value of $timer[1] )?
    3. If my codes is wrong, what's the right codes to measure the cycle time?

    Hi irobot,

    the variable display updates on every change of the timer value. So it updates also the display if you set the value back to "0". Also the update rate of a "VariableInfo" is around 100ms. That's the reason why you see sometimes "0" and sometimes "12".
    Write the timer value to an integer variable and observe this integer value. That will work...

    Regards
    Daniel

    The $TIMER variables increment in 12ms blocks.

    The SPS only executes for (if I recall correctly) about 2ms out of every 12ms IPO cycle. A short SPS program can execute in less then that time, so it might execute multiple loops within that time window.

    What I did to get a rough idea of SPS timing was this:
    in a Global DAT file:

    Code
    DECL INT TimeVal [100]
TimeVal[1] = 0
...
TimeVal[100]=0

    Then, in the SPS:

    The Timer value will be saved at a new position in the array every cycle. An analysis of the array's contents should give you at least a rough idea of the SPS timing. Keep in mind, since the SPS is only executed for a short time out of every IPO cycle, the cycle time of the SPS will change with the length of the SPS, or with how the SPS branches. A long SPS might take so long to execute, that a single loop requires more than one IPO cycle.

    I use array to save the time , they are:

    KUKA Conntroller : _________<br />KSS Software : ____________ <br />Robot Model: _____________

    hm... hard to do a "beat frequency" analysis on that, but I would say that your SPS is taking just slightly less than one timeslice to execute. So most of the time, you only get one $TIMER entry per loop, but once in a while you manage to make slightly more than one full loop, leading to the repeated 36 and 108 readings. On the other hand, I'm not sure why there's a skip from 84 to 108, or from 864 to 888. I can only guess that you have a branch in your SPS that doesn't execute identically every cycle.

Advertising from our partners