web
Vous êtes hors connexion. Ceci est une version en lecture seule de la page.
close
  • AFG3102C - latency of SCPI commands received over ethernet

    I wrote the post below about my embedded system sending SCPI commands to the AFG3102C through ethernet using the vxi11/rpc protocols.

        https://my.tek.com/tektalk/signal-gen/8582d75b-a5ef-ed11-a81c-000d3a5cffcf#86b7fc57-a5ef-ed11-8849-000d3a314d17

    I was able to get this communication working. Generally, I followed the instructions at these websites:

        https://github.com/applied-optics/vxi11
        http://optics.eee.nottingham.ac.uk/vxi11/

    So now my embedded system can send arbitrary SCPI commands to the AFG3102C to adjust the settings.

    My next issue is with the command latency. I would like to be able to update the AFG3102C settings around 20-30 times per second. 

    I have a trigger signal that goes to both the AFG3102C and to the embedded system that sends the SCPI commands to the AFG3102C.

    When the trigger signal goes from high to low, I want:
     
    • the AFG3102G to produce on the output a pulse according to its current configuration
    • the embedded system to send a command to the AFG3102G to adjust the AFG3102G configuration for the next trigger cycle. I'm typically updating the pulse delay using SOURCE:PULSE:DELAY.
    If the trigger happens 20 times a seconds, that's 50 milliseconds between each trigger.

    Is 50 milliseconds enough time for the AFG3102G to receive a SCPI command and make the adjustment in time for the next trigger?

    The embedded system is fast and responsive, it is able to transmit the SCPI command over ethernet within 1 millisecond of the trigger. So the AFG3102G should have nearly the full 50 milliseconds to process the command.

    In my testing so far, I can update the AFG3102G using SCPI commands only around 2-4 times per second. If I go faster than this, the command does not seem to take effect.

    Does this sound right?

    I am assuming that if the AFG3102G receives a SCPI command after the trigger happens while it is outputting the pulse, the change to the configuration will not affect the output of the current signal being produced.

    To note, the triggering of the AFG3102G output at 20-30 times per seconds works fine. I just can't change its settings using SCPI commands sent over ethernet at this rate.

    If the AFG3102G cannot be updated over ethernet 20-30 times per second, is there another TEK waveform generator model that can do this?

    Thanks for any help.
  • AFG3102C control using raw sockets

    I am using an AFG3102C waveform generator to test an embedded system.<br> <br> The AFG3102C is set up to generate rectangular pulses when triggered externally by a signal from the embedded system.<br> <br> There is a short delay configured between the trigger signal and when the pulse is output. This delay is configured using the AFG3102C front panel.<br> <br> The delay needs to be adjusted at various times while the embedded system is running.<br> <br> The timing of the adjustment needs to be somewhat precise and it needs to be made from the embedded system.<br> <br> Since the embedded system is running a real time operating system and not Windows, I cannot use TekVisa via C# or python to communicate with the AFG3102C.<br> <br> It does have ethernet though so I was thinking that I could send the SCPI commands defined in the programming manual for the device though a raw socket.<br> <br> For example, it looks like the delay could be set using the [SOURce[1|2]]:PULSe:DELay command.<br> <br> Is it possible to control the AFG3102C using raw sockets and the text SCPI commands? If so, what port is used?<br> <br> If not, are there any options for using VISA or vxi-11 on an embedded system with a real time operating system?<br> &#160;