#146 AC Simulation Error: checker error, a .DC action is required for this circuit definition

0.0.19
open
nobody
None
5
2014-09-17
2014-02-28
BrendaEM
No

Hi,

Myself and 2 others have a problem with the AC simulation, which reports, "checker error, a .DC action is required for this circuit definition (accounted 2 non-linearities)"

BTW, this otherwise looks like a great application! Thanks!

Windows 7/64Bit and 7/32Bit, perhaps others.

Ref: Thread at https://sourceforge.net/p/qucs/discussion/311050/thread/167b7f5f/

1 Attachments

Discussion

  • BrendaEM
    BrendaEM
    2014-02-28

    I got a simple low-pass filter to work with AC Simulation, so the system seems to be set up correctly.

    However, the attached file produces the same error. Because the Pot is an Verilog part, which omits the word "Ohm" in the parameters I tried it with and without.

    I also tried clearing the .qucs (temp) folder.

    I will try replacing the pot with resistors in a voltage divider configuration.

     
    Attachments
  • Guilherme
    Guilherme
    2014-02-28

    For the .DC error, please add a DC simulation to the schematic...

    Anyway, what is the expected behavior for your circuit? The output looks more like a high-pass to me (attachment)

    Have you tried other examples involving AC simulation? http://qucs.sourceforge.net/examples.html

     
    Attachments
  • BrendaEM
    BrendaEM
    2014-03-02

    Yes, I have it working now. Thank you.

    However, I find it a little unintuitive to need to add both simulations at once. I wonder if the need for this could be better documented in the tutorial or in the error code, or if the AC simulation could call the DC once behind the scene. I wonder how many people will be stumped by this, and will also post bug reports.

    So far, I am really enjoying and are grateful for Qucs. : )

    The circuit is a passive Baxandal tone circuit, which can go quite flat if both pots are set in the middle. It is quite an old circuit, so it can be included as an example. In the version I did, I added a sweep on both pots using 5 steps.

     
  • BrendaEM
    BrendaEM
    2014-03-02

    This can be included with Qucs.

    (I do with the scientific notation truncation point was adjustable)

     
    Attachments
  • BrendaEM
    BrendaEM
    2014-03-03

    (I do (wish) the scientific notation truncation point was adjustable)

     
  • schubi77
    schubi77
    2014-07-26

    I think you did a rather good job and thanks for this really great
    program, but, Yes, I agree to the remark of Brenda.
    If I work with OpAmps I do not put any Simulation DC on it, just
    a symmetric or asymmetric Power supply (which I drawed with two DC
    Voltages 15V, symmetric and grounded and connected to Vcc and Vee).
    In a simulation with AC (for example frequency testing) there is
    absolutely no DC wished. Maybe if you want to look for the reaction
    of your circuit with DC on his input than you can use this function.
    But why does the system handle this as an error, if I have no DC simulation
    and forces me for some rather unlogical additional steps during schematic drawing.
    And its rather frustrating if you designed your circuit and you know
    (40 years of electronics) there is all done what you had to do and
    then the system stops with any DC errors...

     
    Last edit: schubi77 2014-07-26
  • Guilherme
    Guilherme
    2014-08-04

    Whenever it encounter non-itineraries a DC simulation is required.
    Perhaps we change this error into a warning.

     
  • Guilherme
    Guilherme
    2014-08-06

    In this case the error is:

    checker error, a .DC action is required for this circuit definition (accounted 3 non-linearities)
    

    The check_netlist.cpp code issues the error. We could work around the error. Adding a .DC action on the parsed tree and create a warning instead.

            // count analyses requiring a DC solution
            a += checker_count_definitions (root, "AC", 1);
            // check dc-analysis requirements
            c = checker_count_nonlinearities (root);
            n = checker_count_definitions (root, "DC", 1);
            if (n > 1)
            {
                logprint (LOG_ERROR, "checker error, the .DC action is defined %dx, "
                          "single or none required\n", n);
                errors++;
            }
            if (a >= 1 && c >= 1 && n < 1)
            {
                logprint (LOG_ERROR, "checker error, a .DC action is required for this "
                          "circuit definition (accounted %d non-linearities)\n", c);
                errors++;
            }
    

    Similarly, more than one .DC is also an error.

    Perhaps we can handle such situations more gracefully. Needs work...

     
  • Guilherme
    Guilherme
    2014-09-17

    • Group: 0.0.17 --> 0.0.19