Jump to content


  • Content Count

  • Joined

  • Last visited

  • Days Won


Everything posted by mel

  1. mel

    Warning 38142

    Why not run the simulation at 10 kHz? As the out file states, the frequency supported starts from 1e4 Hz (10 kHz), and your measurement device can measure up to 10 kHz.
  2. mel

    FEM meshing

    This is the Altair Feko forum therefore a FEM mesh refers to a tetrahedral mesh where the EM fields are solved. That does not sound right. If you cannot attach the model, please ask your study leader for Feko support contact details in your country.
  3. We strongly recommend downloading the latest Feko. Your license should be valid for this.
  4. If you want to connect the cable to geometry, you should set the solution method on the Cable Harness properties to Method of Moments (MoM). In addition, you will require a valid triangle mesh vertex at the connection point. One way is to just use Imprint Points, snapping to the two ends of the cable path. Note also that direct connections across the shield are not supported. See the suggested circuit diagram modified at connector 2 in the attached model. I have not checked the CISPR document to see if this is consistent with what should be modelled. A polyline refinement on the mesh is also recommended below the cable, and this refinement should be made finer based on the distance to the surface below the cable. Tisch_mit_netzgeraet_last_leitung___ALT1.cfx
  5. It would be best that you contact Antenna Magus for support directly.
  6. mel

    Litz wire

    You would have to model the wires physically, i.e. drawing multiple lines (which will mesh into wires). Depending on the length and number of wires, it could turn out to be computationally expensive. Note that the mesh length for the wires should be similar to the spacing between the wires. Also note that the wire segments only model axially flowing current. If there are secondary effects such as currents varying across the cross section of the wire, this effect would be ignored. My understanding is these Litz wires are used at low frequencies. I would recommend that you also post this question in the Flux forum here.
  7. Please post your query in the Winprop forum here.
  8. This is not possible in CADFEKO, however the surface area will be given in the *.out file near the beginning of the *.out file. "Surface of all triangles in m*m:" This will obviously be the surface area of the mesh, so be careful if you have curved surfaces meshed very coarsely. You don't have to run the solver fully to get this output. You can use the check-only mode for the solver.
  9. You don't have any metallic (PEC) faces in the model, they are all dielectric surfaces. Have a look at the ExampleGuide models "Example-A09-Circular_Microstrip_Patch_with_Coupled_Feed" and "Example-A10-Aperture_Coupled_Patch_Antenna" that will show you how to do this.
  10. This may be a difficult one to answer...on the Altair Feko forum.
  11. The colouring according to face medium will show PEC for coatings since essentially it's still a PEC face, just coated. However if you set the face medium to a layered dielectric, then the face is a dielectric face and colouring will be according to the layered dielectric color. To confirm the coating, use the Highlight option.
  12. The near field, sampled basically at the source, will include the contribution from the source, which will dominate. You should enable the option "Calculate only the scattered part of the field" for the near field. The OUT file shows the coating is applied, not sure yet why POSTFEKO has PEC coloring. Label Rectangle1.Face2998: Electrically thick surface coating Thickness of surface coating: 9.00000E-04 m Layer = 1 Eps_r = 6.450E+00 Sigma = 3.830E+01 S/m tan(delta) = 1.386E+00 Mue_r = 1.000E+00 tan(delta_mue) = 0.000E+00
  13. You mention it is a simple model. Could you please attach the cfx?
  14. mel

    Limits of FEKO SE

    Hi Shaiksa Your question is off topic. This topic has to do with Student Edition limits. Please create a new question by clicking "Start new topic"
  15. The Green's Function for infinite substrates uses interpolation tables. Over frequency ranges the algorithm used is changed at some point over the range. A small glitch could be observed at a specific but narrow frequency range. Adaptive frequency sampling will "zoom in" on this glitch and appear to amplify the glitch. If you rerun with linearly spaced discrete points the results should be the same with or without the far field. I see that you are not a University user. If you are a commercial user please contact your local Hyperworks representative for further support.
  16. For an array of impressed sources in CADFEKO, script record the setup of 1 element and then modify the recorded script to use a FOR loop. Any changes made in EDITFEKO modifies the .pre file which is a file that CADFEKO cannot read back in - CADFEKO only outputs this file. So changes made in EDITFEKO can only be viewed in POSTFEKO which reads the .fek file. To quickly see the changes made in EDITFEKO, run PREFEKO. PREFEKO makes the *.fek file which can be viewed in POSTFEKO without actually running the solver. It is a bit of hack job but some optimisation can be done after changes in EDITFEKO. The optimizer uses the labels of e.g. the farfield, voltagesource, nearfield, etc. For example, the voltage source on a wire (usually written to the .pre file by CADFEKO) appears in EDITFEKO as follows: A1: 0 : Horn.feed.Port1 : 0 : : : 1 : 0 : : : : 50 ** VoltageSource1 The label to optimize in CADFEKO is then VoltageSource1 which would have to be manually typed into the Focus source label box in CADFEKO. If there was a far field request, this would be in EDITFEKO: FF: 1 : 181 : 1 : 0 : 0 : -180 : 0 : 2 : 0 : : : 0 ** FarField1 and the label would be FarField1 or you can use any of your liking.
  17. Here I have set a local mesh size on the faces of the small cuboid of 1/5th of the face length or 0.18/5.
  18. This is not directly possible in Feko yet. I am aware of some users applying a correction factor K to the ideal conductivity based on the Groiss formula. See here.
  • Create New...