Jump to content

JIF

Moderators
  • Content Count

    403
  • Joined

  • Last visited

  • Days Won

    21

Everything posted by JIF

  1. JIF

    WPT link with feko

    Hi This would be rather simple. Simply create the two dipoles and add a port to each dipole. You can then either perform an S-parameter calculation and from S21 deduce the power transfer. Or you could simply add a voltage source on the one port, set the power (you can set that for a configuration) to 0.1 W and add a load (probably your reference impedance load) on the other port. In POSTFEKO, you can then simply plot the power in the load (no additional requests required for that). I would suggest that you have a look at Example B-01, B-02 and B-03 in the FEKO Example guide. The models in those examples are more complicated, but it shows you how the calculation is done. Example A-01 shows you how to create a dipole, but I assume you know how to do that already. ;-)
  2. JIF

    WPT link with feko

    Hello estc, What is the problem? We can help you, but you need to indicate what the problem is. You need to be more specific w.r.t. what you are not able to do. What have you tried thus far? Regards JIF
  3. Hello hykr, It is possible to write out the impedance matrix that FEKO creates. In CADFEKO, this option is enabled on the "Solver settings" dialog, "General" tab. It writes out a *.mat file. Reading the file can be done easily in Fortran or it can be converted to ASCII for reading in other environments. See the website page "How to read the *.mat, *.lud, *.rhs files and *.str files" for more details. If you don't have access to this link, let me know so that I can make it available to you. Update: There is a section in the User Guide in FEKO 2018 and later on "How to read MAT, LUD, RHS and STR files". Regards JIF
  4. Hello Is the data plotted on the Polar graph calculated by FEKO? If so, you can also display other quantities on the polar graph, such as the real or imaginary part of the electric field of a component (theta or phi component). Thus, you should be able to get all the information that you are looking for onto the graph and then export it. I have never tried this and thus my comment is based on what I expect should work. Is there a reason why you are not exporting an FFE file and using that as a source at the AR card? That should be much easier, but assume that the data that you are trying to export originates from FEKO or at least is a far field DataSet. If the data that you are looking at does not originate from FEKO and you only have the gain values, then you will have to work that value back to the electric field. If you know what the losses in the model and at the sources were, then you can calculate the directivity and if you then know the power level at the source, you can calculate the magnitudes of the fields. There is no way to calculate the phase. I would try to avoid this route as far as possible. I have provided some ideas, but without know what you are starting with, it is difficult to guide you. The easiest is if the far field was calculated in FEKO, then use one of the native FEKO formats such as the FFE file or even spherical modes (SPH file) as the source.
  5. Hello, Are you running into problems when following the description in the documentation? I can provide an example, but I don't know how a simple example will make it any more clear. Below is an example file that defines (as indicated in the documentation) the values for Theta, Phi, abs(E_theta), phase(E_theta), abs(E_Phi), phase(E_Phi): 75 0 0 0 1 0 105 0 0 0 1 0 75 140 0 0 1 0 105 140 0 0 1 0 It has only two theta values and two phi values (grid of 4 points). Does this example make it more clear?
  6. Hello, The MAT file that FEKO exports is not a Matlab/Octave file. They share the same extension, but that is just coincidence. FEKO 14.0 does not support reading or writing Matlab/Octave files. The FEKO website contains details about the format of files such as that MAT file (update: the utility is shipped with FEKO 2018 and later versions and there is a section in the user guide that covers this) and also has a utility that can convert MAT flles to ASCII files that you can read (MAT files can be read directly if you are in a Fortran environment). I am in the process of updating these utilities on the website, so please let me know if I should attach them here.
  7. So uninstalling and installing it again resolved your original problem? Did you do anything differently in the second installation that could have caused it to start working? You now have a problem where you say that the model should fit in the student licence limits, but in your installation it does not? Are you able to send the model? If not, are you able to send the *.out file so that we can see what limit has been exceeded?
  8. Hello, I'll try to help you find and fix the problem. The error is correct, since FEKO does not currently support non-ASCII characters, but we need to figure out what is causing the problem and then we can find a solution. Could you please send me the following file: C:\Program Files\Altair\14.0\feko\FEKOenvironmentFromSetup.lua Regards JIF
  9. It is also possible to export the impedances in Touchstone files. Here we are assuming that you are referring to the input impedance matrix. If you are referring to the MoM impedance matrix, then the answer is also yes. This can also be exported for advanced usage.
  10. Hello, You will have to provide more details regarding your setup. What does the script look like? Does the remote machine have access to the model file or is it only available locally? You will have to have the model file available on the machine where FEKO is running.
  11. Hello, I'm not sure I understand your request or problem, so I'll give a general comment on mesh refinement and hopefully that resolves your problem. Mesh refinement can be added to wires, edges, faces and regions. Thus, if the port is on a wire, you could consider adding mesh refinement on the wire. If you don't want to add mesh refinement on the entire wire, but only close to the port, then the best solution is probably to add a point refinement (not linked to geometry) in the vicinity of the port. Note that wire ports are automatically refined (by default) a little and thus it should not be required, but if you need to refine it further, I hope the comment above will help.
  12. JIF

    SAR warning

    Hello, It would be great if you could share the model so that we can investigate the problem properly. From the warning it sounds like a near field is being requested on the axis of a segment and thus the field is undefined (in a PEC region). Did you manage to resolve the problem? If not, please share your model or a simple model that also shows this warning. Regards
  13. Hello, I suspect that you have solved your problem by now, but here is my response in case you are still stuck. The first error should be quite clear. You are trying to simulate a model that uses more memory than you have RAM and the solver decided to run out of core (using the hard disk) and then even ran out of hard disk space. Since I don't expect your model to be that large and your PC is probably not that small, it could be that you meshed too fine. Are you able to share the model? If not, check your mesh settings and model unit - there could be a mismatch that is resulting in your model being meshed 1000 times (mm to m) too fine. It is just a guess. The second error indicates that you have extremely large and extremely small triangles in your model. Again, this could be indicating a problem in your meshing. I can't really advise more without seeing the model.
  14. Hello, I can see what is wrong and I'll help you to correct the problem. I'm quite surprised that CADFEKO managed to start and we will investigate this further. The problem is that somehow the ALTAIR_LICENSE_PATH__SETUP is being set to (what looks like) the contents of your license file, but should either be empty or it should be the path to the licence file. If you have any details regarding how this happened, please let me know. The solution depends on how you have or want to setup the licensing for FEKO. Option 1 (best option if you only have 1 HyperWorks installation): Remove line 5 form the file completely and thus the path to the license file is not set at all. Place your license file in the default location. This means that you need to place your license file (altair_lic.dat) in %ALTAIR_HOME%\security. After this, you should have a file %ALTAIR_HOME%\security\altair_lic.dat. You should now be able to start CADFEKO and all the other FEKO components without problems. Option 2: Change line 5 in FEKOenvironmentFromSetup.lua to so that the value of ALTAIR_LICENSE_PATH__SETUP points to your license file. This means that the path, including the file name should be placed between the square brackets. It would look something like (it is just an example since I don't know where you want to place your license file or what it is named): setEnv([[ALTAIR_LICENSE_PATH__SETUP]], [[C:\Users\Kun Li\Desktop\FEKO\ALM1234.dat]], true); You should now be able to start CADFEKO and all the other FEKO components without problems. Let me know if you managed to resolve the problem with one of the listed options.
  15. Hello FEKO does not currently support non-ASCII characters in the configuration files. From the error message it looks like there is a non-ASCII character in line 5 of your FEKOenvironmentFromSetup.lua file. Could you send the file so that we can have a look at it and propose a solution? Thanks JIF
  16. Hello goudakaes I had a look at the model. For some reason the face in that model has strange properties (not something that you can change) that cause the mesher to think the face is in a PEC region instead of free space. I have never seen a bug like this. If possible, could you provide me with the Gerber file so that I can log a bug report. Unfortunately the model does not really help, since we need to figure out how the model ended up in this strange state. I did find a workaround to get the face corrected. I created a rectangle in the same plane as the spiral, that is larger than the spiral. If you then union these two objects and delete the outer face, the resulting face seems to be correct. Regarding the model and what you are trying to solve, I have the following comments: - 125 kHz is a really low frequency for FEKO to solve. FEKO does have low frequency stabilisation (an option that you need to activate) that can work down to sub-Hz, but it only supports PEC structures (no dielectrics allowed). - I noticed that there seems to be an offset (translation) that you added to raise the dielectric cube from the ground and that the spiral was not unioned with the substrate. You need to union to ensure correct meshing and connectivity and I don't think the offset should be used. - If this model needs to be simulated at a higher frequency (since 125 kHz is too low for FEKO given the use of a substrate), I would recommend that you change the model in one of the following ways: A ) Remove the finite dielectric substrate and model it as an infinite dielectric substrate (planar multi-layer substrate). You can then use the microstrip ports that you already have. B ) Use the finite substrate, but then add metal strips down to the ground and add an edge port (not microstrip) inside the dielectric substrate (not on the boundary).
  17. Dear Tuan Nguyen You are correct, for RCS you need a plane wave. See the equations in user manual (20.6 Far fields and receiving antenna) and note that these assume a plane wave excitation. It sounds like what you are trying to calculate is the reflection from an object due to some source, other than the original plane wave. If the object(s) is far from the source, you should be able to approximate the source as a plane wave instead of using the actual source (if they are well separated, only the field in the direction of the object will be seen by the object). If they are not, I'm not sure that RCS (a far field concept) is the correct way to investigate the reflection effects that you are interested in. Without understanding more about what you are trying to do or investigate, it is quite difficult for me to make suggestion for your simulation. Regards JIF
  18. Hello David, This is not possible with the current version that is available to users, but this feature will become available with the next feature update to FEKO 14.0, nl. 14.0.410. We expect it to be released in a few weeks (2 to 3 weeks). Once you have access to FEKO 14.0.410, you can use the new ExportIsoSurfaceToSTL(filename) method on the NearField3DPlot object in the API to perform the export. This feature will not be available in the GUI. I hope that helps and that you can wait a few weeks for the feature. Regards JIF
  19. Hello Abdullah, It looks like you almost have the full solution. Currently you have a single plane wave with a single polarisation and you add two far field requests. Let me assume your plane wave is currently vertically polarised, you thus currently calculate VV and VH. What you are missing is a second configuration where the plane wave source is horizontally polarised (with the same far field requests) so that you can calculate HV and HH. I'm assuming that you use CADFEKO to create your model (not EDITFEKO) and thus you should be able to modify your model by performing the following steps: Duplicate the configuration that you already have Set the sources to be "per configuration" (it is an option on an icon that is a gear with four colour blocks on it - next to the + icon on the configuration tab) Select the second configuration, open the source dialog and set the polarisation to 90 (assuming it is currently the default 0) The second configuration's far fields will give you the HV and HH polarisations. Regards JIF
  20. I think this question then becomes something that some of the admins of the forum needs to answer. I can simply state what I have done and what I'm seeing. I have enabled email notification in my profile - I set it to immediate and then also read through the long list of options and selected the options that I think would work best for me (I was quite liberal with the notifications, at least initially). I also only followed the FEKO forum and that is why (I think) I received your question (this topic) via email. I replied twice and did not receive notifications - I consider this acceptable for me. This morning, I again received an email with your follow-up question. For me, the noticiation works as I would expect. I hope this helps. Did you check your spam folder? Maybe some of the others would like to indicate what notifications they received (or did not receive)?
  21. I should probably also mention that you can (and should) follow forums, such as FEKO, and also individual topics. If the chatter on a specific topic becomes too much and you are not interested in the topic, you can also stop following a topic. The follow/unfollow buttons are also at the top right hand side of the web page.
  22. Hello Derek Each registered member of the Forum can determine how (and if) notifications should be done. These settings are available when you edit your profile (click on your name at the top right hand corner once you have logged in).
×
×
  • Create New...